Fluctuating TPS Signal

[rabellsam]

Hi everyone,

I am from the Chico State Formula team and we have been having some problems with our TPS readings. First, some background. We are running a 2007 GSXR600 engine with a carbon fiber intake and aluminum throttle bodies to hold the injectors that go into the boots on the engine. Our TPS is mounted on a stock throttle body that we took off the bike and plugged all the holes that has our aluminum restrictor in it and is mounted on our intake. I can add a picture i needed for clarification. We are using a power commander III to do some tuning and there is an option to set the zero and full throttle positions on the sensor. The value that we get for zero throttle when the engine is idling can be anywhere from 1600 to 2600 (for reference full throttle is usually around 22000). The ECU reads a voltage from the TPS so something must be causing some noise in the signal. We have thought it might be that our throttle body was getting some static electricity from the air going in because it wasn't grounded. We added a ground though and that had no effect. We also thought heat from the engine might be causing some noise so we moved the ECU farther away from the engine. We are about out of ideas of how to remedy this and any help would be greatly appreciated.

Thanks in advance,
Sam

[Swiftus]

Sam - Do you know if your throttle position sensor is a wiper-type potentiometer? If so, it may be worn out. Also, double check that the return stop on the throttle body is clean. A spec of oily dirt can change things so easily.

HTH

Jay

[rabellsam]

We are using the stock sensor from the bike and I believe it is the wiper type. We recently switched it with another one off a different bike, same year, same model, and noticed no change. And by the return stop are you referring to the throttle stop screw?

[Shuff]

have you checked all connections/crimps?

[rabellsam]

We are using the stock wiring harness from the bike and haven't messed with that besides lengthening some wires to reach our components. It came from a running bike so I would assume that all of the connections are fine, but you never know.

[Andrew Richards]

First, how do you know that your TPS signal is fluctuating? Do you have a data logger that you are using to monitor engine parameters? If so then post the data. Have you connected an oscilloscope to verify that it is a noise issue? If not, hook one up and post that data. Was the sensor once working asdesired and this is a new issue on an old setup? Or is this a new setup all together that you are still working the bugs out of?

Whenever I think I have a noise issue (I rarely do) I wrap the sensor in question with some good ol aluminum foil....

[rabellsam]

We are using the power commander to tune the engine. It give us throttle percentage and also the ability to set the low and high values for the TPS. I have continuously been changing the values to match what readings we are getting from the power commander. This is just with the engine running and the car not moving.

Haven't used an oscilloscope but I will see if I can get my hands on one.

We have been having this issue since we turned the engine on the first time. This is our teams first car.

Thanks for the reply.

[apalrd]

A few things I would do:

-What is the range of this number? Measure the voltage with a multimeter and see what the voltage of the wire is at 0% and 100% throttle.

-I'm not familiar exactly with the Power Commander, but what kind of ground pins are available? If the ground path for the signal goes through a wire that carries current, it's possible you will get some potential (voltage) difference on the ground to shift, changing your readings (it's really measuring the voltage between TPS and Ground, so if you change the ground you change the reading). Our (standalone) controller has separate power and signal grounds, so all of the sensor grounds come back to the ECU and it measures across them, so any current across the power ground does not cause the sensor readings to drift. We actually have 7 separate ground pins on our module.

-See what happens with the engine off. If it stabilizes then you have either the floating ground issue above, or noise from another system. If it's noise I would look at ignition system first, then the rectifier, and possibly add EMI filtering caps to the power wires of one or both.

-If all of this fails, I have some more obscure thoughts for later.

[rabellsam]

Ok we just went out and drove and I have a couple answers.

The TPS has a range from 0-5v and 5v is at 0% throttle. The power commander taps into the signal wire that goes from the TPS to the ECU. The ECU has a chassis ground and an electronics ground and the TPS is on the electronics ground. The power commander has a chassis ground.

When we were testing I noticed the signal was steady when the car was off, but once we started cranking, it jumped all over the place again. Do you think the power commander being on the chassis ground could have an effect on the signal going to the ECU?

[Swiftus]

If you make the grounds common, your problems may go away.

[Luniz]

+1 on the common ground...

Why are you using this "power commander" thing and not a proper ECU?

[murpia]

Hi,

I think the ground current issue is likely to be the cause, as well. But, it could also be related to a drop in the TPS supply voltage, which the ECU itself is robust to because it uses a ratiometric calibration. You may be able to tell which it is by the direction the error goes when high currents are present in the various ground lines.

'Proper' measurement of a signal (in this case TPS) referenced to a different ground requires a differential measurement. I have never worked with a Power Commander, but I have worked with shift systems that need a clean TPS from the engine ECU. These use a 2-wire differential measurement, so you connect the + line to the sensor signal and the - line to the sensor ground. If the sensor ground floats relative to the shift ECU, this does not affect the measurement.

However, this differential measurement is still absolute, not ratiometric, so it does not fix the second possible cause. Since I have not seen a shift system that also senses and compensates for 5V issues, my assumption is that this is a less common problem.

Hope that helps, Ian

[Kirk Feldkamp]

Dare I ask how much "jumping all over the place" constitutes? If your idle is "2000-ish" and full throttle is "22000-ish", you're probably going to see plenty of movement in all the way up to the 10 or even 100's digit with the engine running/vibrating/etc. In other words, with that sort of resolution, the number you see is going to be sensitive! Does the situation improve when you're holding a relatively-constant part throttle opening? Or is it just that the TPS signal is dropping out or moving erratically? In my experience, if you assume wiring from an unknown origin/background is "good", you're asking for trouble. Diagnosing connections and sensor integrity is always the first thing I go for. Frankly, with a 3-wire pot, there's not a lot that can go wrong, and usually the solution is the most obvious thing possible. All this talk about grounding or noise seems like a diversion from what is more likely going on (bad sensor or connection... or nothing at all). Since you're using the stock ECU, throttle bodies, etc... what did you do with the secondary throttle shaft? I can't remember, but isn't there a second position sensor within the secondary control/actuation module? If you're running a PowerCommander with a stock ECU, it's going to be expecting that stuff to be there, right?

It's also unclear (to me) if you're following the rules with regard to the throttle body placement within the system. Are you using the stock throttle body for actuation, or do you have a separate, single throttle body upstream of the restrictor? Just want to make sure so you don't have any surprises down the road! :)

-Kirk

[rabellsam]

Made a little bit of progress today. Tried tying the chassis and ECU ground together but no change. Started going through the diagnosis in the manual for the TPS and it is getting 5v as it should and all three wires are continuous. The one thing that was off from the specs was the resistance at full throttle. At zero throttle, it is supposed to be 1.1 k ohms and full throttle is supposed to be 5 k ohms. We were getting about 2.5 for full throttle. We also tried pulling the Power commander out just to see what would happen and it didn't seem to make anything better.

Then I had what I hope was an aha moment after I got home. We aren't using the secondary injectors or throttle valves or TPS. The secondary TPS uses the same ground as the primary and we have nothing plugged into that connector right now. Floating ground? That's what I'm hoping. I'll try plugging a sensor in there first thing tomorrow morning.

[rabellsam]

Luniz, we are using the power commander because it is what we had. First year team low budget and inexperience issues. We found out there is a Motec that we can use but we didn't want to risk messing things up before competition. Our car still starts and runs as it is, even if it is not great.

[rabellsam]

Kirk, first off, we have an upstream throttle. We took one off the original intake and are using that with the restrictor just behind it. The throttle bodies we made that go into the boots on the engine have no butterflies in them, their only function is to hold our intake on and hold the injectors.

As far as fluctuations, when the power is turned on, but the engine is not, it fluctuates roughly 30-50. I haven't seen it jump a hundreds place with the engine off. With the engine on, I have seen it move a couple hundred values at a time which can move it from 0 to 2 or 3 percent throttle. These happen all the time, at zero throttle or a constant nonzero position. I don't think it is a sensor issue or a problem with the stock harness. There have been two sensors and two wiring harnesses and two engines that have all been mixed and matched and we have had the same problem the whole time.

Since we made the throttle bodies going to the engine, we didn't put and throttle valves in them. The stock throttle body we are actually using the throttle on has had the secondary throttle valve pulled out and the holes have been plugged. Now the STVA and STPS are unplugged. There ECU might be unhappy about that.

[rabellsam]

Just plugged in the STPS and STVA with no effect, will continue trying the other suggestions.

[rabellsam]

Looks like we got it figured out. The Power Commander had a chassis and ECU ground that were both grounded to the chassis. Switching the correct ground to the ECU fixed the problem. Thanks for all the help.