Misfire issues, narrowed down to MS I think

[John L.]

I'm running Megasquirt v3.0, MSnS-e 029q2, MT 2.25 release. It's an 88 Ford Thunderbird Turbocoupe 2.3 Turbo. All of the TFI wiring is shielded and separated from the power wires as much as possible. I'm using the VB921 (driven from LED17) to drive the TFI module, along with a pullup resistor.

I was having the problem before switching to 029q2 so I know it's not a software issue, all settings have been checked and double checked, all ignition parts have been replaced. The first place I started a thread was here on Turboford.net:
http://www.turboford.net/ubb/ultimatebb ... 2;t=041864

However, to summarize, the car misses and runs slightly rough at idle, the miss is most noticeable under high load as it's a violent random jerk, the tach often jumps as I'm losing coil charge apparently. I narrowed it down to MS (at least I think so) because if I pull the spout pill and set the timing to 20 degrees full-time the car runs perfect under load and runs much smoother all around. When I give spark control back to MS it goes back to the same thing, although the car is still quite driveable, it just runs rougher at idle and has the annoying miss. It also sometimes detonates when it's missing.

I've checked all connectors, I've played with all the wiring, and non of that duplicates the problem.

I do however have a pretty severe voltage fluctuation as read by Megasquirt, I had the logs but deleted them without thinking. I will get another log of the battery voltage if necessary, but essentially it's a rapid oscillation from 10-14 volts typically, and sometimes dipping as low as 8 and going as high as 16, so I'm thinking bad alternator. I will soon have the alt swapped so I can rule that out also, it was in the plans for the car anyway.

Other than battery voltage, my logs have looked perfect with only a very occasional tach spike which I cannot relate to the missing at all, it happens much less than the miss does. AFR even seems to be just fine as read by an LC-1 wideband.

I'm probably forgetting something and if so I apologize, thanks for any help you can give,
John Lewis

[jsmcortina]

You may have more tach spikes than you think as the datalog will only occaisionally catch them. Work on fixing those spikes and you may cure the misfire.

James

[John L.]

Okay, I have some posts on TFI tach spikes and some possible solutions, cchristopher was having a problem that sounds related. I'll look into that and all the other possible causes and report back, meanwhile I'm taking any and all ideas. The voltage fluctuations on the power wire really concern me, as soon as I find the crank pulley needed I will be converting to a better alternator.

[John L.]

I have a log during which the problem occurred, I set the Timer Interval to 5 ms to grab more samples. No tach spikes to be found, which like you said they could still be missed by the log, but that's why I set the interval so low, in hopes of catching them, but yet there are none at all. There should at least be some. I have Batt Voltage on this log also so you can put that on your view in MLV. Thanks for taking a look,
John

[John L.]

I posted my progress in my Turboford thread, excuse me for not saying more but I'm getting exhausted with this and it's tough posting between two forums plus talking to people one on one:
http://www.turboford.net/cgi-bin/ultima ... 864#000025

Still no solution...

[jsmcortina]

I'm using the VB921 (driven from LED17) to drive the TFI module, along with a pullup resistor.

Could you explain exactly how you have this wired as this isn't the method described on the TFI setup page of the MS1/Extra site.

James

[John L.]

http://www.msefi.com/viewtopic.php?t=16 ... ight=led17 It's done as suggested in this thread, although it doesn't have to be that way and I worry that may be causing problems. I might go ahead and switch it to run directly off of LED14/17. I guess you and Scott thought that would be okay(with the VB921) but saw no reason to do it that way, so maybe there are issues this way.

Essentially, the top of R26 triggers the VB921 via a 330R(totally unnecessary I think). The VB921 drives the SPOUT.

[John L.]

I switched the wiring so the IGBT is no longer used, the SPOUT is driven directly off D14, with a 330R pullup on IGN where the jumper wire runs to it. It starts and runs just the same, but still has the same problem of a random jerk under load. I also added a fuel filter on the MAP line as a damper and that made no difference. Datalogs still look good, but the tach spikes are sometimes showing up again, but I don't know what's causing them. I removed the noise filter on the power wire and that made no difference really, but I still have power spikes as shown by Megasquirt, especially under load. I guess I'll have to rewire the switched 12V in my car entirely to see if that's the problem. I replaced the under hood relay module (IRCM) because I think it had a bad relay, and I've checked all the wiring, and even added an additional relay to supply my 12V switched circuit in the car, but still have the jerk. I've even added two more grounds directly from MS to the chassis, but nothing has really made much difference.

[cossie_storm]

Hi
i have some of the same problem.

I'm thinking of trying out EDIS, so maybye that will solve it..

[John L.]

It would be great if you could describe your problem also, as no one else has ever had this type of problem that I can tell. At the moment I can't really blame MS.

I want to switch to wheel decoder , but not until I figure out this problem.

[John L.]

I decreased the resistance value on the Tach Input to 1K by putting a jumper in place of R12 (was the 390 ohm) and leaving a 1K in D2 (or D1, can't remember). It seems to run much better in neutral but I haven't run it under load yet, will find out in the morning. My theory is my PIP wire has a little bit of resistance from age and such and was causing the tach spikes. Maybe the tach spikes were the problem all along.

[FoundSoul]

Just wanted to touch base with you and see if the latest change made any difference under load?

Also-- what do you have Dwell set to?

[John L.]

Reducing the input resistance may have helped some, it's not as often that it happens now, but it still happens sometimes.

Dwell is set to 50% duty cycle.

Like I said, it could still be my car, but that's the most difficult thing to troubleshoot, as Turbocoupes are amassed with hundreds of wires to its many accessories and blessed (*sarcasm*) with Ford late 80s wiring.

[FoundSoul]

How much dwell are you getting at idle and higher load conditions when it misses? A datalog may help there-- I usually use static dwell rather than the calculated dwell based off of duty cycle. Usually something in the area of 6ms cranking and 4ms running.

[John L.]

Not sure how to datalog that, is there a variable for the dwell that's being commanded or is this something that requires an oscilloscope? I need an oscilloscope either way, but just haven't seen one available yet that would work for me.

[cchristopher]

John,

Watch it with a timing light. I noticed that when using MS, and watching idle with a timing light, it bounces quite a bit. With the stock computer in it is rock solid. My wiring is all solid and my ignition components are new.

When idling my car will be fine for about 30 seconds (after first start) then it will start to misfire for no reason. When it misfires the rpm will drop and the afr will fluctuate. My battery voltage is also pretty solid. I will see if I can get a datalog of it tonight.

[John L.]

Mine's not that bad at all, I don't notice a misfire at idle now, and I hardly ever have the solitary jerk under boost. My timing does scatter a lot though, even though my distributor is in very good shape with very little play at all in the shaft.

[FoundSoul]

It is a variable that's being commanded and should be in your datalog (ALT-L in MegaTune to log).

[John L.]

I couldn't find that variable in my datalogs, and changing to static dwell doesn't seem to make any difference, it just changes from day to day, one day it might be fine and the next it might jerk once when I hit boost.

[FoundSoul]

You could be having another issue then, possibly the spark blowing out under boost sometimes--- or maybe intermittent knock due to other variables, IATs for example.

If it's the spark blowing out sometimes running a slightly narrower plug gap can help. Preignition can be caused by running plugs intended for an NA vehicle (hot plugs) in a turbo car, you generally need to go a step or two colder for a turbo car if this is a conversion. As for knock due to IATs does it seem to occur more on hot days than cold? You should be able to see this in your datalogs too...