The car between 2500-4000 rpm, under load, bucks and kicks, afrs read fine, so does rpm, and all the other sensors.

I searched hard (on here and on my car) and found a small vacuum leak at one of my block off plates, but that's been corrected. And it still happens, The idle is steady but occasionally stumbles for a second and then goes back to normal.

new plugs, new wires, everything I could think of

megasquirt v2
board v2
'89 rx7 turbo 'vert


Thanks in advance.

 

sounds like you should look into the tps !

 

that's what I'm thinking, it's unfortunate the car's in my garage two hours away
but, rode my motorcycle over and checked the grounds, vacuum leaks, etc.

The tps values are 52 for the low side
and like 138 on the max.

these vary though, It'll be between 50-68 if I were to go out there and "set" it again using the tools in megatune

and the high changes as well

 

Take a datalog of the behavior and post it. Post an msq as well.

Ken

 

Do u have a 2nd gen? I'm facing a similar problem.

Replaced tbody tps and sort of checked the vacuum.
Adjusting the tps smoothed things out, but still bucks and bounces the rpm on its bad day

 

If it's misfiring, AFR should be reading extremely lean, not normal.

For perspective, roughly the same AFR will change by 3 or 4 units depending on if the engine is running smooth or braptacular. An estimated 12:1 ratio will read out at about 15:1 when the engine's being all bridgeporty.

I have a lot of problems with the FC TPSs sticking. There will be a tip-in lean stumble, followed by a rich stumble when the TPS finally moves. That or the TPS voltage will stutter a bit, something I worked around by raising the minimum rate of change to something like .2v.

 

Yea, the 2nd gen TPS, especially after 20 yrs, is not the best design. When you measure the closed throttle position, wait a second after letting your foot off the pedal, since if you still have the damper, then the throttle will take a second or two to fully close. Also, if you still have the thermowax & warm-up hardware on the throttle body, these will change the closed position of the throttle based on the temperature (until the engine is up to temp anyway).

A very light smear of light oil around the plunger tends to help for me; although oil there can attract dirt & dust, which in the long run isn't going to help you at all.

What AFR are you seeing, b/c depending on where in the map you are, 16:1 AFR can cause some bucking (lean is more likely to do this), especially if you have the overrun fuel cut off on and are going into & out of it. I've also noticed that you can get some hunting & jerkiness when the IAT gets heat-soaked, such as re-starting an engine after 5-10 min that was just running & at temp for a while. For this you need to play with the IAT-based enrichments, but the only complete solution is to re-locate the IAT sensor to someplace it won't heat soak as badly

 

A datalog and msq would make it so we don't have to guess at what is causing your problem.

Ken

 

2011-04-17_14.54.54.zip



Replaced the TPS, solved my problems,
noticed that "TPS ACCEL Enrich" would blink on when it stumbled at idle.
also the min/max were never the same I'd occasionally see -10%tps

Seemed to have solved all my problems,
came to a stop and the idle started dropping to 0, jumped back up, down.
on both my tach in the car and the ecu

Problem is I can't cause the car to recreate the scenario. I was quick and have a data log of it. This has happened three times, all after different lengths of driving, all at operating temp, fan on and fan off. Considering it could be a dying Crank Angle Sensor. Shielded wire, grounded at the megasquirt end is used to carry the signal to the megasquirt.

Caught the tail end of the issue, but it was steady, and the car just came back to normal idle.

showed a repeated lost sync reason: 17

I searched and saw that this could be noise on the primary trigger, or not seeing the secondary at all. I'm thinking that if it was based on noise I'd see this more often than "once in a while"

 

The only things that can cause that lost sync error are the issues I mentioned. If it's not happening that often, it's more likely a loose wire type of issue or similar than it is to be a noise issue, but there aren't that many situations that can cause that kind of error.

Ken

 

I'll look through my harness this weekend with the multimeter, wiggle things around, see if I loose continuity. Double check the connectors that the CAS uses and make sure there are no cold solders.

 

I had a cold solder in my cas connector, so I repaired it a few days ago and things were working great.

Last night, my car's pulsewidths occasionally blipped for just a moment to 0.0. causing a stumble and then it went back to normal. Well later last night it decided to get progressively worse. Surging, always stopping at idle, and then just after a few moments, die. The car now just starts for a moment, idles at 800 for seconds and then cuts off. (so no log) The RPM signal didn't skip like a connectivity issue. The TPS accel enrich and/or TPS decel lights would flicker when the "stumble" and pulse width problem occurred.

I just replaced the TPS with a good one, but last night I noticed that there was a new value for the "max" tps setting. I distinctively remember 158 being the max for this new tps, but it was set to 205. I was using Tuner Studio's "Ve Analyize Live" to tune so maybe the tps% difference caused the car to lean itself out too much?

I'm going to grab the VE table from Arroncake's writeup and try to start it with that. If that's the case then I'll increase the fuel in my warmup, although this occurred with a fully warmed up engine.

 

Flashed my firmware, to 2.1.0d
got my car started again, idles erratically.
new tps is acting funny after a brief period of stability

I have a log, the pulsewidths are still jumping up.

posting Data Log and the MSQ that ran my car up until last night.
still tuning using VE analyize live
2011-04-17_14.54.54.zip

 

You've got 157 sync losses in that log.

 

2011-04-25_19.36.18.zip


looks like I posted the same data log twice, here's the correct log reflecting the harness correction and the MSQ posted above

 

That log is filled with TPS accel enrichment events. Your TPSdot value is all over the place, while your TP value is relatively constant. I've dealt with this by putting a capacitor on the TPS input signal. I've often wondered why this isn't part of the MS design.

 

rate pw adder/ms
210 2
400 5
800 10.5
1540 15.0

I haven't really messed with my TPSaccel/MAP accel parameters yet,
The TPS threshold was initially set to 200% and the graph seems to be very high, but that's just at first glance.

I should probably just have it disabled while tuning (wouldn't the 200% disable it?). If not the the software is probably compensating for the accel enrichment on the ve table. I'm reading that section of the megamanual tonight. and a few posts I found on here.

do you have a recommended value for the capacitor?
or does the megasquirt see the signal in Hertz?
then I'll just figure out a low pass filter @ 170hz 5kohm

 

MS just reads the voltage coming in on that line, its not expecting a pulsing signal (unless you're actually pulsing the gas pedal anyway). The 200% is 200% of the throttle range/second, which is still possible to see while driving, especially with noise

 

thanks for explaining that value to me

it's obviously noise, but I'm not familiar with how noise gets worse over time. the driveability of the car just plummeted over a few hours, from tuning and cruising to not even starting.
or maybe Tunerstuido's VEAnalyize Live didn't disregard a/e

I'll set the threshold to something ridiculous so it doesn't engage at all (1,000?), add a narrowband filter to the tps signal, to eliminate and see where I'm at.

 

I don't think the re-start issue is noise; noise tends to be much more rpm-dependent than time-dependent. My guess is that you don't have the warmup enrichments correct & were compensating with the VE table being rich, and VEAnalyze Live leaned out the table to the point where its too lean to start. Can you compare the VE tables from before & after this last tuning session, and see what changed, how much, and where on the map it is?

 

I've yet to tune the Warmup enrichment table, as I thought you had to have a reasonably right VE Table to get it right.

VeAnalyze has a filter which disables it below 160*F or when Warm Up Enrichment is engaged. Unfortunately I just got a new laptop as my old dell died, so the only reference I have is Aaroncake's "Starter" base map / VEtable

I've got it started this morning, idles the same as in the last datalog, with some serious TPSDOT fluctuations with no tps movement, as well as the TPS Accel Enrich light coming on during stumbles

 

I'm going to disable the accel enrich for the time being and get my VE table sorted out, Eagle7 advised me to add a 10uf cap to the board, in parallel with c9, this cut down the noise a bit, about by half. I'm lowering the TPS lag settings in the general settings, so the value will have a bit more of a resistance to change, hopefully smoothing things out. If I can't get it out, my MAPdot signal is way smoother, so I'll give mapdot a/e a try