Code 2.89 Problem

[DoD70]

I just moved up from 2.87 to 2.89 and now my uS is not functioning properly. When I power up, it only cycles the fuel pump on and off every few seconds. The tach in MegaTune while this is going on bounces from 0 to 8000 and back. The car refuses to start. What have I fried?

[DoD70]

Not sure what I did but it is working again. Just like it was before. Which is to say that it still will not rev past 5000 rpm or so. I was excited about moving up to 2.89 when I read that this problem was addressed in the new code but apparently something is still not right with my setup. I have swapped out the diode with a 1N4148 like I read in the VR sensor thread. C30 is still in place. Racing season starts soon. Where do I go from here?

[cmonref]

Did you load the INI for 2.89? Here's what Lance posted for someone else about 30 minutes ago: [edit: well 6 DAYS and 30 minutes ago ...]

"....... Use 2.890 instead. Also use the INI that goes with 2.890. You will find the code and INI here: http://www.microsquirt.info/uscode.htm>

[DoD70]

The correct .ini is loaded. The problem with the engine not revving past 5000 or so does go away on its own after the motor has been running for 20-30 seconds but it is not WUE related. Cold motor or hot, WUE on or off, it acts the same.

I have other issues that may or may not be related. Last year I had a stalling problem that I traced to a flaky MAP signal, accel enrichment and lag factor settings. Even with the MAP sensor disconnected, the MAP value would change constantly, rapidly and randomly. This movement would trigger the accel enrichment, flood the motor and cause it to stall. I masked the problem by tuning for TPS only accel enrichment. That was last year. Now the TPS signal is acting the same way as the MAP even when the TPS is disconnected. Wiring is good, connections are soldered and shrink sleeved. Not sure where to go on this one.

[grippo]

The problem you had with the new code is that when the code is freshly loaded it has no way of knowing what hardware it is in: MS2 or microsquirt. That makes a difference as to how it sets up its outputs so as not to leave injectors on, coils charging. So it cycles the fuel pump and rpm to let you know you need to either download your msq or set the ECU Type. It started working because you did one of those two things. So what you experienced there had nothing to do with your problem or any flakiness in code or hardware.

As far as the rev limit of 5000 for the first 20-30 seconds, does it keep doing this in the sense that once you get past 5000 rpm, does it run ok ? When you drop down below 5000 and then try to go back up, does it again have a problem ? Do you have to hold the throttle down and it sits at 5000 for 20-30 secs then goes past or do you keep pushing the throttle up and down until it goes past ?

Did you do a datalog when this happens ? It doesn't need to be long - just datalog the problem one time and look at the output or post it. We are looking for whether the rpm drops to 0 or the trigger count +/- changes which would indicate it is a wheel trigger problem, although I have never heard of what you are experiencing. If its not a trigger problem we can look at what is changing between the time it is sitting there trying to go past 5000 and the values after it gets past 5000.

[DoD70]

You were right about setting the ECU type and downloading the .msq file. That was what I did to fix that problem. Unconsciously of course...

As for not wanting to rev past 5000 rpm, the datalog below shows that it actually does not want to rev past 4000 (5000 was an estimate) and it does seem to be connected to WUE after all. I was wrong in my earlier post. You will see in the log that just before the coolant temp hit 160F that the rpm will not spin past 4000 and just after 160F it does. When I start the motor where the coolant temp is already above 160F, it will not spin past 4000 until the ASE has completed its cycle which is about 20-30 seconds.

The good news is that the problem with the jittery MAP and TPS signals that were falsely triggering the AE and causing the motor to intermittently stall from too much fuel is solved. It turns out that the 5vdc supply for the sensors was dirty for whatever reason. I installed a large cap (4700uf 6.3wvdc) across the +5 and signal ground and the problem went away. Yeah, it's big but that's what I had laying around. Just to be sure I filled the fuel cell, fired the motor up and let it idle until the fuel ran out. It idled for over an hour and never skipped a beat. As for the wiring, the MAP, TPS and CLT sensors are the only ones wired to either the +5 or sensor ground. Disconnecting the sensors made no difference in the jittery signal I was seeing while in the acceleration wizard and the AE was randomly triggering by itself. The cap put an end to all of it.

[DoD70]

Worked on this all week. Still can't get it to run right. I'm at my wits end. Anybody have an idea why I can't spin past 4000 rpm until the WUE is done?

[Bruce Bowling]

Worked on this all week. Still can't get it to run right. I'm at my wits end. Anybody have an idea why I can't spin past 4000 rpm until the WUE is done?

OK, there is nothing per se in the code that is different when WUE is off, this is why this is a puzzler.

Thinkgs to try:

1) Make sure the WLED signal wire (pin 17 on ampseal) is not grounded or going anywhere. In fact in the s/w, I would try to re-assign this output to be something else (Port I/O section) other than WUE, this will eliminate any affect on this from a hardware stance.

2) I believe the WUE threshold is 160 degrees, and from what I read if the temp is 160 or above you can rev. So, if the temp is 158 or 159 it will not go above 4K RPM? Is it this granular? Its important because if it is then it really suspects of something like #1 above - a hard threshold of work/don't work.

3) No rev limit mode set anywhere?

- Bruce

[grippo]

This is very interesting. In the rpm attachment I plotted the rpm from your datalog and as you can see, every time you try to run high rpm there are dropouts where the rpm goes to 0. These are resynchs caused by VR sensor noise or dropouts. At the end they cease and you can go way past 4000 rpm. This seems to show that they are only indirectly correlated with warmup. The trigger plot is even more interesting - it shows that prior to 1500 (line no. in excel) you are missing VR pulses, and then starting at 1500 you start getting a lot of extra pulses from the trigger wheel. However, these are being thrown out by the code, as they should be, and you are able to rev freely.

This is really weird - I know that VR sensors are temperature sensitive, but not like this. The plots lead to several questions - why is the trigger failure mechanism changing like this. A closer look at the data showed me that the rising trigger count is probably due to the missing tooth gap which Bruce and I and many others know well can start inserting an extra very low amplitude false tooth in the gap above a certain rpm - typically around 4000. This should be removed by fidling with the VR circuit components, but with the fix in 2.890 it is being removed by the code and quite well - there are no misses after 1500 even though the trigger count is skyrocketing.

So the real problem here is why there are missing teeth at colder temperatures. These are not being compensated for because they are probably random and the code will not compensate for more than one consecutive miss or extra pulse. With the missing gap extra pulse, this is very consistent and is always just one. The missing pulses during warmup are a totally different mechanism and I have never seen it so related to temperature. I can't conceive of anything in the code that might explain this. As Bruce said, the sections where the VR pulses are detected and processed have no connection to warmup temperature or warmup mode, but I will double check to see if I can find anything.

[grippo]

In looking at the code I believe I see something that mgiht explain a lot. The pulse tolerance used in VR sensor checking is changed after warmup. So this probably the explanation for the change in trigger+/- behavior as soon as warmup temperature was reached. Try setting this tolerance the same as the normal running tolerance and then see if you can go past 4000 rpm before warmup.