How hot is too hot

[Toehead]

Today was a real scorcher.

Get in my car, try to start. Fuel pump primes as normal, but no start. I assume the RPM signal is lacking. Pull the megasquirt out for the dash and it is HOT, to the point that it is slightly uncomfortable to touch.

Let it cool, the car started as normal and I drove all the way home without a problem.

How hot is too hot, and what components may be thermally sensitive on the tach input circuit?

Thanks guys.

[naviathan]

Sounds like you need to look into getting one of those solar powered fans that you can roll up in the window. It can get extremely hot in a car during summer, but I wouldn't think anything in a MS would be fazed by it.

[Toehead]

I have a thermometer in my car. It was 180-190 F in there.

[naviathan]

Yeah, that's too hot. You need to put some windshield sunblockers in and leave your windows cracked to let air flow through.

[Toehead]

Only problem is the rampant thunderstorms.

Is the megasquirt normally that sensitive?

[Toehead]

I hope not... The stock ECM never complained about the heat!!


I need to resolve this issue This is my every day driver, and if it can't reliably start i can't drive it to work...

[Toehead]

Anybody have a suggestion for a testing procedure?


I'm planning on using the stim and a heat gun.... Maybe a thermocouple too so I don't kill anything. Heat it until symptoms show.


i am still not sure whether or not this is the car or module. I need a weekend when I can do some experiments.


please chime in guys I'd really appreciate some help in getting this issue resolved

[Matt Cramer]

Stim and a heat gun should work. I'd try data logging one of the hot starts first, however.

[naviathan]

I wouldn't have expected the MS-II to be that sensitive, but then again, you are hitting damn near 200* inside the car. Sounds like you have the right idea with the heat gun and the stim though. Keep posting the results, I'm curious about this one.

[UK Turbo]

I fixed a 2" computor fan into my casing and picked a 12v feed from the pcb. Couple of holes in the other end and she runs cool.

[naviathan]

Sounds like a good solution, but the problem shouldn't exist in the first place. Toehead is right, if the stock ECU can handle the heat the MS should be fine as well. Everything is built with 1/4-1/2 watt components and I know from experience they can take quite a beating.

[Toehead]

Ok.

New updates.

I have confirmed that it is the megasquirt that is failing to receive the rpm signal. The module is brand new, and once the car is started works flawlessly.

If the car wont start, I can get it to start every time by cooling the megasquirt with a gel icepak (or a cool can). I have been doing this to start it. it will run all the way home without problem.

HOWEVER!
The megasquirt, even when it will not start the car, will work perfectly on the stim. I cannot replicates symptoms on the stim.

I have the MSnE HEI mod for the the ignition input. I am thinking that the module is not providing a completely clean ground, and when the optoisolator is hot it becomes more sensitive.


Solutions? (transistor mod?)

Last weekend I brought the squirt into work, and used the clean room and completely redid the whole board under a scope. It is spotless.

[Sprig]

...I can get it to start every time by cooling the megasquirt with a gel icepak (or a cool can).

The megasquirt, even when it will not start the car, will work perfectly on the stim. I cannot replicates symptoms on the stim.

This would indicate a possible issue with the car wiring. WHY cooling the MS adds confusion to the issue.

I have the MSnE HEI mod for the the ignition input.

Solutions? (transistor mod?)...

R12 is now 1K ohm? You have an earlier stim or did you modify it to work after the MSnE mod?'

Can you use a can of cool air to isolate a single part?

Ok, so step back and take a fresh look.

First, double triple check the HEI module wiring and ensure the coil polarity is correct.

Second:
Confirm D1, D2, C12, and C30 are unpopulated.
Confirm R13 is 4.7K ohm, and has +5V at one end.
Confirm C11 is correct value.

Third:
For this, and the problems you have been experiencing, I would personally use the shotgun approach. Replace U3, R12, R13, and C11 all at once.

NONE of these parts should be affected by cooling them down slightly. This behaviour just doesn't make sense.

Good luck,

[Sprig]

... cooling the megasquirt with a gel icepak (or a cool can). I have been doing this to start it.

BTW, the only thermally sensitive components on the board are the polyfuses.

You could jumper the 5V polyfuse for testing.

I also overheard that D19 can be affected thermally if it has been comprimised.

Thinking out side the box...

[Toehead]

I know it is not a problem with the wiring. I did the transistor mod... I suppose i'll see how it works tomorrow.


R12 is now 390 Ohm.

With the HEI mod, it was 1 K ohm. I did not modify the stim and it worked to show RPM.

I was thinking that the only thing that could even be sensitive on the input circuit was the optoisolator.


It is weird behavior, but it is repeatable. Cooling the megasquirt (just the megasquirt) will start the car every time.


Could it be the db-37 plug? I tried moving it around and it was still a no go.

Edit- I do have an early stim, without the ignition led.

Today when it wouldnt start the megasquirt was at 185 degrees F!

[Sprig]

...R12 is now 390 Ohm..With the HEI mod, it was 1 K ohm.

I did not modify the stim and it worked to show RPM.
Edit- I do have an early stim, without the ignition led.

This is an IMPORTANT piece of information that is KEY to troubleshooting your problem.

The transistor mod may appear to "fix" your problem. I'm curious your resuls. However, the transistor mod seems to be a "band-aid" for something else amiss.

Orginally it appeared to me that the transistor mod was the way to go, but additional consideration of the whole enchillada, so to speak, we determined that the MSnE mod was the best overall solution.

I was thinking that the only thing that could even be sensitive on the input circuit was the optoisolator.

I remember that electrolytic capacitors are extremely sensitive to thermal conditions, but have forgotten how sensitive the small ceramic capacitors are.

It is weird behavior, but it is repeatable. Cooling the megasquirt (just the megasquirt) will start the car every time.

ReRead my addtional post. D19 may be comprimised and subject to variations in thermal conditions. BOTH polyfuses ARE affected by thermal changes.

Could it be the db-37 plug? I tried moving it around and it was still a no go.

This was my first guess. However, the MS board uses male pins on a female connection. Cooling the MS should in theory decrease the manlyhood of the pins and make female connection less then likely.


IF the transistor mod "fixes" your hot start issues, I'd be curious if you want to look into this further.

I'm curious how much of a change in trigger offset is required between these two modifications.

I'm glad I don't have these problems so far!

[Toehead]

D19 is overvoltage protection for the 5 volt line correct?


If D19 was compromised, it could be leaking voltage and the 5 volts would be less the 5 volts. Correct?

Would I not see this on the TPS calibration? The TPS calibration does not change at temperature.

It seems to be normal. In fact, everything else works perfectly, and the board appears to be completely normal (aside from the RPM problem)

Another thing I should probably mention.....

Before the HEI mod, this was not a problem. I had another problem with hot cut-outs of the ignition (due to module voltage drift). The HEI mod cured that problem, but now left me with this one!

Thank you for all your replies!

I will keep everybody posted.

[Sprig]

D19 is overvoltage protection for the 5 volt line correct?


If D19 was compromised, it could be leaking voltage and the 5 volts would be less the 5 volts. Correct?

Would I not see this on the TPS calibration?

Correct.
Correct.
MAYBE! Might be just slight enough voltage drop that the sensor calibration doesn't drift out of normal tolerance, but still drops enough to keep the optical circuit from functioning.

Thank you for all your replies!

You problem is unique and doesn't really make sense. I honestly don't think there is a high probability that anything I write will be the actual fix. However, do hope that the words keep you thinking and motivated to keep working out a solution despite how frustrating something like this can be.

[Toehead]

So far so good....


The transistor mod appears to be working....

Fingers crossed!

However, I don't have my timing light so I cant check the offset as of yet!

[Sprig]

R12 is now 390 Ohm.
With the HEI mod, it was 1 K ohm.

Something about this thread was bugging me, so I came back and took a look.

Here is the link to the mod page:
http://www.megamanual.com/ms2/inputHEI.htm

I doubt the tranistor mod worked because it is more sensitive to lower voltage signals.

Did R12 get sized by pulling numbers out of a hat? There is quite a bit of diffference between 390 and 1300 ohms in the two circuits. If I did the math right, assuming 1.2V drop across the transistor when full on, the optical coupler gets over 2.5X the current as the original circuit design.

Simply reducing the value of R12 may be the solution.