By on October 3, 2015

Volkswagen is working on a fix to bring their “defeat device” equipped cars back into EPA compliance. The SCR upgrade option is very costly and another possibility is a software fix. The latter option would likely come in the form of an ECM calibration that would work similar to “test mode” at all times, possibly robbing power and fuel economy.

The crew over at TFLCar attempted to emulate this test mode on a dyno but fell short of collecting reliable results.

While I applaud the idea of finding an affected car and trying to generate real-world test results, the tests and analysis in the video do not correlate to an actual EPA test.

The first red flag in the video was the conversation with what appears to be an emissions station technician. He makes a point of mentioning that they have to disengage traction control and that is likely what puts the car into test mode. He throws around a bunch of acronyms including ESP and ASR and then tells us that they disengaged it on the Jetta to do their test. This is not possible as the Mk VI Jetta does not have a button to disable traction control, so they would not have a way to disable it short of retrofitting a button to every Jetta that rolled in.

There is a big focus on running the car on a loaded, 4-wheel drive dyno and how they are able to simulate on-road driving before then disengaging the valve for the rear dyno roller to simulate an EPA test with just the front wheels rolling. The first test completed successfully without any errors and they share the horsepower and torque results along with a set of corrected results. The correction factor is usually put in place to correct for barometric pressure and temperature. As Denver is much higher than sea level, their correction factor causes the figures to go up by a small margin.

They discuss how the corrected results are very close to the numbers stated from the factory but fail to understand or explain that factory figures are measured at the crankshaft while theirs were measured at the wheels. If they were measuring on the exact same device with the exact same variables put into the system, their results should have actually been 10 to 15-percent lower due the loss of power through the drivetrain. Since they do not know what device was used to measure the power at the factory or what factors were involved the comparison to factory, figures would be meaningless anyway.

Once the first run was completed successfully, the technician disengages the rear roller and tells us that he is going into two-wheel-drive mode. The technician was likely trying to communicate the change on the dyno, but the message comes across as if he has made a change to the Jetta which is only available in front-wheel drive. The test initially fails as the car goes into limp mode due to a missing signal from the rear speed sensors. Eventually, errors pop up on dash showing that ABS, stability, and traction control have issues.

The technician makes the incorrect assumption that this must mean that the car has switched into the “test mode” and proceeds to run the test in a single gear. It is very unlikely that the car switched into test mode as it requires a strict set of variables to be set, such as steering angle and barometric pressure, which cannot be met easily in such a setting.

The test mode also requires running one of the EPA test cycles, such as the FTP-75 test, in order to engage the test mode. This tests requires starting from a cold engine and then running the car for 30 minutes in various gears and speeds to simulate an urban driving cycle. Since TFLCar’s test was ran in a single gear, it never approached anything mimicking the EPA test.

The second test did show a drop in horsepower and torque, but that can easily be explained by the ECM programming and dyno configuration. Since the second test was ran with errors for ABS, stability, and traction control, it is very likely that a calibration exists within the ECM to limit torque in such a condition in order to make the car safer.

The dyno can also play a role in skewing the results since the load on the dyno is changed when the rear roller is disengaged. Once the load factor is changed, torque will drop, especially as the car starts moving in the lower end of the RPM range.

The best way to think about the load change is to think about a car in neutral and how the engine is able to rev and seemingly not produce a lot of torque since there is no load on it. Once the engine is in gear and loaded, it produces more torque. These changes can be measured by running a car in various gears and seeing the change in the output numbers. The load on the engine will change as the gear ratios change and produce varying torque results.

Power and torque are going to drop if the fix from Volkswagen involves a reflash that is similar to the test mode calibration, but the only way to get an accurate result of exactly how steep those drops will be is to perform the test on a car before it is reflashed and then run the test again with the same exact variables and settings once the new calibration is in place.

Get the latest TTAC e-Newsletter!

Recommended

53 Comments on “TFLCar’s Jetta Dyno Test Doesn’t Show Us “Test Mode”...”


  • avatar
    HerrKaLeun

    I think most people underestimate what it takes to accurately test. It isn’t just hooking up some hose and push start. You need expensive equipment and technicians knowing how to use it. Test has to be reproducible and defensible in court.

    Past comments alleged some journalist or car magazine just should buy some $5K tool to test all cars. BS, you basically need to hire a certified engineering firm that has better equipment and knows what they are doing.

    In this case, they still were on a roller stand. the cars accelerators kind of know the car isn’t moving….. I think Bozi did a good job pointing out the other flaws.

    And it wasn’t clear if the car went into actual emission test mode since they didn’t run an emission (part load) cycle. it looked more like the VSC slowed down the car as it showed the “ESP” symbol. i don’t claim to really understand their test, but don’t think it reflects what we can expect once they reprogram to permanently meet emissions. And they didn’t measure NOx, so this doesn’t really tell much. NOx was the whole point….

    • 0 avatar

      On that $5,000 tool: A professional portable emissions measurement system costs around $100,000.

      • 0 avatar
        heavy handle

        Is Snap-On not considered “professional”?

        https://store.snapon.com/Flexible-Gas-Analyzers-Kit-Hand-Held-Gas-Analyzer-5-Gas–P821505.aspx

        I’ve seen this tool in action. While probably not “EPA grade,” it’s good enough. Only trouble is, it’s not cost-effective for most shops. Maybe it could pay for itself in California if your shop specializes in bringing pre-OBD cars into CARB compliance.

        • 0 avatar
          Rudolph

          https://store.snapon.com/Hand-Held-Gas-Analyzers-Kit-Hand-Held-Gas-Analyzer-5-Gas–P821508.aspx shows 3390.00 USD — Item: HHGA5BP

          ▼ Error message ▼
          https://store.snapon.com/Flexible-Gas-Analyzers-Kit-Hand-Held-Gas-Analyzer-5-Gas–P821505.aspx returns error message
          ▲ Looking for something? ▲

          We’re sorry. The Web address you entered is not a functioning page on our site.

          • 0 avatar
            heavy handle

            I guess the old SKU has been superseded. The link worked a week ago. Maybe they sold out.

            The link you provided (HHGA5BP) works. It’s a similar machine, at a slightly higher price.

            I still don’t think that journalists need this type of tool, but it comes-in handy when there’s a suspicion of cheating. Your local Snap-On dealer can perhaps tell you if any local shops have one, in which case you just need to pay the shop rate (or whatever you negotiate).

      • 0 avatar
        HerrKaLeun

        Agreed, some people here on the comments (and elsewhere) claimed one can just buy one for cheap… and an average journalist should just be able to test. Apparently all people who never tested anything before :)

        And you need a portable tool to test while actually driving. And of course done by people who know that stuff on a repeatable cycle. If your test can’t be peer-reviewed or defended in court, it is pointless. It will always take some government agency, or some lab with deep pockets to perform an actually good test.

      • 0 avatar
        George B

        Mark, It is alleged that Volkswagen is out of compliance by more than an order of magnitude. Less expensive tools should be able to measure a huge relative change.

    • 0 avatar
      Luke42

      Why not just hire a certain lab at WVU? They got the equipment for another grant, and graduate students work cheap!

      The other research schools that I’m familiar with (Virginia Tech, University of Illinois) would be happy to do this kind of work for grant money, too. And the costs can be shared by a number of sponsors.

      Then when you review a car for a week, take it by the lab for a few hours to have it spot checked for flagrant emissions problems. I’m sure dyno test numbers would be popular with traditional gearheads, too.

  • avatar
    CJinSD

    Much is said about not being able to devise a real world test for verification of compliance with emissions limits. I don’t think it should be that much of a challenge. The original city and highway EPA dyno protocols were based on a route driven in a California city. The accelerations, cruises, and decelerations were all modeled after two actual trips taken in cars on real roads. I think they cropped out any time stationary, which is why engine stop/start doesn’t help EPA ratings, but otherwise the dyno trips resemble real trips taken 45 years ago.

    Chance are that the actual roads used have changed a lot, but that doesn’t mean you can’t go to a closed circuit with similar climate conditions. Let’s call it Fontana Speedway. Build a test rig mounted on a pickup truck with a boom supporting a hose that can mount to the exhaust of the car in need of testing. Some experimentation can be done to see if this works better with the vehicles running one in front of the other or side by side. Then you just drive the same protocol of gentle accelerations and decelerations as would be used on the dyno.

    It is well known that the EPA’s modeling is awful. That’s why the economy numbers published on window stickers are greatly reduced from the numbers seen on the dyno. They can use those same correction factors to determine if the emissions achieved at Fontana are within expectations for how a dyno tested car will perform in the real world. Problem solved.

    • 0 avatar

      You don’t even need a person to drive a test track these days. At the Ford Proving Grounds, I sat in the back of a Ford Transit that was driven by a robot connected to a computer and GPS. It could reproduce the same route, over and over and over again, until the tank ran dry. It was a violent drive (they were testing for durability) but a lot of cars exist in pretty harsh environments.

      • 0 avatar
        CJinSD

        That’s exactly what’s needed for reliably performing the test. You could have another robot driving the pickup truck.

        • 0 avatar

          There’s no reason, in this day in age, why you couldn’t do that. The only question is *should* you do it? The lab provides a level playing field, something an open-air track can’t.

          • 0 avatar
            CJinSD

            The main reason to do it would be to obviate GPS-based cheat modes that tell the car to perform differently while stationary. It wouldn’t be necessary for every car, but it would provide verification of guilt or innocence when cheating is suspected.

          • 0 avatar
            HerrKaLeun

            I think drag of the car gets better taken into account on a track. Same car with better aerodynamics should pollute less, no?
            The lack of drag is my main complaint about the dyno test. i think they add some artificial load or so, but still.

          • 0 avatar
            qfrog

            Under what conditions is a car allowed to emit higher than normal (legal) emissions? Shouldn’t leaving some degree of condition variability add to the realistic nature of the testing? IF the car fails an on road open air emissions test under conditions that the car would reasonably be expected to experience on a regular basis, then what is the problem? The car failed and would be out of compliance in real world use thus the test revealed that there is a problem needing attention. The mfg should have tested the car in an climate cell on a chassis dyno to replicate plausible real world scenarios.

          • 0 avatar
            CJinSD

            The only legal limits are the ones prescribed by the test. There’s no legal limit for a car at full throttle under maximum load, for example. The tests are supposed to represent a normal service regiment for a car to determine what its average emissions are likely to be.

    • 0 avatar
      Lack Thereof

      The problem with doing a “real world” test, is that everything from air temperature, to humidity, to barometric pressure, can affect tailpipe emissions. Unless you are testing in a climate-controlled test booth, it is impossible to get consistent results. A car that fails the test on one day could pass on the next if the weather changes.

      And for the record, the EPA test DOES accurately model a particular drive. It just happens to be a drive that rarely occurs in reality, unless the “reality” you speak of is being stuck behind a Citroen 2CV in Los Angeles traffic circa 1972.

      The EPA has tests that model more modern driving, and they do perform them as part of the MPG testing. However, due to political pressure from automakers and the like, they do not use those tests for any emissions standards. Automakers have fought every change to the EPA emissions drive cycle, under the argument that it would be “moving the goalposts.” So the EPA’s only politically viable option is to set ever stricter emissions limits on the “stuck behind a 2CV in 1972 LA” FTP-75 drive cycle, rather than more moderate emissions limits on the modern US06 drive cycle.

      • 0 avatar
        CJinSD

        Was it a Citroen 2CV that set the performance benchmark? I’d thought it was a VW bus, but I might have just guessed that was slowest legally available vehicle at the time the test was developed. The 2CV stopped being sold here in 1967.

        The correction factors used to create EPA fuel economy window sticker ratings are pretty large, so I’d think they could be used to extrapolate a range of real world emissions that would allow track tests to be assessed. If a car passes, then it’s good. If it fails based on the track test, then further examination would be called for.

        • 0 avatar
          Lack Thereof

          I doubt they benchmarked any specific vehicle, be it the 2cv or the T2 transporter, but the obvious intent was to have acceleration and top speed values low enough that the slowest economy imports of the time, including those not yet released to the US market, could achieve them. Their 3.3 MPH per second acceleration limit may very well have intentionally undershot the Transporter’s capabilities, because industry figures at the time were swearing up and down that performance of pollution-controlled cars would plummet back to flathead-6 levels, and government officials may very well have believed them.

          As for the MPG correction factors… those huge MPG correction factors are only applied to certain test results. Automakers have a choice when they test for EPA fuel economy. They can either run only the old, slow tests required for emissions compliance, and apply the massive correction factor, or they can also run the new, more aggressive high speed / hard acceleration tests, and average the results of those tests uncorrected, with the corrected results of the old tests.

          Car & Driver did a pretty thorough feature on the various ways automakers can choose to navigate the fuel economy tests. http://www.caranddriver.com/features/why-is-the-epa-so-bad-at-estimating-hybrid-fuel-economy-feature

  • avatar
    gasser

    Let’s just run the cars with a giant plastic bag attached to the exhaust and then analyze the bag contents.

  • avatar

    WHO CARES???

    The people who have these illegal TDi models love them. They bought them with good, liberal, environmental hippy intentions, but realized that they’d been duped. Someday everyone will realize that the Global Warming lie has duped them as well.

    In the mean time, all the punishment of VW will be undertaken by unproductive middle-men with their hands out…many of whom are directly responsible for setting ridiculously punishing regulations in the first place.

    • 0 avatar
      VW16v

      Talk about living in a bubble. Your taste in autos is great. But, sad and entertaining all at the same time when your rant on about factual data regarding the environment. Must be nice to not think about the affect on one’s health or others when it comes to global warming/climate change.

      • 0 avatar

        There is no such thing as “scientific FACT”

        Science is a system and method of measurement.

        There are THEORIES and LAWS – but NO FACTS – simply because we lack the technology and instruments to measure “everything”….things far beyond human understanding.

        Phenomena.

        Global Warming is only an accepted fact by PROGRAMMED, MINDLESS SHEEP.

        I however am fortunate enough to live above & beyond their simplicity.

        -HELLCAT 300 COMING SOON to an Ozone Layer near you!!!

        • 0 avatar
          VW16v

          Entertaining stuff. Thanks.

          • 0 avatar
            WhiskeyRiver

            NASA made the news this week talking about Mars. Apparently there was once an ocean there but “catastrophic climate change” eradicated it.

            One might ask what would cause that. How did the Mar’s climate change without internal combustion engines and evil polluting humans? If Mar’s climate changed then didn’t the other planets change too? What’s the common denominator? The sun? What else?

            Some things are just natural events and we’re powerless to change those.

            That said, nobody I know wants to see air quality in Los Angeles return to the 1960’s smog-laden problems of the 60’s. There’s a balance and we’ve done a reasonably good job with that balance.

            You do realize that there’s an element of the population that thinks the Earth would be better off without any humans at all?

            Balance folks. Balance.

            Take a deep breath…

          • 0 avatar

            WhiskeyRiver

            My problem is POLLUTION =/= GLOBAL WARMING

            CO2 is not a pollutant.

            Sulfur, Lead, NOx and many other elements pollute, but CO2 is absolutely not.

    • 0 avatar
      VW16v

      If you can think before typing. At least look it up before you type it. Diesel exhaust contains carbon dioxide, carbon monoxide, nitrogen oxides, sulfur compounds, formaldehyde, benzene, volatile organic compounds, persistent organic pollutants (POPs), methanol, and other gases. Greenhouse gases released by this exhaust affect climate change. Period. Your statements make it hard to believe you have any formal education. a.m. Radio is not an education.

      • 0 avatar

        And yet I do.
        And I continue to make them.
        And you waste your time continuing to reply to them.
        While I cash in.

        Who’s the fool?

        • 0 avatar
          VoGo

          BTSR,
          Please share with us the name of the institution of higher learning where you became so educated in climate science. I bet there are thousands of people on the internet who want to follow in your footsteps and become a deeply misinformed mortgage broker from Queens who loves to spout off on subjects beyond his comprehension.

        • 0 avatar
          VW16v

          Stick to making high school grade YouTube videos. They are not so bad. Your scientific and political views are purely entertainment for most. A geology degree? Yeah, my brother in-law drives a cement truck. He laughs and says he also has a geology degree. Because he transports rocks.

          • 0 avatar
            Frylock350

            @VW16v,

            I get you are trying to make the “rocks for jocks” argument here with the geology degree; but you’re actually highlighting your ignorance regarding the field. It deserves more respect. Its actually the most relevant to studying prehistoric climate than any other discipline of science. Geologists can use evidence encased in rocks to tell us about temperature, oxygen levels, etc. A geologist can tell us that the fossil fuels we burn are not “dino juice”, but Carboniferous plant matter laid down millions of years before the first archosaurian ancestors of dinosaurs evolved. In fact early reptiles were just beginning to diverge into synapsids (mammals) and diapsids (reptiles, dinosaurs and birds) when our fossil fuels were living. A geologist can ascertain an enormous quantity of observations from dumb ol’ rocks. Geology is like a time machine where the flux capacitor is the layers of rock observed.

            To be clear: I’m not arguing the global warming isn’t very real; just defending a branch of science that’s far more important than you realize. I personally think the evidence shows that global warming is inevitable no matter what we do as we are still technically in an ice age; but why try to accelerate it (which we definitely are)?

  • avatar
    VW16v

    Sorry this was a reply to bigtruckseriesview. Talk about living in a bubble. Your taste in autos is great. But, sad and entertaining all at the same time when your rant on about factual data regarding the environment. Must be nice to not think about the affect on one’s health or others when it comes to global warming/climate change.

  • avatar
    EAF

    “…the only way to get an accurate result of exactly how steep those drops will be is to perform the test on a car before it is reflashed and then run the test again with the same exact variables and settings once the new calibration is in place.”

    Agreed x 10^10! Ideally the before & after would be measured on the same exact dyno/facility, on a day where ambient conditions are identical, and in fact on the same exact cold car.

    I would love to see VW’s algorithm for the cheat. There were talks about the hood being propped up, steering angle, etc but I don’t believe any of that was conclusive.

    • 0 avatar

      I think most of that was speculation based on what happens during the EPA test. However, as far as I know, there’s been no official word from Volkswagen or Bosch on how it is triggered.

      • 0 avatar
        Wheeljack

        I’m surprised Bosch has not been taken to task for their role in this. Regardless of who was paying the bills and what they were asking for, Bosch should not have written software that made the vehicle compliant only under a very specific set of parameters that match the EPA (or other worldwide regulatory agency) test protocols.

        • 0 avatar
          Lack Thereof

          My understanding is that Bosch engineers did not write the programming. They provided programmable ECU’s, VW engineers wrote programs for those ECU’s, and the Bosch flashed the VW provided program onto new ECU’s for delivery.

          • 0 avatar
            Wheeljack

            Ok – good info. I didn’t realize VW had adequate internal expertise and figured they relied on Bosch to provide a complete fuel/aftertreatment/controls solution like many other OEMs selling diesel engines do.

        • 0 avatar
          brn

          My understanding is that most vehicles have a “test mode”. They do it to keep the nannies from kicking in while the car is being emissions tested. If Bosch wrote the code to recognize emissions testing and to go into “test mode”, that’s normal and expected.

          They key is what the car does once “test mode” is recognized. VW chose to do more than disable nannies.

    • 0 avatar
      Lack Thereof

      Documents from the EPA specifically mention that air temperature, barometric pressure, engine run time, and steering wheel position are used to trigger test mode, but failed to go into any specifics beyond that.

  • avatar
    Wheeljack

    I kind of wonder if VW can make the vehicles compliant with just software. If they deliberately used cheat software, then it follows that they may have undersized some of the other aftertreatment components (i.e. the NOx adsorber catalyst and EGR cooler, etc.) knowing that they wouldn’t have the same duty cycle as parts that were in more continuous use.

    If they can make the vehicle compliant with just software, then I would think the vehicle would be in a near-constant state of post-combustion dosing and regeneration cycles to keep the NOx cat scrubbed or burn off soot in the DPF. If that’s the case, the fuel economy will suffer tremendously and expect constant and pervasive coking of the EGR system, variable geometry turbo and DPF from all the soot being generated.

    SCR is probably not a viable solution due to the extensive tear up to the vehicle and the cost. Even the estimates from a previous article of $300-ish a vehicle up front (if designed in from the start) sounded low to me. The solution may lie in a larger NOx adsorber catalyst and a dedicated exhaust mounted doser to spray the required hydrocarbons into the cat to support regeneration. This would eliminate the need to fire the engine’s injectors post-combustion, and it should cut down (if not eliminate) all the potential coking issues with the upstream components like the EGR valve & cooler and the variable geometry turbo. The vehicle will still suffer some loss of fuel economy, but a larger catalyst should reduce the frequency of regenerations since there will be some “passive” regen occurring at times. All of this would probably require a new ECM with additional circuits and drivers and maybe even a small “smart” module to control certain functions since it’s not all just software involved with the fix.

    • 0 avatar
      Lack Thereof

      Bingo. Finally, someone who understands what’s going on here.

      They can make the vehicles compliant with just software, but how that effects the long-term reliability of the components involved remains to be seen.

      Larger sized / more durable NOx absorbers are very likely to end up being part of VW’s eventual recall on this.

    • 0 avatar
      George B

      I predict that most of the Volkswagen TDI cars will never be made compliant. Volkswagen will pay a huge fine and recall letters will be mailed, but owners who like the way their cars run will not get their cars modified. There will be some buy back program to help TDI owners in the CARB states where emissions requirements are tougher.

    • 0 avatar
      HerrKaLeun

      Wheeljack: you nailed it. Exactly my thoughts.

    • 0 avatar
      brn

      You mean parts might brake down prematurely? On a Volkswagen?

      I suspect they’ll have to provide some kind of crazy extended warranty on the affected components.

  • avatar

    It’s likely that the EPA already has data that could be used to determine the difference. When they emissions test a car it’s on a dyno and hooked up to all sorts of sensors. They likely record stuff like wheel horsepower and fuel consumed. Then, we’d just have to compare that data with real world results and we’d have a good idea of the difference.

  • avatar
    Waftable Torque aka Daniel Ho

    I commend them for getting 300k hits in less than a week. That’s like 10x what even Consumer Reports got from their VW discussion, and far above their usual YouTube take rate. Perhaps they can do some Minecraft music videos to pad their numbers?

  • avatar
    Rudolph

    I believe I will nurse and maintain the 2002 1.9l TDI •

  • avatar
    GLInick

    I agree the test is bogus (posted some of the specific flaws on YouTube and tfl’s page).

    But I think it is worth noting that the video is confusing and unless you watch it a few times it makes you think they all worked together on this test. There are 3 parties and its not clear if they all knew what the objective was. The incorrect statements are all made by TFL (voiceovers and on-cameras from Smirnov). It seems that they were driving the “experimental design” (note the air quotes).

    The emissions manager seems to be confused about traction control, and if they knew anything about the scandal they’d know the difference between a PM and NOx test (and done the latter). The dyno shop seemed to also be well below minimal technical know-how.

    But I agree, tfl is a typical media site of dubious crdentials that cares more about building a buzz for itself, and not a whole loy about delivering any real knowledge to anyone.

Read all comments

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • Moparmann: And THAT is why I change my own oil (aside from the fact that it is a simple task that I enjoy doing!) I...
  • Pig_Iron: Agreed, considering how far behind Ford is in EV design and development, this project is nothing but a...
  • burgersandbeer: Count me in too. The latest styling has really grown on me. With NA engines and a traditional roof...
  • Moparmann: I think there are a lot of us out here that are suffering from an advancing case of “OLD”, but...
  • Moparmann: I despise no show socks, if I’m gonna wear ’em, then you gotta see ‘um!! :-)

New Car Research

Get a Free Dealer Quote

Who We Are

  • Matthew Guy
  • Timothy Cain
  • Adam Tonge
  • Bozi Tatarevic
  • Chris Tonn
  • Corey Lewis
  • Mark Baruth
  • Ronnie Schreiber