By on February 1, 2010

Update: a portal to all of TTAC’s articles on the subject of Toyota gas pedals is here:

Toyota uses two different electronic gas pedal designs in its cars. The version built by CTS (lower) is the subject of a massive recall, and the 2.3 million units in affected Toyota cars are to be “fixed” by the insertion of a steel shim. This CTS design is also being modified for new Toyota production, currently suspended. To our knowledge, Toyotas built with the other design (by Denso, upper) are not subject to any recalls or NHTSA investigations,. We have spent the last two days tearing down both units, and familiarized ourselves with their designs, reviewed Toyota’s “shim fix”, and replicated the fix ourselves. Toyota’s planned fix will undoubtedly reduce the likelihood of sticky pedals in the short term, but after examining both units, we are convinced that the CTS unit is intrinsically a flawed design, and poses safety risks in the long term, even with the fix. The only right action for Toyota is to acknowledge the long history of problems with the CTS-type unit, and replace them all with the superior Denso or another pedal unit that lacks the intrinsic flaws of the CTS design.

Before we briefly review the key design differences, we must acknowledge that Toyota is ultimately responsible for both designs. CTS has stated that its product was built to Toyota specifications. What we don’t know (or understand) is why Toyota has two such fundamentally different units in production. Is one unit cheaper to build? Or was CTS tooled up to produce its unit because of other similar units it builds for other manufacturers? What we do know is that the CTS unit has been used in Toyota products since 2005, whereas the Denso unit has been in use since well before that time. Toyotas sold in Europe are also subject to a similar recall, and based on the description of the issues and the unit, it appears that it is the same or similar design as the CTS unit, but we do not know if it was built by CTS or another supplier.

The key component in question is the friction arm of the CTS. It is both essential and desirable to have a certain defined degree of friction in these electronic gas pedal assemblies. The amount of friction is designed to be some degree less than the return spring, so that when the pedal is released, it returns to the closed position. But the friction (hysteresis) makes it easier to maintain a steady throttle setting, and relieves strain from pushing against the spring continuously. It simulates the intrinsic friction that is present in the traditional throttle cable as it passes through the cable housing.

The two units generate the desired degree of friction in very different ways. In the Denso unit (above), the return spring (steel coil) is squeezed on both sides of its housing. It rubs against the plastic housing as it compresses, which generates the desired amount of friction. Both sides of the full length of the Denso coil are in continuous contact with the rubbed are, spreading out the contact area size. And the metal to plastic interface seems to be relatively unproblematic.

The CTS unit is a fundamentally different design. The friction is generated by two “teeth” (A) that extend from the friction arm, and ride in two grooved channels of the housing (B). The friction arm is an extension of the pedal itself, and moves as the pedal is moved. Both the friction arm, its teeth and the surface it rubs against are plastic. Notice the small area of contact (dulled gray spot on tooth). This is the fundamental source of the problem with this unit, and one that Toyota has not come clean about. The friction unit assembled, showing the teeth engaged in the two grooves, is shown below.

In Toyota V.P. Jim Lentz’ appearance on the Today show, he claimed that issues with the friction arm go back to only October of 2009. Not so. According to a letter from Toyota to the NHTSA , in 2007 Toyota changed the plastic material used in the friction arm (from PA46 to PPS) in response to problems similar to those occurring now.

Furthermore, Toyota has been facing similar issues in Europe going back to 2008:

Toyota has been modifying the friction-arm (CTS) type assembly since 2007. Yet to our knowledge, the Denso design has never been implicated in any sticking-pedal issue, and has presumably been in production for some ten years. Why didn’t Toyota change over years ago?

Toyota claims it now has the solution to the pedal problem. Later this week, Toyota will be sending shims that will be inserted under the friction arm of the CTS-built pedal to reduce its tendency to stick. We understand how this fix will work, and have replicated it. It does reduce the degree of friction; the exact amount will depend on the height of the spacer. Our one-eighth inch spacer made a fairly dramatic difference in subjective friction, but we could not test it installed in a car to see how different it would feel on the road.

Regardless of the thickness Toyota chooses for the shim, real and perceived friction will by necessity decrease to the detriment of pedal feel. The original designed degree of friction was obviously chosen to maximize the balance between the two forces at play; any change can only deviate from that, and away from that original ideal balance. We believe the odds are high that drivers will feel the difference, and that some or many may not like it.

Furthermore, the CTS-built unit is prone to continual wear and change in friction level over the long haul. I do not claim to be an expert, but having two small plastic surfaces rubbing against plastic does not strike me as an elegant, reliable or durable design, and one that is presumably subject to long term deterioration from natural and unnatural causes. There are a lot of twenty and thirty year-old Toyotas on the road. But it’s difficult to imagining this assembly still functioning as intended that far down the road, nevertheless even five or ten years from now. Toyota’s well-documented de-contenting is graphically on display here. Yet Toyota is apparently staying with this design, with some further modification, for ongoing new car production.

Whereas the Denso unit (above) is not exactly inspiring in solidity of its all-plastic design and build, it seems to lack the most serious flaw of the CTS unit. The smooth metal coils rubbing against the plastic housing seems less prone to deterioration and change in friction level. There are no known issues or problems associated with it.

We are calling on Toyota to replace all CTS-friction arm type gas pedal assemblies with either the Denso unit, or another proven design that lives up to Toyota’s legendary quality and the longevity expectations of its loyal owners. “Propping up” an intrinsically inferior and historically-proven inferior design with a piece of metal stamping is not going to restore Toyota’s tarnished reputation. The two units are interchangeable; Toyota should do the right thing  and switch production over, and insist on replacing all the CTS-type units even after they have had their temporary fix. A Band-Aid will stop the hemorrhaging for the moment, but nothing less than a transplant will do for the long haul.

Get the latest TTAC e-Newsletter!

122 Comments on “Why Toyota Must Replace Flawed CTS Gas Pedal With Superior Denso Pedal...”


  • avatar
    Corvair

    Very well put argument.

    Will anyone weigh in on the design features used by other OEMs? Given that trucks have had throttle-by-wire for a longer period, how have they worked their designs to avoid the problem?

    • 0 avatar

      I disagree entirely with the assumption the pedal is defective – this is an Audi 5000 story all over again. All the cars are automatic, none of the cars displays the problem when tested, every driver seem absolutely convincing. There were deaths claimed and lawsuits filed. Years later through deceptive psychological testing they prove the driver was mistaken about what their feet actually did. Everyone who sued lost.

    • 0 avatar
      ruthtruth

      I believe a sticking pedal did not seem to be the problem:
      from: http://wheels.blogs.nytimes.com/2009/12/08/toyota-tops-complaints-of-unintended-acceleration-report-says/

      “There is most definitely a problem with Toyota’s drive by wire / engine management system. In 2005 I bought a new Avalon. It frequently suffered a delayed response to throttle input – sometimes not responding at all and at other times over-responding with a powerful RPM surge after a 3 to 5 second delay after throttle input. Once after kicking-down the transmission refused to drop back to a higher gear when the throttle was released, and as engine RPM rapidly increased the vehicle surged forward for about 10 seconds until I managed to get the transmission to shift into a higher gear by kicking the throttle back and forth a number of times. No carpet was involved – this was a throttle/DBW/transmission problem. On another occasion the transmission disengaged itself while driving as if it had shifted itself into neutral, and only re-engaged after I pulled over, turned the ignition off and restarted.
      On two occasions the car unexpectedly pulled me into traffic almost causing me to be broadsided. Toyota denied that there was a problem with the vehicle, and after 6 months I traded it because I considered it too unsafe to drive.”

    • 0 avatar
      forever2

      Interestingly, many Audi vehicles have adopted an ‘engine stall’ feature on some models when the foot brake is applied very hard for a some specified duration.
      This makes a good safety feature for a few reasons, but I can’t not help but wonder if one of the reasons is to put to bed the legendary ‘unintended acceleration’ events (real or not).
      All conjecture, but maybe food for thought.

      I am not convinced this CTS pedal is a as bad as it appears on the surface. With all due respect to the analysis presented, putting it in a real world setting would be more conclusive.
      Still, the option of using the Denso ‘version’ would seem to be the better solution, to erase all doubts. It could be there are not sufficient supplies, or Denso assemblies are no longer in production, and tooling up would result in even more delays for a remedy. I guess this might be proven if Toyota goes back to the Denso in future models.

      As others have said, I do have to wonder if other manufacturer’s are using similar parts.

  • avatar
    John Horner

    Brilliant work Paul, and spot on.

    It is very damning that this issue with the CTS pedal, moisture and friction control are not something new at all. This is an old issue Toyota has been dorking around with for several years now. The “fix” reduces the baseline friction of the pedal assembly and does very little to ensure that said friction remains constant over the years ahead.

    Now if only the NHTSA would do its job and force Toyota to do the right thing.

    This is but another example of how the silly notion that “the markets self-regulate” the behavior of massive corporations for the benefit of all just isn’t true.

    Now the next question is this: Are the CTS manufactured pedal assemblies for other car makers of the same design as the one used by Toyota? That will tell us if the root of the design goes back to Toyota’s engineers or CTS’. Obviously the final responsibility lies with Toyota, but it is very interesting to understand where this design came from and who else, if anyone, is using it.

    P.S. I suspect that the engineers thought that any long term degradation of the friction characteristics would be on the side of less friction and that when it became annoying (sometime after the warranty is up!), the customers would be buying replacement pedal assemblies.

    • 0 avatar
      gslippy

      I cannot agree that the notion of self-regulating markets is silly. Toyota, Ford, Firestone, and a host of other big corporations have all felt market pain after ugly public safety recalls. Toyota will never make this mistake again, even if there were no lawsuits. Customers, dealers, shareholders, bloggers, and the MSM are all hopping mad – how will more government regulation make a difference in this circus of derision poured upon Toyota?

      Toyota is subject to the laws which regulate automotive safety issues, and appears to be abiding by them. I’m sure the NHTSA will become involved.

      The safety fix for the Pinto fiasco was nothing more than installing a formed piece of 1/4″ thick plastic between the gas tank and rear differential. I owned 3 of them. Could it have been made even more safe? Yes, but it would have been unreasonably expensive and time-consuming to do so, to resolve an issue which occurred so rarely, but with tragic results.

      Toyota is in the same “bind”. I’ll guess this fix will cost them $1 billion, for something whose incidence is rare and certainly depends upon the convergence of several factors.

      They may not have more confidence in the Denso pedal, and neither do I. I’m not convinced that coil springs rubbing sideways on plastic is a great way to generate pedal friction. Such a design is subject to the vagaries of plastics, too, such as thermal expansion, creep, fretting, moisture absorption, etc. I’d much prefer a friction stack, which the CTS unit approximates. Simply saying that the Denso unit “appears to be less problematic” is not sound engineering. It’s just Monday morning quarterbacking.

      Economics always plays a compromising role in every engineering design, and the fixes when those designs fail. Everybody suddenly becomes an expert when an airplane crashes, a building collapses, or a throttle mechanism jams. For most things, more regulation will only cost more money and drive companies out of business that provide us the products and services we want.

      Toyota has given the consumer what it wants – a quality vehicle for a good price. But somewhere in their specification and qualification program they made a mistake, which they are now correcting. Very few people want to pay for gold-plated safety. which is why everything from airline tickets to toasters are affordable.

      Every product, every car we drive is filled with compromises between quality, price, and time to market – regardless of brand. Government regulation of every choice isn’t necessary; the market will do much of it very well.

    • 0 avatar
      don1967

      I would argue that the markets are indeed self-regulating, but that “market forces” include Paul Niedermeyer and, to some extent, NHTSA (as a democratically-created voice for consumers). It’s all part of the ongoing negotiation between buyers and sellers.

    • 0 avatar
      texlovera

      I’m with gslippy & don1967. Toyota is about to take a huge bath over this issue. If they don’t get the fix right, they will lose customers for anywhere from a few years to life, and that will cost them an order of magnitude more than this recall will. Just ask GM & Chrysler what happens when you lose consumer confidnece, especially in the Internet Age.

    • 0 avatar
      Robert.Walter

      Markets are self-regulating … but only on a macro-scale and is subject to over-shoot.

      This is why, in the micro-time periods, we need regulation and enforcement.

    • 0 avatar
      kps

      That will tell us if the root of the design goes back to Toyota’s engineers or CTS’.

      Although patentese is not my native language, the CTS design appears to be covered by US Patent 7,404,342, held by CTS.

    • 0 avatar
      richardschulze

      Hey Gslipy, you think car companies self regulate? You should look at Fords long term history. Having a really embarrassing screw up has never stopped them from doing it over and over again. The Pinto was a fine example, which was followed by Ford Crown Victoria having the same problem (bursting into flames from rear end collisions – killing numerous police officers). Ford had the dubious distinction of the world’s largest recall on their ignition switches, followed by the next world’s largest recall on cruise control switches. And let us not forget the Ford/Firestone debacle in which Ford was wrong and all the other vehicles that had the same tires were never recalled because only the Ford vehicles were experiencing the tire problems.

      You should look at the history of the Cruise control switch recall (SCDS, NHTSA recall 05V-017). This recall covered 10 million vehicles. The investigations started in the mid 1990s. The first recall was in 1999, but the conclusions about the cause were incorrect and so was the fix and the wrong parties were accused (Texas Instruments and DuPont). The next recall did not occur until 2005 (six year later). The recall was rolled out over an almost four year period, the last wave was in August of 2008. The cruise control system in question was first installed in 1992.

      Ford denies to this day that their cruise control switches cause fires, or at least their legal council does. NHTSA was heavily involved with this recall but despite pressure from government, Ford continued to deny the problem.

      Ford also likes to hide behind a rule of law called the Economic Loss Rule, which protects them from subrogation on vehicles that out of warranty, hence the delay tactics.

      By the time Ford had recalled the cruise control switch in 2005, almost all the vehicles in question were out of warranty and therefore not Fords responsibility. The manufacturing date of the latest recalled vehicles was 2002. By 2005, these vehicles were three years old and falling out of the three year 36,000 mile warranty. The very last vehicles recalled, were the first vehicles that had the problem, Ford Crown Victoria, Lincoln Town Car, Mercury Grand Marquis. By the time these cars were recalled, they were between 12 and 16 years old.

      How about that for being good custodians of the public trust? Gslipy, know what you are talking about before stating an opinion. These are just the high profile cases. Ford has also had a long history of “runaway acceleration” just like Toyota is now having.

      For the rest of you doubters out there, runaway acceleration is a real problem. It did not become a problem until vehicles were equipped with either cruise control or fuel injection with automatic idle control. There is now an up tick in runaway accelerations because most car companies have switched from throttle cables to electronic throttle control, such as the recalled Toyotas. Most new cars no longer have a mechanical linkage between the accelerator pedal and the throttle body. Computers control the throttle opening with a servo motor. This is the first step in drive-by-wire. Just wait until there is no mechanical linkage between the steering wheel and the front wheels (soon to come).

  • avatar
    educatordan

    Hmmmmmmmmmmmmmmmmmmmmm. Has anybody done a cost analysis of how much money Toyota supposedly saved per unit by using the CTS style design over the other? I’m now interested to see what a GM unit looks like taken apart. How bout the new Buick’s drive by wire unit? I wonder if you’re tuning to an older clientele like Cadillac or Lincoln it affects what type of assembly you use. Those drivers would be used to a typical throttle cable for most of their lives.

    • 0 avatar
      Mike66Chryslers

      The public will probably never have enough details to compare the pedal assemblies on cost. To my knowledge, Toyota has never divulged why they even use two different designs. Perhaps the CTS design was considered superior because of the steel shaft and brass bushing, or the friction mechanism provides a more consistent pedal feel between pedals (when new, of course) than the Denso unit, which relies on springs rubbing on the housing.

      Many people are claiming that this fiasco is evidence of Toyota’s declining quality due to decontenting, but it’s quite possible that they viewed this design as superior to the Denso.

  • avatar
    Daanii2

    Although not precisely on topic, this post makes me wonder why we still all use brake and gas pedals and steering wheels. They are holdovers from the days when they were needed to create mechanical force. Now a joystick assembly would be much better, cheaper, and safer.

    Given all this with Toyota, though, no wonder no carmaker wants to get out of the Dark Ages of driver interface design into the modern age.

    • 0 avatar

      Talk to people who want to yank the gear lever. If we cannot even stop producing manual gearboxes, what to say about joysticks!

    • 0 avatar
      Sinistermisterman

      Speaking as someone who can’t drive anything on Gran Turismo on my Playstation but who can pilot a reasonably powered car around a circuit with varied and reasonable success, this idea would be great for some people, but for those of us who learned to drive a car before learning to play on a Playstation I’m not sure it’d work!

    • 0 avatar
      Garak

      “Now a joystick assembly would be much better, cheaper, and safer.”

      Have you ever driven a forklift with 100% electronic controls? The complete lack of feel is uncomfortable, and losing all control when a fuse goes out is not something you’d wish to experience in a car.

      And think of the lawsuits after the first fatality.

    • 0 avatar
      twonius

      I actually had a class where we discussed by-wire control systems for cars. As of right now the degree of redundency necessary for safety critical systems makes it prohibative (steering, brakes).

      Also it’ll be hard to get people to trust it at first. Even switching to hydrualic brakes from mechanical took a while.

      This gas pedal inscident will probably set it back another 5-10 years.

    • 0 avatar
      Mike66Chryslers

      Sorry, doing everything with a joystick is NOT safer. From an ergonomics perspective, it is beneficial to separate tasks so that one of your extremities is not overly taxed.

      A round steering wheel makes a great deal of sense for piloting a vehicle. You can grab it anywhere on its circumference and it will react the same. It’s easy to turn it slowly or quickly to turn the wheels at the desired rate. It’s difficult to hook it on something by accident and cause the vehicle to do things you didn’t want, unlike a joystick.

      Just about the only thing that makes little sense today is a gearshift lever on automatics. I see little reason that the gear selector shouldn’t be pushbuttons on the dashboard. Heck, Chrysler had this up to 1964, and that was an entirely mechanical system!

    • 0 avatar
      ponchoman49

      Time to pull your head out of the playstation junior. There is a reason a car has a steering wheel, brake pedal and shift lever. It’s called the design of the human body. Cars are not imaginary simulated images driven by joysticks

    • 0 avatar
      Robert.Walter

      On Steer-by-Wire … Joystick v. Steering Wheel … Which is better?

      Modern EPAS systems, whether column or gear mounted, can, if they go crazy, exhibit self-steering tendencies …

      In the design of the s/w and the hardware, every precaution is taken to avoid this (same could have been said for the ePedal too though).

      We did some tests, simulating these faults, and I can tell you, if you ever want to see a CEO turn ashen-faced, and ask what additional resources you need, just simulate system instability with the wheel swinging +/- 45°, and a reversed-input to -output relationship (not at the same time)…

      When an electric motor is strong-enough to turn the front wheels on a fully laden axle without any driver’s inputs (as in self-parking systems), it is more than strong enough to overcome the driver’s attempts to overcome it (and even strong enough to break a human arm under the right conditons.)

      Current EPAS steering systems are not pure steer-by-wire, they still have a mechanical (intermediate) shaft connecting the st. column to the st. gear, but the reality is that should one of these systems go nutso, whether you have a wire-connected joy-stick, or a mechanically-connected steering wheel, you will not be able to control the car.

  • avatar
    RoadRage

    Im very curious who built this part for Europe. It’s so easy to throw CTS under a bus, but we all know that the design specs were engineered by Toyota. Was it cheaper to built it this way? I don’t know. Why is the design so different then the Denso Unit? Again, i do not know. One thing I do know is Toyota will never man up and say that the CTS part was designed to their exact specs. This is the tale of two cities.

  • avatar
    segfault

    As I suspected, the shim installation will be the first in a long line of band-aids for Toyota owners. First, the zip-tie floor mats, then the shim, then replacement of the entire pedal assembly, then ECM reprogramming/replacement, then ETC replacement. I remember returning to the dealer every couple of months for repairs when I owned a VW. It’s one of the reasons I don’t have a VW any more.

  • avatar
    Brian E

    It’s irrelevant. The actual fix is the ECU patch that’s being applied to kill the throttle on brake/throttle overlap. I think it’s more than likely that the throttle pedal recall, like the floor-mat recall, is a smokescreen designed to shield Toyota from possible negligence suits arising from the lack of this throttle kill-switch. Nothing else could justify having mechanics at thousands of dealers repair millions of throttle pedals, a process that will inevitably create at least one unintended acceleration incident due to a botched fix.

    Furthermore I think you’ll see other manufacturers who use DBW throttles but don’t have a brake/throttle overlap throttle kill program issue recalls of their own to repair pedals or floormats – and updating the ECU at the same time. I expect Honda to be the first to announce this.

    • 0 avatar
      psarhjinian

      The actual fix is the ECU patch that’s being applied to kill the throttle on brake/throttle overlap.

      Probably not, if only because it’s still possible to have the car accelerate in such a situation, if the driver were to, for example, panic and pump the brakes and/or experience pedal confusion. I’ve seen stick-shift drivers make this kind of error in a panic.

      But you’re right about other manufacturers who lack the brake throttle-cutoff. Even if they implement such a feature, there’s still blood in the water, and the PI sharks are circling. Any manufacturer could be fair-game for such a suit or an expansion of an existing class.

      Where this gets interesting is the following question: when does content, or a lack thereof, constitute a suit-worth safety issue? Brake throttle-cut-off is the tip of the iceberg: what about cars that lack stability control, even though it could be added? Anti-lock brakes? Side airbags? Pre-collision systems? Tire-pressure monitoring? Where do you draw the line on what should or should not be implemented?

    • 0 avatar
      Geotpf

      What ECU patch? There is no such thing.

  • avatar
    mistrernee

    I really don’t like not having direct control over the vehicle I am driving… though this is a mechanical problem (apparently) I would still much rather have a cable attached from the pedal to a throttle body. I would gladly pay the extra money it costs to build the car with such a system in place even if Toyota won’t.

    Throttle by wire has nothing to do with emissions and everything to do with dollars and cents and cheaper speed control and traction control.

    I wouldn’t be surprised if some auto manufactures start advertising that they don’t use throttle by wire systems in their cars… (edit: if there are any left)

    Especially in the newer economy cars I have driven, the computer is very intrusive and it’s immediately noticeable with a stick shift.

    • 0 avatar
      gzuckier

      just got a honda civic si; the throttle by wire is a bit odd. in particular, it keeps the revs up when you get off the pedal, even in neutral, which makes fast shifting questionable. an emissions fix, i assume. (my older throttle cable civic did the same to a much lower degree, using the fast idle mechanism i assume) getting totally off the pedal while in gear and coasting, then getting back on slightly is most odd, in that there is no effect at all until you get up to the correct throttle setting for what the engine is doing, which definitely feels different from what happens with a linkage. reminds me somewhat of cruise control, which is something that i don’t use because it makes me nervous, given that i live in heavily trafficed areas.

  • avatar
    baldheadeddork

    Nice work, Paul. But this doesn’t square what happened with CHP officer Saylor and the ES350, which Toyota says didn’t need to be recalled because cars made in Japan were only built with the Denso assembly.

    • 0 avatar

      It doesn’t square with a number of other things. BTW, we don’t know what happened to that guy. Maybe his last words here were “watch this”.

    • 0 avatar
      Paul Niedermeyer

      @baldheadededdork, I’m utterly convinced that his problem was the floormats; he had two of them on top of each other. I know for a fact that the floor mat problem was real, maybe not all of the issue, but a big portion of it.

    • 0 avatar

      The Saylor crash was attributed to floor mat interference with the accelerator pedal, and the prior customer/witness that used that loaner car had the same problem. It was exacerbated by the keyless start/3 second hold kill switch issue. That particular problem resulted in the floor mat recall. I haven’t heard anything about a modification of Start button behavior yet.

      This recall is accelerator pedal related, but not the same issue. Whether the pedal is the sole cause of this problem is open to debate, but I think the Saylor crash was more likely due to floor mat interference, given the witness statement.

      @PN:
      Excellent work throughout this story. Keep it up!

    • 0 avatar

      Saylor’s case was most likely due to floor mats meant for the RX that the dealer had put in an ES loaner vehicle. That and panic in an unfamiliar vehicle with unfamiliar controls.

      The problem with this and any other pedal fix is that it’s not yet clear if the pedal is the main source of the problem.

      The CTS part actually looks like a more expensive part, with it’s partly metal construction.

      Toyota probably issued performance requirements, rather than the specific design used to achieve these requirements. Denso and CTS might then have adapted some of their existing designs to meet these requirements. It would have tested the part, though.

      DBW is useful for emissions, fuel economy, traction control, stability control, throttle response tuning, and no doubt other things. Cables aren’t coming back.

      The problem with the kill switch for enthusiasts is that there are times with performance driving where you want to apply both the brake and the throttle. The baby’s about to go out with the bathwater here.

    • 0 avatar
      Geotpf

      To repeat what Michael Karesh said, there are two seperate problems here:

      1. Floor mats, especially using floor mats from a different model of car.
      2. The CTS gas pedals.

      The incident with the CHP officer was due to #1 only.

  • avatar

    Toyota is making a big fuss about the extraordinary lengths to which
    they are going to in order to solve this problem. Recalling so
    many cars, keeping the (some of the) dealers fixing through the night,
    and suspending production on new models makes it seem as though
    all stops are pulled to make the vehicles safe.

    by Toyota’s own admission, replacing the pedals costs no more than
    modifying them, yet they will not provide new pedals except on a
    “case by case, at the discretion of the dealer” basis.

    They could do themselves a big favor right now by replacing with
    the Japanese pedals and rewriting the code so their would be a failsafe default to no throttle.

    Please Toyota, less PR and more disclosure and action. Just like
    GM could have eased their own pain by thinning the herd down greatly long ago, now is the time for Toyota to make a very drastic change.

    • 0 avatar
      Disaster

      The problem with replacing all the pedals is production capacity. Suppliers are necessarily lean operators (look up JIT) and they do not have the capacity to produce millions of throttles in a short period. The cost to tool up, and the time, would add considerably to the cost of the recall and the time it takes to implement.

  • avatar
    psarhjinian

    The only right action for Toyota is to acknowledge the long history of problems with the CTS-type unit, and replace them all with the superior Denso or another pedal unit that lacks the intrinsic flaws of the CTS design

    This is a good piece, but it’s a little idealistic in it’s aim.

    The problem is volume. Recall the Ford cruise control recall, now many years old and still behind schedule because it’s simply not possible to double (or, in this case, increase production by orders of magnitude) for a limited time.

    CTS is not set up to build the Denso pedal. No other supplier is set up to build it. Denso itself couldn’t fill an order of several million within years. So what is Toyota to do: stop production until they fill all recalled vehicles with the new pedal? That would be financial suicide, and you’d still have cars out there with the flawed design regardless.

    What we need know is if the shim is sufficient to address the issue for the expected life of the rest of the assembly, or the vehicle itself.

    • 0 avatar
      Paul Niedermeyer

      Use the shim for now, and change production at CTS to the Denso type unit, and replace the CTS units ASAP.

    • 0 avatar
      psarhjinian

      Use the shim for now, and change production at CTS to the Denso type unit, and replace the CTS units ASAP.

      But you’re still looking at years to do this, no matter how you slice it

    • 0 avatar
      Brian E

      I can be cranking out tested throttle pedals made to the Denso spec in a month if Toyota wants them. If Toyota wanted to explore this option, they’d put out an open bid, provide the tooling design under NDA to bidders who meet some minimum qualifications, and choose a set of suppliers who meet their needs for price, quality, and volume. It’s all a matter of money and hassle.

  • avatar
    redrum

    “This is but another example of how the silly notion that “the markets self-regulate” the behavior of massive corporations for the benefit of all just isn’t true.”

    That is not true, you’re making a straw man argument. Anybody who’s studied economics will tell you a lot of assumptions are made when describing market behavior in order to isolate variables. One of the assumptions in a true free market involves “perfect information” — knowing everything about the product being purchased — and obviously in these situations the consumer was kept in the dark.

    • 0 avatar
      psarhjinian

      One of the assumptions Communists make is that people won’t be power-hungry opportunists. As such, the Communist ideal is a nice theory, but utterly unworkable in reality because it relies on an unreal ideal.

      The perfectly free market is a similar philosophical fiction because it assumes a ground state of equality and opportunity that stops existing as soon as some players in the market get more leverage. This is why you have regulation: to try and bring the market nearer to that state, rather than allow it to enter a self-destructive feedback loop.

  • avatar
    Bill Wade

    The more I look at the CTS design the less I like it. Wouldn’t debris, wear, heat/cold and aging of the plastic affect the friction?

    Looking at this unit I would assume the older it was the less friction because of wear but apparently there’s other factors negating this.

    The Denso unit seems to be a very simple design that’s probably less prone to environmental and aging issues.

    Then again I know little about materials technology.

    • 0 avatar
      Paul Niedermeyer

      Bill, initially, the CTS unit looked better designed to me, but as I understood the actual issues involved, I totally came to see that the Denso unit is a more elegant solution to the requirements.

    • 0 avatar
      Disaster

      Like Paul alluded to, simpler is always better, when it works. Less parts means less to go wrong. Two excellent disciplines, in this area, are “Design for Assembly” and “Design for Manufacturing.” Every engineer should be required to be educated in Boothroyd and Dewhurst’s work.

      http://www.amazon.com/Product-Manufacture-Assembly-Revised-Expanded/dp/082470584X/ref=sr_1_1?ie=UTF8&s=books&qid=1265127735&sr=8-1

  • avatar
    cjpistonsfan

    Is it possible that ramping up production for the CTS units was a lot easier/less costly than ramping up production for the Denso units? That’s one reason I can think of as to why Toyota would be avoiding going with Denso’s across the board. It would also fit the admission by Lentz that rapid growth could possibly have contributed to this quality issue.

    Another thought that I like even less is that Toyota does suspect that there is a deeper electronics system problem that won’t be helped by the pedal fix, and therefore they don’t want to waste money changing the design when they know it won’t help long-term. Let’s hope that’s not the case and is simply a fun idea for the conspiracy theorists to talk about.

  • avatar
    blue adidas

    Toyota has screwed up royally.

    First, the design of the pedal shape was flawed and poorly tested, and stuck to the floor mats. They used zip ties to fix. Lots of people buy aftermarket floormats so the design needs to safely accommodate this.

    Second, they have a poorly designed and poorly tested accelerator assembly that they’re fixing with a shim.

    Third, There is no fail-safe brake system that are used in many European cars

    Fourth, there is no ignition shutoff system that works with the necessary immediacy

    Fifth, there appears to be a flaw in the design of the electric throttle control.

    That Toyota has awarded CTS with various quality and zero-defect awards shows that the components were built to their specifications. For Toyota to now throw them under the bus is comical. If this is one cause of this problem, it’s one of many causes. Toyota didn’t test it enough because, evidently, other CTS accelerator components with other automakers aren’t having this issue. Most people are seeing right through this.

  • avatar

    rumor is the F150 has the same problem and it’s not the pedal, it’s that damn drive by wire software. if Toyota’s fix doesn’t work, they are forever screwed.

  • avatar
    oldguy

    Paul:
    Not only are you a great writer, but your obvious engineering skills are beyond reproach. I’ve got an idea, why don’t we just find the closest big tree, round up and hang everyone that may have ever worked for Toyota.

    Dave

  • avatar
    Sinistermisterman

    Excellent article. It’s investigative writing with logical and intuitive thought processes involved like this that makes me glad I read articles on TTAC as opposed to reading some of the regurgitated pap that passes for news in other automotive news outlets. Thanks Paul.

  • avatar
    TomJones

    Does anyone know what other car models from other vehicle manufacturers have the CTS pedal? Will there eventually be an industry-wide recall, or are the pedals being supplied to other manufacturers significantly different that this will not be necessary?

  • avatar
    Disaster

    I’m an engineer who was involved in several safety system designs, product recalls (not my own designs) and who has designed an electronic drive by wire pedal (which I have a patent.)

    This report’s conclusions, are lacking a complete understanding of how these mechanisms might wear, age and fail and therefore I can’t agree with the conclusion (it is too easy to say the thesis is correct because the design is experiencing failures when you do not understand exactly how that is happening.) I also take fault with calling this evidence of cost cutting. If anything, the CTS design is more costly (adding a separate mechanism to produce hysteresis.)

    Here are some facts to chew on that might help better understand possible failure modes of BOTH designs and why one may or may not be superior to the other.

    1. All parts that rub together will experience wear, however some materials make better bearing surfaces than others because they have lower friction, or wear less, or wear exponentially less past a certain point.

    2. Materials have different expansion rates based on temperature. Plastics expand and contract more than metals.

    3. Plastics expand and contract with water content, which they absorb from the surrounding atmosphere.

    4. Plastics have considerably different elastic modulus than metals, higher strain per equal stress. Elastic modulus and yield and ultimate strength of plastics (deformation) can be considerably enhanced with fillers, like glass.

    5. Plastics are often used as bearing materials because of their lubricity and low wear characteristics. The consistent lubricity of them allows them to be used dry (sans lubricant) which removes a failure source.

    In the Denso design we have two dissimilar materials, plastic and steel, rubbing against on another to form the required friction and hysteresis. In the CTS design it is done with two plastics which may or may not be the same.

    Without extensive knowledge, and testing, it is very difficult to know how these two different bearing surfaces would age and how that aging process would effect their friction. Any comments on it are purely conjecture. It is even more of a 20-20 hindsight stretch to describe one of these designs as inherently inferior to other.

    Having said that, and given the facts I can think of several long term wear issues with BOTH designs. Any design that needs to have a consistent friction, which must stay within a maximum range to insure pedal return, should be vigorously tested before release under worst case conditions of temperature, moisture, and dust and must include accelerated aging of the plastic material.

    Toyota isn’t at liberty to jump to quick conclusions with this issue like the author of the article did. They must thoroughly evaluate what went wrong, design a new, better test that can consistently repeat the problem, and then design a solution that passes that test. Having been in their shoes I can relate to how difficult it is to create a test that meets this criteria. However, if they do anything less than that, they may very well find themselves going through he entire process again.

    • 0 avatar
      TomH

      Nicely done. My guess is that your rebuttal will not be part of the PI attorneys case who use this article in their pocket lining lawsuits.

    • 0 avatar
      gslippy

      I agree. I even think the CTS unit may be superior in concept. Maybe we just don’t know of any Denso failures yet.

    • 0 avatar
      Mike66Chryslers

      +1 Disaster! I agree this article is jumping to conclusions.

    • 0 avatar
      guyincognito

      Well said.

      I also have to echo what Michael Karesh said above. Toyota likely provided specifications for attributes like cost, feel, return speed, and durability to CTS but not the design itself.

      Further, as psarhjinian stated, it is likely impossible for Toyota to quickly change over to the Denso design.

      While I do appreciate the reporting TTAC has done to date on this issue, I think this recommendation is misguided.

    • 0 avatar
      Morea

      >> 5. Plastics are often used as bearing materials because of their lubricity and low wear characteristics. The consistent lubricity of them allows them to be used dry (sans lubricant) which removes a failure source.<<

      On an otherwise excellent post but I would like to take exception to this point.

      Plastic friction is controlled by a large number of hard-to-pin-down variables mostly due to the "blooming" of small molecules to the surface over time. These include absorbed mold release compounds, anitioxidation/antiUV additives (often designed to seek the surface), presence of monomers/low mass oligomers from the polymerization process, residual catalyst, polymer degradation over time producting low mass oligomers that subsequntly move to the surface.

      Polymers are molecularly porous materials. To coin a phrase, "what goes in will come out" over time. This will change the friction/wear properties.

      To further think on this, the outgasing of such additives by the carpeting and soft materials could infiltrate the pedal assembly and change the friction characteristics. Check the interior windows of a new car in the hot sun after a few days. This is all coming out of the plastics in the interior. Hard plastics largely avoid this which may explain why more manufacturers are using them. (And gives TTAC reviewers something to complain about.)

    • 0 avatar
      Paul Niedermeyer

      Disaster; You have some excellent points, and I don’t pretend to be an experienced engineer. Nevertheless, the key point is this: The CTS pedal has been modified several times, due to its design and materials used on generating friction; the Denso hasn’t.
      The CTS unit seems intrinsically to be prone to the kind of problems being experienced, and the Denso doesn’t. I’ve gone in depth to explain them in my argument. And the CTS pedal is failing. That tends to support it further, no?

    • 0 avatar
      Robert.Walter

      Disaster:
      I agree with many of your comments. That said, I think TTAC has done a service by trying to understand and expose the issues. As is true of any investigation, there will always be positives and negatives amongs the theories, suppositions and analysis. This is the purpose of leveraging the extended team (B&B) to search out the truth. I appreciate having such a forum where we all can exchange ideas.

      Different topic: Sealing is one thing that has bothered me with these designs. In my first posting on this issue, I suspected that the materials in the CTS design might be hydroscopic enough, and the ePedal assy might be close enough to the floor vent that there would be both an affinity for and a source of moisture… we’ve since seen reports that this is true.

      Faimiliar w/ Arizona Dust? I’ve been wondering how the friction characteristics of the two designs would be modified by the infiltration of this material (I think the CTS design would suffer more), yet we’ve not heard of this as a concern/issue.

      What’s your take?

    • 0 avatar
      NotSoFast

      Really good analysis and discussion. I’ve got a patent on a special screw thread using dis-similar plastic materials and designs that are much like the CTS unit in the thread shape (but otherwise not at all similar in function or use), so have had experience aplenty with plastics.

      It strikes me that the Denso unit has a major flaw as well, which is that it relies on steel rubbing against plastic, and although the surface area is much larger than the CTS unit, there seems to be a high potential for excessive wear at some point given the millions of cycles a gas pedal has to undergo over a few years.

      Isn’t there also a large potential for problems with wear in the Denso unit? As a designer, I never would have used a plastic surface as a friction “retarder” due to the potential for the wear.

  • avatar
    cardeveloper

    I’m concerned about the material change to PPS. That material is not known for it’s lubricity, and tends to exhibit “sticky” behavior. It’s dimensionally stable and will not absorb moisture like the nylon material. I’ve successfully used 4/6 nylon on high wear parts.

  • avatar
    william442

    Wow. We have come a long way from the days of carefully checking the throttle return spring on the big Holley between quarter miles.Have we progressed at all?

  • avatar
    BuzzDog

    Before everyone jumps on the “If we were still using throttle cables, this wouldn’t be happening” bandwagon, don’t forget that old-school cables also suffer from a moisture absorbtion issue.

    It’s called a “rusted accelerator cable,” and it used to happen fairly often.

    • 0 avatar
      gslippy

      Even I have mistakenly yearned for the old throttle cable, in spite of my one experience with one that stuck open while driving (due to ice).

      E-throttles do have advantages for function and packaging, even thought my only experience with one was lousy (02 Passat).

    • 0 avatar
      Mike66Chryslers

      I had my own “unintended acceleration” event when my drivers-side motor mount broke. When I pressed the gas, the engine cocked up and yanked on the throttle cable. This positive feedback loop (engine twists more throttle cable yanked harder) very quickly resulted in WOT! I had to think fast as 500ft.lb accelerated my car towards the tight 90* turn at the end of our quiet suburban street. I shut the engine off and stopped the car. After quickly figuring out the problem, I drove the car back home in reverse, so the engine torque would push it DOWN on the drivers side.

  • avatar
    Jeff Waingrow

    Disaster’s letter makes clear just how complex the designing of even the simplest systems can be. In a way, it’s a miracle that so many things work as well as they do. Yet we never dwell on these accomplishments, do we? We press our buttons and pedals and expect everything to work perfectly all the time and forever. Away from the areas of design, engineering and manufacture, where else are human endeavors so successful so often? I sense a disconnect.

  • avatar
    Philip Riegert

    Why don’t more pedals pivot from the bottom instead of from the top? Seems to get rid of any problems sticking the accelerator pedal with a floor mat. Porsche uses it along with many other European companies. What’s good for the goose …

    • 0 avatar
      racebeer

      My guess would be cost, from both a component and assembly perspective. You would have two assemblies doing the work of one that is integrated …. unless you actually put the electronics at floor level (turn the CTS module upside down, and you sort of get the picture). From an environmental contamination viewpoint, that configuration might be more expensivve to design.

      But, what do I know?!?!?

    • 0 avatar
      Mike66Chryslers

      If the pedal is attached to the floor, floor mats are MORE likely to ride up the pedal and weight it down if they slide forward, instead of bunching-up underneath the pedal, preventing you from pressing it down. Which condition do you think is worse?

      My Chryslers have gas pedals that pivot from the floor. I put rubber floormats in the one that I drive most often, to protect the carpet. I used to have to cock the drivers floormat to one side because the gas pedal was in the way. Eventually I got out a knife and cut a notch in the mat so it fits around the pedal.

    • 0 avatar
      Toad

      Pedals that pivot on the floor collect dirt, water, salt, etc. from the drivers shoes and floor debris at the hinge point. Look at a heavy duty International truck and you will see that both the brake and accelerator pedals pivot on a floor mounted hinge, and both are very prone to sticking if they sit unused for a while. The pedals will go down but not return to the up or neutral position. Anybody dealing with these trucks after they have sat knows to spray the pedal hinge liberally with WD40 before driving the truck.

      Very exciting when a 500 hp 19,000 pound truck has brakes that decide they will not release or the accelerator stays stuck on the floor.

      “Hanging” the pedals above the footwell allow the valves, hinges, springs, etc to stay dry and protected under the dash.

      Like my grandfather said, “The best thing about the good old days is that they are gone.”

    • 0 avatar
      Disaster

      It is partly an ergonomic issue. With a suspended pedal you can move your heel forward to hold the pedal in place…without even tilting your ankle. The driver can “create” a comfortable pedal angle by locating their heel wherever they want. It is less fatiguing. With a floor mounted system different pedal travels are all associated with different pedal angles. Those angles change relative to the person’s seating position too. Some people have limited range of motion and in some cases the angle might be quite uncomfortable…especially to hold for any period of time.

    • 0 avatar
      chuckR

      Porsche may have good reasons – beyond design inertia and tradition – for having a floor mount e-pedal. I have no idea what those might be. Edit – maybe heel and toe? The Porsche e-pedal fastens with a single slot/tab and one torx head screw. Just like the CTS and Denso units there is a cable connection out the top of it. Sprint Booster makes throttle response remappers that plug in line at the cable connection for Porsches and many other cars including the suspect Toyotas- probably, many drivers would object that resulting tip-in is too aggressive….

      A former co-worker long ago passed along the observation he heard that any competent engineer can design a water pump for a Rolls, but it takes a genius to design one for a Chevy. Same thing probably applies to a whole lot of auto bits and pieces and you can substitute your choice of expensive and mainline cars for Rolls/Chevy. In this case, bet the Porsche e-pedal costs more to produce in any quantity, its not even a bet that it sells for a lot more.

    • 0 avatar
      tced2

      Cars before some time in the 80’s(?) had accelerator pedals attached to the floor. Sometime in the 80’s virtually every (American) car I owned switched to accelerator pedals detached from the floor – and the lever to which the accelerator was attached was pivoting up on the firewall. My current vehicle (2005 Acura TSX) has returned to the accelerator attached to the floor. I have seen explanations that the “feel” of the pedal is superior (from Acura/Honda). The car is a drive-by-wire but the accelerator is still operating a cable to a “device” mounted on the firewall under the hood.

    • 0 avatar
      Disaster

      Forgot to mention one other reason…quite possibly the main reason, that pedals are suspended from above. This design puts the sensor and wires up above the floor where they are less likely to see contamination and foot damage.

    • 0 avatar
      Csnyder

      Bottom picot pedals are NOT the solutuion. Top suspended pedals came into use BECAUSE bottom pivot pedals had a serious propensity to stick. The “hinge” was down in the wet carpet, soaked with salt and water, which caused them to rust and seize. Lubricating them was normal maintenance on those old systems – and even when lubricated they would stick.

  • avatar
    phargophil

    In the past I’ve been an engineer as well as some others addressing this topic. Designing for a constant level of friction is very tricky as has been outlined above by more articulate people. Friction almost by definition mandates wear. Wear removes material which lessens friction. Spring loading to take up gaps caused by wear also varies because the spring lengths change. Simulating a mechanical feel in any substitute is very difficult.

    Next question is, what other vehicles use either of these design concepts, and what are their safety histories?

  • avatar
    guyincognito

    Agreed that the brake override will be part of the fix. However, this feature is not necessary unless there is potential for a stuck pedal. Also, it would not necessarily eliminate the problem of a stuck pedal. What happens after you release the brake and the ECU starts to get a WOT command again?

  • avatar
    50merc

    The best fix is a return-to-idle software command when the brakes are applied. Cruise controls already work that way. But if I were running Toyota, I’d also offer everybody a Denso-design (not necessarily Denso-made) pedal assembly. There’s no need to worry about production bottlenecks; once tooling is replicated the assemblies could gush out of factories worldwide by the jillions.

    And as others have said, what pedal assembly designs are used by other manufacturers? We shouldn’t have to wait for trial lawyers to investigate this as part of their sniffing for the sweet smell of class action awards.

    • 0 avatar
      Disaster

      When making engineering big changes one has to look at the current systems, how they are used, and what the ramifications of change might be. Is there ever a time when a person might need to apply the brake and throttle together…or close together? How about when rocking a vehicle out of a ditch? How about when holding a vehicle on a steep hill, while engaging the clutch or torque converter? These are things that have to be thought through thoroughly before making snap decisions.

  • avatar
    golf4me

    Lots of good points in the story and comments. I must point out, though, that the Denso version is not steel on plastic, it’s whatever the steel is covered with vs plastic. Probably enamel, which is very hard and very smooth and will not likely deteriorate, nor wear the plastic in any appreciable way in this application, whereas the CTS pedal, with two of the same materials will, especially if they are not aligned perfectly.

  • avatar
    Morea

    Some information on PA46 plastic:

    http://www.plastic-foam-product-manufacturer.com/46-pa46.htm

    Note that it is low moisture absorbing but Toyota cites moisture absorption as a potential problem. Hmmm…

    PPS is polyphenylene sulfide:

    http://en.wikipedia.org/wiki/Polyphenylene_sulfide

  • avatar
    ASISEEIT

    I just recieved an E-mail from a friend with an article about Toyota’s accelerater problem and that it might not be “Just” a pedal assembly problem! Steve Wozniak(Co-Founder Of Apple Computer) according to this article owns a 2010 Toyota Prius(Which Is NOT a Recalled Vehicle) and he can get his accelerater to stick over and over again during the use of the “Cruise Control”. He states he has been contacting Toyota and the N.H.T.S.I. and has had little or no reply. Could this be an electrical software problem? Wozniak thinks yes!

  • avatar
    ponchoman49

    I have never fully warmed to drive by wire electronic throttle control. And this is a good reason why. Another is that a good old fashioned cable assembly provides better feel and is more reliable in the long run and will be far less costly to fix. The fact that I am relying on two pieces of plastic to “simulate” friction is not very confidence inspiring. This is yet another case of why do we really need electronics in this part of the car over simpler more reliable and proven setups that work? I have to chuckle when I visit my firend at his vacuum repair shop. People are always bringing in expensive electronically controlled vacuum cleaners that all of a sudden quit working! Wanna gues why they quit working? Because electronic components do NOT belong in a dirty dusty vacuum cleaner that is subjected to lots of movement and abuse, varying temperatures and conditions. See any similarity here?

  • avatar
    Truckducken

    Major design no-no. The two parallel tabs and channels are a recipe for major force effects driven by minute misalignment, wear, or contamination. Anyone ever heard of ‘stiction’? If my Sienna had this assembly, I’d snap off one tab and keep driving.

  • avatar

    Thank you for the review. Having dealt with material and design failures like this as part of my work… this is very enlightening.
    Are they going to use an epoxy to affix the shim? Have they tested that fix over a lifetime of thermal cycling, will a dealer be prepping the surfaces adequately? Will the shim be grit blasted? Most polymers are tough to glue to as well. If they drill + bolt – will they be intiating environmental stress cracks in the plastic?
    Hokey ME quick fixes for a material/design problem.

  • avatar
    Ronman

    Excellent Report… i don’t think anyone else has done what you guys have…. keep it coming… and Toyota better watch themselves…will be patiently waiting for the aftermath…

  • avatar
    HalfMast

    Excellent analysis, Paul! But let me take it a step further to identify the problem and difference between the two units.

    The CTS unit:
    The design goal here appeared to be for continued friction, even after the plastic wears down significantly. Looking at the placement of the return spring attached to the friction arm as a lever, this implies they accomplished this by applied a direct force on the lever arm. So as the teeth wear down, they are pushed against the channels to continue providing friction.
    The problem with this design appears to be that as the teeth are worn down, they will decrease slightly in size and be pushed further into the channels. Now, you have increase contact points with plastic that has not yet been worn down, which will have a different coefficient of friction than the plastic that has been in constant contact. As the teeth move through the channels, this new plastic is a lot more likely to “grab” and provide a moment of much higher friction than expected, possibly overcoming the return spring. Of course, what plastic they choose to use for these parts is a key peice to their wear properties (thus the change in material in 2007).

    The Denso unit:
    Because the source of friction here is the tangental contact of the spring itself with the side housing, this unit will not have the CTS’s “feature” of continued friction after significant wear. Instead, as the housing is worn away, the percieved friction will slowly reduce. There is no direct force to push the friction components together, only the tangental contact with the housing. After time, there may be a significant decrease in friction than when it first came off the lot. Someone who drove that car throughout it’s life likely wouldn’t notice. But more importantly, it’s failure mode is “open”. Worse case is a complete lack of friction over time, not a sudden increase of friction like the CTS unit.

    That is my continued analysis based on what you described and the pictures. I can see why the CTS unit was designed the way it was, but I think they over-engineered it and created the failure mode that we are seeing.

  • avatar
    Slack

    I may be in the minority here, but I would submit that the folly here is not Denso V. CTS. The folly is replacing a cheap, reliable steel cable with finicky, complicated electronic parts. Can someone please explain to me why, other than giving the engineers something to do, and the dealerships more things to fix, drive by wire is so hot? You can actually make the argument that fly-by-wire makes planes more safe while at the same time lighter. What is the corresponding argument for cars, whose numbers on our highways dwarf airplanes (in the air, even in New Jersey), and whose operators have far, far less training. And once they figure out how eliminate the physical connection between steering and braking and the cars’ wheels, watch out.

    • 0 avatar
      Robert.Walter

      I don’t agree. As cars move from mechanical throttle-based units to mostly electrically driven units, it will be necessary to have ePedal technology … and, it would still likely be desirable to have a hysteresis feature in that “throttle” system.

      This is a case of bad conceptulization due to limitations in material selection and production processes.

    • 0 avatar
      Slack

      I guess my question is more basic. What is the rationale for moving from mechanical to electronic throttles? Nothing with moving parts is ever going to be completely free of problems. But from what I have read, Toyota has had major problems with throttles ever since they moved to electronic. What is the benefit, is it fuel efficiency, is it cost, is it weight, or is it just part of the effort to make cars even more complicated to increase dealer repairs? I’m not an engineer, nor to I play one on television, and I’m not a Luddite. But from my layman’s perspective, drive by wire has always seemed like a stupid idea. And nothing I’ve read about this latest situation changes that. Just because we can does not always mean we should. Someone please explain the advantage.

    • 0 avatar
      psarhjinian

      I guess my question is more basic. What is the rationale for moving from mechanical to electronic throttles?

      Lower emissions, better fuel economy, more power, stability control.

      Oh, and you get a net reduction in things like sudden, unintended acceleration.

    • 0 avatar
      Slack

      Indeed, as we have seen. Thanks for the info.

  • avatar

    Very nice write-up.

    But I will disagree with your conclusion that the Denso pedal assembly is the superior one. Looking at the two friction mechamisms, it seems that the CTS assembly is superior for its ability to compensate for wear, something the Denso assemble lacks. The presence of the angled grooves suggests that as the surfaces wear the movable surfaces will “sit down” into the grooves on the circular part attached to the pedal, thus maintaining better a constant amount of friction.

    I suspect the problem, other than the surface friction characteristics when working in a humid environment, is the angle of the grooves isn’t optimal which allows too much contact as it wears. Suspicion only; I’m a chemist, not an engineer.

    • 0 avatar
      Paul Niedermeyer

      Frankly, I was surprised at both of them. Given how critical they are, and how important it is to have the right amount of friction, I expected something more sophisticated. But in relative terms, the Denso seems superior for the reasons I gave.

  • avatar
    Brazilian owner of Corolla

    My wife has a Corolla model 2009. In last september she had an episode of sudden acceleration and a near accident in a street of our city. She was lucky to deviate to an empty avenue and managed to stop by pushing into neutral, pulling the parking break and turning off the engine. After some days in dealer, Toyota of Brasil stated that the matress provocked the incident. My wife disagreed with them.
    Now the episode is clear, the problem is worldwide. Toyota of Brasil denies that there’s problem with cars produced in Brasil.
    The pedals of brazillian Corolla are produced by DENSO. Local newspapers reported four more incidents. One led to accident with total loss of vehicle and driver’s light injuries. Only cars equipped with CTS peddals are being recalled in US, Europe and China!!!!
    I think that the greatest question is: are the pedals responsible for the problem? Brazillian Toyotas have problem of sudden acceleration and use DENSO pedals! We are anxiously waiting for these anwers.

  • avatar
    crash sled

    Nice write up, Paul, and excellent discussion and comments by all. I’ve spent some years in automotive product development, following years in another unrelated engineering field, and have great respect for real car guys. You are collectively brilliant, and make magic. If only we could get the beancounters out of your way (if only… right!).

    That said, and I may not fully understand the issue or operation here, but it seems to me that the “fix” shim is designed to eliminate/restrict travel down into the grooves. By definition then, it’s not a part that will now “wear in”, no? If your cartoon is correct in concept, they’ve eliminated “wear in”, if there was ever any designed into the system.

    The original part had a little nub that finally restricted travel down into the groove, but the shim now eliminates travel beyond the crown of the grooves. They herein seem to acknowledge that the geometry of the arrangement, perhaps coupled with material issues, forces them away from the proposed “wear in” concept (if it ever existed), and into a “fixed” concept. I’d agree this will affect and change pedal feel, over time, but then so will the Denso concept potentially.

    Also, the arrangements and concepts for the magnetic field sensors are different between the 2 pedals. Could this part of the problem? The CTS has a lever arm passing through fixed arms, nice and unencumbered data flow, hopefully. The Denso appears to have a rotary arrangement, presumably, in the round green colored area. The CTS lever arm geometry (and thus its magnetic field and readings?) might appear to have potential to change as its “groove” system wears in. The Denso might also change as the pivot point mechanically wears, but in a different way. In any event, this is another area of exploration… the differences in how these signals are produced and recorded, and whether and how age and use affects each.

    I’m not uncomfortable with differing components and commodities in vehicle engineering, it’s common practice, and we’re always balancing QCWF… always… and this is the inevitable result. If Toyota has surrendered on any of these, shame on them, but there’s no definitive info that they have.

    All due respect, but I have to agree wtih the above commenters, that you should wait before making firm judgement on all this. Too many times, the complexity of auto engineering has twisted teams around, and forced a retreat back from a prematurely-taken path, at any and all points along the vehicle program life cycle. Toyota is likely following a defensible path, even if perhaps belated. I also agree that calibration and control could be a culprit, or part of a culprit. Forget the enthusiasts, and force an automatic idle where it’s needed. The enthusiasts can buy an aftermarket defeat for this, if required.

  • avatar
    Slack

    Disaster wrote: Like Paul alluded to, simpler is always better, when it works. Less parts means less to go wrong.

    I’ve been monitoring this discussion for a few days now and all I can say is this, were it not for the quest for fire with drive by wire, this discussion would not be taking place. I stand on my earlier observation that replacing a simple spring loaded steel cable with finicky electronic parts that have to be made to mimic the action and feel of said steel cable is trouble waiting to happen. Can’t wait to see what they affix the steel shim to these plastic CTS assemblies with. Anyone know? Will it be epoxy? If so, can’t wait until batches of shims start falling off because the epoxy wasn’t mixed right, wasn’t mixed at the right temperature, got contaminated by dirt, or the plastic or metal surface wasn’t prepared correctly, and so on and so forth.

    • 0 avatar
      Csnyder

      The dedicated friction shoe design used by CTS would appear from a basic engineering view to be the superior product.The two plastic parts rubbing together at low speed should have an almost unlimitted lifespan as there is very litle heat generated by the friction, and the wear resistance of properly selected plastics can be VERY high.

      However, the RIGHT plastics need to be used – nothing hygroscopic that then expands with moisture content for instance.

      The Denso design is actually a bit scary. That steel spring is wearing against the plastic case of the assembly – on both sides. IF and WHEN it wears excessively, the coils of the spring will go through the plastic case, and the possibility of it jamming up are quite high.

      The shim is a pressure-friction fit between the case and the friction shoe from the looks of it – so no adhesive issues.

      As far as cable throttles go, their record is not stellar either. MANY cables have stuck over the years. Cables fray. Water get into them and the freeze. Linkages (clips) fall off. Return springs break, come off, and jam. You need to look at the history of throttle controls. When did the cable become king??? What caused it???
      You only need to look back to about 1968 or 1969 and the massive recall by GM of all cars back to something like 1958 due to throttles sticking WIDE OPEN when engine mounts failed. Mechanical linkages went over-center when the engine lifted off it’s mounts under accelleration. It was a massive recall, and to do it right GM would have had to replace EVERY ENGINE MOUNT EVER PRODUCED from the first implementation of the side mounted engine mount in the late fifties or early sixties, on every product the built except the corvair.
      The fix? They supplied a loop of cable to bolt to the engine (exhaust manifold on V8s) and loop through the upper control arm pivot to restrain the engine from lifting off it’s mount.

      They re-engineered the mount to have interlocking hooks on the upper and lower plates so when the rubber came unglued the engine could not move far – and then the went to cable throttles to make sure they could never go over-center IF the mount allowed the engine to move.

      GM learned something that time — an experience that sadly has been too rare since.

      Going back to cables or mechanical linkages would be a definite step backwards.
      Getting the TBW system properly sorted out WILL happen -brake actuation over-riding throttle is a REQUIREMENY on electric assist bicycles – why should it not be on cars??????

  • avatar
    rockford_il

    Great comments everyone regarding the mechanical design and material aspects!

    Has anyone compared the electrical outputs of the CTS sensor to the Denso sensor? Are the sensors even interchangable? Are they using the same magnetic sensor technology?

    Autoblog’s gallery shows that there are 2 electrical outputs from the pedal. Are the electrical outputs always continuous and tracking together properly during pedal actuation? How does the ECU handle ambiguous / discontinuous data from the sensor(s)? Does the addition of the shim affect magnetic performance / sensor sensitivity?

    The photos also show a laptop connected to the car. If the Toyota diagnostic was also silently re-flashing the ECU software while he tests the pedal…would the Toyota technician know?

    Autoblog gallery:
    http://www.autoblog.com/gallery/how-toyota-is-repairing-sticky-accelerator-pedals/#27

  • avatar
    mpoccia

    OK, one obvious problem w/CTS assembly (aside from its cost/complexity vs the Denso). Look at the Denso springs – outside is wound RH, inside is LH. The coils cross each other and so would have difficulty catching on each other to cause a lock-up. There also seems to be some sort of “shield” between the coils for extra safety. (Smart thing to do given the wrap of the spring around the arc, which increases the propensity for coil interference.) CTS springs are both RH wound, and are therefore sitting there waiting to catch and jam. For years, it has been absolute standard practice to wind coaxial springs in opposite directions as Denso has done. Just goes to show that there’s no such thing as a no-brainer.

    Also, the way things usually work these days, Toyota probably didn’t design the two components differently – more likely, Denso designed one and CTS the other.

    So, even if Toyota’s evaluation of problem assemblies they’ve seen is correct, I’m still betting that coil interference will cause some pedals, some times, to fail to return, sticking down. There’s a chance that they have some sort of lubricious coating on one or both of the springs, but the design is dancing with the devil. Springs are cheap and turnaround is fast. There’s no reason not to toss one of those as part of the fix, replacing it with one wound LH. As long as the coax springs are same-wound, this problem will continue to rear its ugly head. Murphy may be dead but his Law still holds.

  • avatar
    crash sled

    I think you’re mistaken. The CTS unit appear to have one LH apring and one RH spring.

  • avatar
    mpoccia

    Mea culpa, crash sled. Our retirement-age eyes aren’t what they once were. Zooming in on the pic, I can see from the light/shadow pattern that the inside coil is, indeed, LH wound. My apologies for wasting forum time and bandwidth, and also to Toyota and CTS.

  • avatar
    crash sled

    No apologies required, mpo. You’re a man after my own heart.

    My philosophy for dealing with automotive OEM’s is to seek out the heaviest baseball bat you can swing at greatest angular velocity, and apply same to the appropriate OEM’s head. A nail pounded through the end of the bat is an optional but recommended accessory. Sometimes, if you swing hard enough, you might get through to them.

    Not that I think the NHTSA is capable of hitting in the big leagues… because they’re not, which is why we need guys like you stepping up to the plate and hitting, even if you foul one off once in a while. Keep swinging!

    • 0 avatar
      mpoccia

      Well, thanks, but I at least owe apologies to the CTS and Toyota folks, whom I (essentially) called out as having done something very dumb, failing to catch it before it was released for production, and then failing again in the face of failures – all on my own misreading of the pic. As a retired design/development engineer, I’d be entitled to an apology if the shoe was on the other foot.

      As for the OEM relationship, that could get very complex, in this case. Does anyone know what/how much design responsibility Toyota has in the US? Can anyone here authorize an engineering change without running it through Japan? What kind of test facilities do they have here? (I’m assuming that the CTS pedals – at least mostly – affect US assembled cars.) The answers to these questions might not affect my confidence in the quality of a design, but they would affect my confidence in the quality of a quickly-done design FIX. Other factors: How was the “condensation” theory tested? How was the “fix” tested – including simulating wear on the mechanism?

      Also, they DO have a a bat with a nail. Toyota is the customer, and can shop somewhere else (Denso?). Unfortunately, at this stage, that’s almost a nuclear option. The damage to Toyota in having to bring another supplier on-line with lots of quantity might be greater than the entire value of CTS. Their only viable options are: Help CTS with whatever resources they can bring to bear, and/or handle the redesign themselves. In either case, the buck stops at Toyota. Their easy choices disappeared once the design was released for production.

  • avatar
    Slack

    I’d like to get this group’s opinion of this:
    http://www.edn.com/index.asp?layout=blogpostPrint&blog_post_id=630052463

    • 0 avatar
      Csnyder

      I’m not going to agree with the whole thing,but the point about proofing software is definitely valid. Today’s programmers are sloppy because they can be – and a lot of code is generated by the 3rd (or higher) generation programming languages in use. If everything had to be hand coded in machine language, with proper embedded documentation, there would be a lot less code used to do the job, and it would be compact, concise, and correct. It would also be more easily corrected when problems arise.

      That said, we are STILL not sure it is a code problem, or even an electronic problem, at this point.

  • avatar
    mpoccia

    Well, I don’t think that situation’s going away. Maybe the best bet is to get on a very familiar basis with your “manual” overrides, particularly brake, gearshift, ignition switch, and (if you happen to be as lucky as I am) clutch. Oh yeah – even the parking brake – you never know.

  • avatar

    I once read an article about Denso being Toyota’s secret to success. I beleive it mentioned that it was Denso who taught Toyota it’s quality control practices back in the day. However, the final product is Toyota, so they get the credit.

    It’s funny that an error within Toyota is caused by switching to parts not made by Denso.

  • avatar
    mpoccia

    Toyota’s “secret” to success was extensively studied and described in 1990 in “The Machine that Changed the World.” (Paperback is at amazon.com – $11.18, sorry, I’m new to posting and couldn’t figure out how to link up.) It’s been quite a few years since I read it, so I don’t recall a mention of such a role for Denso, but the specifics QC practices, while important, were but a part of a re-thinking of how to do business, initiated by a desire to save the company. It’d probably be interesting to read (or re-read) it and compare that Toyota to the company we see today (especially for those who may be more familiar with Toyota’s current philosophies (actual, not necessarily stated) and practices today.

  • avatar
    katekebo

    As a side note it’s worth to review the infamous case of Audi 5000 and it’s problems with sudden acceleration. I think the Audi case bears many similarities with current Toyota case.

    Audi was successfully making inroads into the US mid-size luxury market. Then, a series of consumer complaints about sudden unintended acceleration started to flow in. The NHTSA ordered an investigation and Audi concluded that the accidents were caused by drivers errors, with a contributing factor being the accelerator and brake pedal layout, which was different from American cars. They recalled the cars and modified the pedals. But the complaints continued and media frenzy exploded, culminating with a “60 minutes” report that ended up destroying Audi’s reputation. Audi took a defensive stand and contested the allegations, adding more fuel to the fire.

    Eventually, after many years of investigation, NHTSA could not find a defect with the car and quietly concluded that the most probable cause of the accidents was …. driver’s error. A separate investigation by a private company reached the same conclusion. CBS quietly admitted that the car shown in their “60 minutes” episode was rigged (a clearly unethical behavior for which they never apologized).

    Audi’s biggest mistake was to take a fairly aggressive stance and confront the media and NHTSA during initial phases of the issue. They exposed themselves to the media and public outrage. The learning is that it doesn’t matter who is right – what counts are perceptions and public opinion.

    There is no doubt that Toyota screwed-up. They did not identify and address a growing problem in due time. But we are just seeing chapter 1 of the entire story.

    By the way, today Audi (part of VW group) is not only the fastest growing luxury brand (ahead of Mercedes, BMW and Lexus), but their parent company is on track to become the largest and most profitable car company n the world.

  • avatar
    Slack

    I am shocked, shocked to learn that Toyota is now admitting that their little shim and epoxy fix might not “fix” the problem after all. But of course, they are still denying it’s an electronic problem. Puhleeeeeeze.

    http://www.boston.com/business/articles/2010/02/23/toyota_apologizes_for_handling_of_safety_issues/

  • avatar
    JHanko

    Lots of very good information here. The way I see things, the Denso unit definitely has a more favorable failure mode, losing friction instead of increasing to the point of sticking. I’m suprised I haven’t seen any mention of this, so I’ll throw it out there: What about a viscious fluid for the friction simulation? Similar to the doors on old cassette decks. I have a 15 year old cassette deck that the door still opens at a nice, controlled speed. I realise that extreme cold temps would affect things, but I would think a good spring pressure could compensate for this. The viscious fluid setup would basically have no wear whatsoever. Just a thought…

  • avatar
    rjh892

    Nice write up Paul. I have two concerns that make me believe that electrical problems are the root cause.

    #1. A recent accident reported involved a lady who left a dealership after getting the shim installed and floormats adjusted. Needless to say, she crashed into a snowbank after her car fully accelerated without her control, the moment after getting the so call pedal issue “fixed.”

    #2) I have also been hearing reports of people complaining their brakes would not work. Granted, yes, if your throttle is fully opened, your brakes wouldn’t stop a moving car. But it is hard for me to believe that if you fully depressed your brakes that the car could still accelerate to over 100mph, as some people have claimed. These are little 4 cylinder engines, plus brakes are usually “over done” or are bigger than needed to compensate for extreme situations.

    I am becoming a bigger believer in an electrical problem that is causing the throttle to fully open when undesired. How, or are, the brakes being over-ridden when the issue develops? Or, is it the sensor in pedal due to moisture or just a glitch in the ECM?

  • avatar
    Moparman426W

    I just plain don’t like the feel of drive by wire systems. it feels like there is no connection between you and your engine.I wish my ram’s hemi had a throttle linkage.

  • avatar
    VibeOwner

    I’m no chemical engineer, but I am a quasi-Toyota owner by default and suspect the recall is a band aid at best.

    Wouldn’t a better “friction” component be a material like teflon that has a low friction/expansion coefficient to begin with against itself or against steel? If so, it seems neither design would have been used – I guess it might have cost Toyota $25 instead of $20?

    Considering the fact that the distributors in these cars use a position notch wheel to determine engine speed (with backup) and Photo-transistor pairs – wouldn’t this be a better approach? There’s wires and power already going to the pedal – and it seems crude to use hall effect sensors for something that’s been proven for years in scanners, printers, industrial robotics – the list goes on and on. This design could provide the ECM with precise pedal positioning, as well as direction and speed feedback of pedal movement much more precisely than a magnet.

    With current flowing at different rates around the dashboard (and even which direction you’re driving) couldn’t that affect the sensor? The hall effect design requires overwhelming that possible magnetic “noise” interference.

    Not saying that any of these thoughts are necessarily better, but I’m interested in hearing other’s thoughts.

    Thanks for taking the time to explain and show the detail

  • avatar
    george70steven

    The transmission disengaged itself while driving as if it had shifted itself into neutral, and only re-engaged after I pulled over, turned the ignition off and restarted.
    online car insurance quote


Back to TopLeave a Reply

You must be logged in to post a comment.

Subscribe without commenting

Recent Comments

New Car Research

Get a Free Dealer Quote

Staff

  • Contributing Writers

  • Jack Baruth, United States
  • Brendan McAleer, Canada
  • Marcelo De Vasconcellos, Brazil
  • Vojta Dobes, Czech Republic
  • Matthias Gasnier, Australia
  • W. Christian 'Mental' Ward, Abu Dhabi
  • Mark Stevenson, Canada
  • Cameron Aubernon, United States
  • J Emerson, United States