I got the latest firmware for the older generation tec3r. It is from 2010 I believe.
Yes, I too have a suspicion that it is a bug in the FW.
Electromotive can upgrade the unit and prepare it for wintec4. But it requires me to ship it to them. Looks like that is something I need to do this winter.
Alternatively replace it for a more modern system.

I thought it would be a good idea to upload some print screens from two logging events. One with OK revs, one with bad. (Done while revving with a stationary car.)

So here goes:

First, the good one:
http://i132.photobucket.com/albums/q...psq1niro9l.png



The the bad one:
http://i132.photobucket.com/albums/q...pspylgmiey.png


I can't make sense of it. It looks very similar, only the bad one shows the slight hesitation in the RPM, and that the AFR goes really lean...
Hope someone else can figure it out?

I'm a Megasquirt guy, but here is something to check....

Sometimes things like accel enrichment and idle control are turned on/off by TPS set points. I found that occasionally that my TPS didn't always return to 0 (zero) due to linkage, etc. When I changed the TPS set point in a few of the setup screens a little further from true 0 ,mystery/illogical things started to go away.

Anyway, just something for you to consider.

Good luck!

Otto, I'd like to see your bin file before possibly commenting on you data log.
Johan

My major mystery with this system on my 2.8L race motor was due to the TPS. If it started at, say, 1V, on depressing the pedal it would drop, say to 0.5V. This was most pronounced driving the car, where there was a load. Needless to say it was hard to get it going, but when I coaxed it out onto the track, where the RPMs never fell below 3,000, it worked OK.

Why the potentiometer behaved that way was a mystery, but the solution was to make, say 1.25V be the nominal "closed" position, with what WOT measured as the WOT position, and let it draw the line between those two points.

In fact, with advice I hogged out the mounting screw holes in the TPS so it could be rotated away from what seemed a bad area. In the fullness of time I spent a lot of $ (for the TWR TBs it is some slightly rare German Ford part) for a new one. It had much the same characteristics, but at least I was looking for something like this.

Yours probably is fine, but one more thing to check - if the throttle or the sensor does not always return to the exact same place, that might cause your hesitation?

Thanks again guys!

Well, as I mentioned earlier, the TPS is pretty solid, always landing between 0.90 - 0.94V.
I can't imagine that this would make a huge difference. I am a technichian in electronics, so I know that making a voltage divider like the TPS is using a resistor and potensiometer is never going to be 100% accurate. So I can't imagine the Electromotive guys designed the system so that it cant handle some "tolerance" deviation from the potensiometer itself and the battery voltage...
Anyway, I have played around with the setting of the Idle voltage, and it looks like it does nothing. I suspect all it is ment to do, is start the "idle programme", wich I am not using. I am using PMO ITB's so no such controls apply.

I could send you the .bin file if I knew your mail adress. Send me a PM...

If there's a new FW and you're confident the TPS is good to go, I'd try the new FW. Of course, I work hard stuff to easy. ;)

In my tecGT with the older software, Electromotive had built in a .08 buffer on the TPS that they do not really clarify. So, in effect you need to add this to your TPS zero point to ensure the EXU considers the throttle closed. So, for example, if your TPS zero's at .94, you would need to set your TPS zero to be 1.02 (.94+.08) to ensure the ECU recognizes the TPS is at idle. You may want to try changing this value. You can always change it back later...

Looking at your graphs I see your EGO corrections seem to be quite high in both directions in good and in bad graph sets. You may need to keep tuning your AFR map and/or your fueling map. In the good set of graphs the AFR seems to bounce around within what you would think are good values (between 11 and ~15?).

When I look at the bad graphs, it looks like the AFR's are out of whack and it looks like the EGO correction is much more in play. Is it possible your O2 sensor or your EGO functions are not working consistently?

If you trust your fuel map enough, you might try turning of EGO correction. This would force the ECU to rely upon your fuel map without correcting it, and I think it would more or less also ignore the O2 sensor, I think. May be worth looking at your options here, as between the two graphs sets, it seems these graphs are quite different.

The corrections may look large, but they are only max 3% from map. So no real big difference is made by that. Its just enough to pull in some adjustment to the AFR when cruising. I have also tried turning i off. But no difference to the issue.

I tried that voltage when playing around with it earlier.

The second set seems to show the AFR's are off quite a bit compared to the first graphs.
I would look in that area - O2 sensor, fuel pressure or fuel regulator, fuel filter. The readings in the second set of graphs seem to show more lean conditions than the first set.

You need a good TPS reading all the time as it triggers your enrichments. The car does not have to be running. Pull up the main screen and watch the TPS voltage, it should be consistant all the time. It should never vary at closed throttle. If it does, it's bad. Replace it and data log again.

PM sent
Johan

your AFR is significantly leaner in the 'bad start" trace. At 1st glane, the other parameters seem to be consistent. Can you monitor your fuel pressure during these logs, to verify that the pressure is constant and in spec?

regards,
al

Great to have this much response in here. None so far from the electromotive forum...

I have monitored the fuel pressure gauge when activating the accellerator by hand in the engine bay. The fuel pressure needle is dead steady... always. So that looks to be OK.

Some thoughts to the Idle voltage theory:
If the Idle voltage really is supposed to be dead accurate down to the precise second decimal every time it goes back to idle, then I can't imagine I'm ever going to find a TPS good enough. You can't expect the potensiometer resistance be 100% accurate from one movement to the next, can you? I mean, there will always be some discrepancy...
If my memory serves me, I observed the same thing when I used the previous TPS on the 3.2 engine. It "lands" at random between for instance 0.86 and 0.82 each time the pedal is released. In my case now it is between 0.90 and 0.94. But it does not vary once it has landed. Then it is steady.

In an effort to solve this issue I have now made a complete rewrite of the entire map, using different UAP and TOG. Aditionally I tweaked the MAP figures in the left coloumns, so the map now effectively has a more detailed "image" of the area that is used for normal operation of the engine. (if that made any sense? ;) )
I also turned off the closed loop and tweaked the accell enrichments.
Anyway, I took some more logs, and it looks like you can now see the difference even more from a good one to the bad one:


Good one:
http://i132.photobucket.com/albums/q...ps2ccrdo1e.png



Bad one:
http://i132.photobucket.com/albums/q...psxlsdxe3o.png

Looks like they both activate the injectors equally from the PWON figure, yet it still runs lean. THAT is probably the real key to this issue. But what can cause that? I mean, it looks like the ECU is activating the injectors correctly. (but are they really activated like the log says??)
-Very strange indeed...

I am guessing you are using batch injection and you may have two injector wiring leads - one for each bank of cylinders. You may want to trace the injector leads back from the injectors to the ECU to make sure one is not crimped, cut, or in some other way "bad". Are both injector harnesses (assuming there are two) connected into the harness solidly? Both getting good voltage? Not sure how those are pulsed by the ECU, but perhaps there is an issue in the signal for one of the injector banks. Hope you find the issue soon.

Vacuum leaks can cause different running. Lawd knows I learned that the hard way!

Sorry, I haven't closely read all of the messages in this thread. What jumps out at me: sequential injection, and 50% of the time it runs well. Are you using a crank trigger or cam trigger for injection timing? If crank, there is a 50/50 chance your injection timing is using tdc versus bdc to start the timing sequence. A very noticeable effect in my old megasquirt days. Spraying while air is rushing past the intake valve versus 180 degrees out and the fuel ends up wetting the cylinder walls. Same injection times, different afr's. Something to think about. Good luck!

All good points, I'll need to investigate some of that a little closer. :)
A little explenation to your comments;
I'm using sequential injection, and they are syncronized by a cam sensor. It is working fine, I know this because I was thinking the same thing as you, nicd. :) Belive it or not, but I have the same issue even if the cam sync sensor is disconnected. -Wait, maybe this is what the "bug" is!? (In the ECU?) Maybe it appears to sync up, but the reality is different?

A vakuum leak is tricky sometimes, but I can't see anything wrong on the MAP though. I imagine I could see something there. Besides, there is really only one place a PMO ITB can leak; the brake booster connection from runner no. 2. (Or to the MAP sensor, but it seems to be solid)

So I am going to check out the possible leak anyway, along with the power supply to the injectors, checking the resistance in the wires all the way from the ECU.

Most screens look consistant, but there is quite the difference in pulse width at idle. There id also a difference in battery voltage, this will change your pulse width. Go into the file and get rid of the battery voltage compensation and see if you see a change.

Post a copy of your accell enrichments.