I just thought I would cross post these captures here in the PicoScope forum from the Technical Duscussion area so those who are not members here yet could have the opportunity to see the kind of performance possible with PicoScope on some modern drive by wire signals.

Victim:  2006 Chevy Cobalt 2.2 Vin F.


Brief system description:

The gas pedal is an Accelerator Pedal Position Sensor (APP) which serves as a throttle request from the driver to the ECM.  The ECM then controls the throttle plate position with a reversing electric motor in the Throttle Body Assembly.  The throttle plate is spring loaded and has a neutral 'rest position'.  The motor must pull against the spring to close or open the throttle plate from this position.  Feedback about throttle plate position is provided to the ECM by two Throttle Position Sensors (TPS) which work with opposing voltage outputs.

We are going to focus on the TB assembly signals.

In the following captures the channels are:

A: TPS1  (voltage falls when throttle plate opens)
B: TPS2  (voltage rises when throttle plate opens)
C: TAC Close (voltage on close side of TAC motor)
D: TAC current (probe is oriented so that positive current is a close effort and negative current is an open effort)

The captures in this post were done with a PicoScope 3423 running 6.0.6 software.  We are collecting samples at 1MHz per channel on each of four channels with two second screens.  It would be impossible to duplicate these captures with any other automotive scope.  The TAC control is modulated at 10KHz!  So, we are looking at signals with widely different frequencies with all of the detail available.  The data files are enormous.  One is 120Mb.  A single screen is 6-10Mb.  So, we will just do some images.  

The first image is on key up.  Note how the ECM closed the throttle plate to check the min TPS value then finds it's 'happy place'.

The second image is a snap throttle capture with the engine running.

Note that the motor modulation is reflected in the TPS signals as the TPS picks up the RFI from the motor action.  Also note the digital pulses on TPS2.

[smiley=nerd.gif]

http://www.autonerdz.com/images/tar140.gif



In the following images, we have exchanged the Channel A TPS1 for the TAC Open side of the motor to show more of the motor control strategy.  Again, the first one is a key up sequence and the second is a snap throttle with engine running.  Note how the voltage is applied to one side of the motor and modulated on the other side to achieve the desired position.

http://www.autonerdz.com/images/tar141.gif


Here is one more image set below.  The first one is a capture after the key has been shut off and the TAC system is doing it's self test and pulling the plate closed to again check the minimum throttle value.  Then the motor control shuts down and the throttle plate spring opens the plate to the 'rest' position.  

The second one is several seconds later when the TPS2 circuit finally powers down.

The third one is a mega zoom in on the TAC motor control modulation as the plate is pulled closed.   Note the TAC Close voltage is high and the TAC Open voltage is modulating.  The TAC current is a positive value average indicating the effort to close.  This current modulates with the TAC open signal, rising when the voltage is low (from TAC Close side voltage) and falling when voltage is high (opposing TAC Close voltage).

The time cursors frame one modulation signal cycle at 100us.  This is 10,000 Hz!!  You could not make this capture time and still be able to zoom in to see this detail with any other automotive scope.  That is because the samples are only one microsecond apart for two million samples per channel during the two second screen capture.  This kind of scope performance is unprecedented in our industry.

http://www.autonerdz.com/images/tar142.gif


To illustrate the zoom power I have made some boxes to frame the area we are progressively zooming in on in the following images from one of the above captures:

http://www.autonerdz.com/images/tar138.gif

http://www.autonerdz.com/images/tar139.gif

http://www.autonerdz.com/images/tar143.gif


Get the idea?  In that last zoom, we have measured one of the pulses at 8 microseconds (8 millionths of a second).  The scope has collected 8 samples during that time for a detail in resolution never before seen in an automotive scope.  These captures could not have been duplicated with any other automotive scope available anywhere.   [smiley=disimulo.gif]

Hi Tom
 When reading something like this in the forums it can be a real pia having to scroll all the way to the bottom so I can scroll to the right and read the right side of the sentence in the very top line.  Is there something I can do in my settings to to avoid having to scroll from top to bottom and back every time I want to move to the right and finish a sentence?

Sorry  :o

Those images were from my wide screen  I shrank them down so that should fix it.  

Thanks Tom. It's not just your captures that do that. What you did doesn't completly cure it but it helps a lot. Ican read far enough to the right that I can fill in the blanks for the most part. The forum doesn't fill the whole screen though. Changing the pixle setting on the desktop wouldn't change it.  I guess I'm way behind the times with this 18" monitor?  Below is what I see looking at the forums.

---

Ronnie,

You must be running 1024x768 on your monitor.  I eliminated the forums container side borders for more space.  That should fix it for you and everyone else.  Thanks for the help.  The site is designed for 1024x768 or higher resolutions so I wanted this to display right.  Getting rid of the wasted side space should help a lot.

Of course, always maximize your browser window for best results.

Since you are a member, you can also read the original thread here:

http://www.autonerdz.com/cgi/yabb2/YaBB.pl?num=1180471365

Tom,

Can I download a pico software "demo" to view your pico files?

You are right, those signals are really fast...

Originally GM TAC failures were software, voltage, and component related. Now we are seeing wiring faults. Sometimes those wiring faults are easy to find by simply wiggling the TB connector and listening to the response. Replacing the pigtail takes care of those. However... then there are those (typically 2003-2004 trucks) where the harness is failing VERY VERY VERY (did I mention very?) intermittently.

The only way to fix those is to replace the ENTIRE harness from the TAC module to the TB. You can only get a pigtail for the TB end. The other end MUST be rewired completely from the wire to the metal terminal. I replace the puny wiring with 18 gauge wire.

I would guess that the ONLY way to catch these problems with a scope would be with a pico but even then, we're most likely talking about intermittently degraded signals, not MISSING signals... So my hat would be off to someone who could catch that! Not saying it could not be done, just that it would be difficult.

The newer vehicles STILL use the puny wiring and so as they age a bit, we'll see even more wiring problems. I am looking forward to seeing a pico capture of one of these intermittent faults!

Lastly, I would strongly recommend AGAINST piercing the TAC wires... While there are other options, one could get a TB pigtail and put terminal ends on it to connect to the vehicle wiring.

Jim,

Here is the demo link. http://www.picotech.com/demo.html

Thanks for the heads up on wiring issues.

Oh, by the way...we didn't pierce any wires on that rental car. We did manage to break it though.  :)

Thanks Carl.   ;)

Jim,

Be sure to choose the automotive Pico 6.  You can also get the latest demo from this page:

http://www.autonerdz.com/picokit.htm

PicoScope 6 In demo mode cannot show the high speed streaming capabilities used for these captures with the demo signals.  You have to have the hardware to experience that.  However, it will open up the psdata files so you can examine them just as you would if you did have the hardware.  Just choose the appropriate device when prompted.  Example: choose 3423 for the four channel demo or to open a 3423 capture.

As you can imagine, the files for these captures with 8 million 12 bit samples are HUGE.  This one is about 115MB.  I am not posting a link to this here so as not to overload my server, but I will make it available by request.  Just email me and I'll send you a download link.   8-)

Then you can open it and zoom, measure and get a really good feel for the unbelievable.   :o