• Hello Guest, welcome to the initial stages of our new platform!
    You can find some additional information about where we are in the process of migrating the board and setting up our new software here

    Thank you for being a part of our community!

Megasquirt Erratic Spark

TR Conn

Member
Joined
Dec 13, 2006
Location
Long Beach, Ca
Erratic spark output
Megasquirt 2 v3 on a B20.
Something is fried in the ECU. Engine won?t run no matter what crank input. I?ve tried distributor -, points trigger, 36-2 with VR sensor, and with a Hall sensor. In all cases the tooth logger and composite logger have normal looking patterns.
The problem is the output to the coil, it is very erratic.
When attached to the stim (low end 5 pots) I get a stable tach reading, but the LED for ignition fires 2 to 4 times per second at 800 rpm. I raise the rpms to 5000 and the ignition output is just about the same. The LEDs for the injectors operate as they should.
I?ve checked over the board several times for soldering, correct parts in the proper places, jumpers where they should be.
My only guess is that something is fried; hopefully it isn?t the daughter card.

Looking back, it seems that my problems began around the time I started using the BIP373 for the spark. It worked well at first, but began giving synch errors. I figured the Bosch 124 would be better, but it had the same problems.

Any help or insight would be greatly appreciated.
Thanks, TR

My setup:
B20
MS2v3
Tunerstudio V3 1.05
MS Extra 3.4.2
36-2 w Hall sensor (both from DIYAutotune)
Bosch 124 module
Locked distributor
 
Need pics of your main board (top & bottom).

How is the MS grounded? Which pins grounded where?

Does it pass the grounding test at the end of section 3.2 of the Extra Hardware manual?

Where is your coil mounted and what type?
 
1) In Tunerstudio, with the stim, try the Testmodes - Inj/Spk - Coil Test, and see if the LED pulses correctly. If this is OK, check that you don't have any Ignition - Noise Filtering enabled.

2) I think you can pick an alternate pin for spark (WLED?) - check the manual or Tunerstudio to make sure, and to find the re-wiring instructions. I'd use an external ignitor/powerstage to reduce the electrical noise within the MS box.
 
Erratic Spark

Thanks for the replies,
All of the grounds terminate on the intake manifold.
The wiring is as the diagram indicates, though I will check later today.

It appears to me that the problem is internal to the ECU since I have the same problem while using the stim.

https://imgur.com/VS6Srwl


The coil is a stock Volvo/Bosch unit. I’ve tried it both with and without a ballast resistor.
It is mounted on the firewall in the stock position (for a PV445)

Using test mode: LED can be turned on and off but will not blink as it should do.

I’m utilizing the Bosch 124 ignition module, I disabled the BIP373 internal coil control because I thought excess noise was the problem.

I’ve also replaced U7 with no change.
Would re-flashing the daughter card have any effect?

lZ171cm.jpg


g0jk4bD.jpg


Again, thanks for your expertise
TR
 
Last edited:
Wash your board! isopropyl alcohol and an old tooth brush.

The solder flux will absorb atmospheric moisture and become conductive.

Soldering on that board is questionable, possibly a poor choice of solder?
Solder joints should have a shinny appearance and flow.
Crystallized blobs are low heat or contamination or crappy solder.
Avoid lead free crap.

Kester 60/40 solder
https://www.amazon.com/KESTER-SOLDER-32117-24-6040-0027-Diameter/dp/B00068IJPO/ref=sr_1_2

I did not see an answer to the question regarding the test in the last paragraph of 3.2 of the hardware manual, this is critical.

Which wide band?
 
Last edited:
Is the spark consistently erratic?
I had a microsquirt that fired the ignition 4x per cycle. Swapped in a known good one, loaded the same tune, and the issue was gone.
The OG microsquirt was basically not repairable... and that sucked.
 
Thanks for the follow up.
I have scrubbed the board and scraped between close soldered pins. The stuff is very difficult to get off.
The solder is 60/40 but quite old. I'll order new stuff and re-solder what ever looks sketchy.
And yes, it passed the ground test.
Keep pressing on...
TR
 
It appears to me that the problem is internal to the ECU since I have the same problem while using the stim.
...
Using test mode: LED can be turned on and off but will not blink as it should do.

Yeah, that seems like the ECU isn't working correctly. From your picture, you're using JS10, which is wired directly from the CPU chip on the daughter board. If the ign wire was going to an external power stage (e.g. 124) but was shorted out to +12v, or connected directly to the coil "-" tab, it may have damaged the CPU pin.

Fortunately, you can use the D14 alternate pin for a single coil ignition. Change the setting in Tuner Studio, and follow the D14 / SparkA PCB wiring instructions.

One more oddball to check first - under the main Ignition Options, make sure you don't accidentally have Dual Wheel enabled. The Dual Wheel mode uses JS10 for the cam input, and I don't know what MS will do if it is also configured as the IGN output.

You could also try reloading the MSextra 3.4.2 firmware, but I doubt it will change anything.
------------
For soldering cleanup, you can buy a can of PCB Flux Remover that work's better than IPA.

For solder itself, 60/40 leaded solder is so much nicer to use than lead-free. AFAIK, it never goes bad -- the spool on my bench is at least 40 years old and works fine.

If you don't have one already, a temperature controlled soldering station makes it much easier to do nice work. Set it for ~700deg. 40 to 50 Watts is plenty for PCB work, but a little wimpy if you want to use a fat tip and join 10gauge wires together. Try jameco.com for cheap "soldering station"s.
 
I have gone full sequential with coils with integrated drivers so I have no experience with conventional ignition coil control using the BIP 373 or external amplifiers. I will let others advise on component failure issues with that particular configuration. However, your circuit board solder condition may be 'the' issue.

You said that
It worked well at first, but began giving synch errors.
It is hard to tell from your photos whether the stuff on the solder connections is residual flux that was never cleaned off or whether it is bloom that has formed on the solder joints over time. How long was it from working well at first to the problem appearing. If it was a significant period of time that stuff may be corrosion bloom. Flux by itself is typically not a problem (other than hiding bad solder joints) because rosin flux is normally non conducting. Bloom is different and can be caused by exposure to moisture. Lead / tin solder can form a bloom; but, it typically seems to be constrained to the surface of the solder. However, if you failed to clean off the residual flux from the initial solder work some 'informed sources' indicate that with exposure to high humidity the fluxes can form corrosive compounds which will attack the solder joint. This can be corrosion pitting which may result is solder joint fracture. Depending on which informed source that you consult the bloom may be partially conducting. If the stuff on your board is corrosion bloom it looks like it is bridging some solder pads which may be contributing to erratic operation.

If the stuff on your board is just residual flux, I would recommend using a spray on flux remover and a tooth brush. IPA will work; but, IPA will also soften and start to remove any conformal coating on the board if it was applied. Depending on the type of conformal coating that may also occur with the flux remover. If that is corrosion bloom, it is hard to find any guidance from industrial sources because that is typically not a problem they are interested in because it does not appear to be a problem they experience with current production methods. The internet wisdom seems to be to use a mix of baking soda and distilled water for clean up. Baking soda is caustic and normally not the hot ticket for dealing with corrosion; but, it is also slightly abrasive which may be what provides the cleaning action. You are kind of on your own with this one.

With the solder joints cleaned you can inspect them for damage / fractures. If you decide to reflow the solder joints you should remove any conformal coating that was applied to the board before attempting to reflow the joints. I know that with acrylic conformal coatings, if you leave the coating in place while trying to re solder the coating melts and fouls the soldering iron tip and you will be perpetually cleaning the tip and your solder joints will look like merde. After the repairs, clean the repair areas with flux remover to prevent future problems.

Once you have fixed the board, for the long term you should apply a conformal coating to the solder side of the board to prevent solder corrosion. However, don't do this until you have completed and confirmed your repairs. A new layer of conformal coating will make any future repairs or modifications difficult because it has to be removed in order to make a good solder joint.
 
Last edited:
WOW! I know so little.
Thanks for the information.
I've ordered new flux and flux cleaner.
I have cleaned the board with IPA, but it didn't seem to do much.
I have a good soldering station, but have been using it at too low of a temp.
I'll give the baking soda treatment a try then reflow all of the joints I can get to.
I have a bad habit: if something doesn't work, I try something different AKA a workaround.
This time I will get to the heart of the problem and fix it.
I'll keep you posted...
TR
 
The set temperature of the soldering iron will in part depend on the wattage of the heating element. If you have a lower wattage heating element you may have to set the temperature a bit hotter because as soon as you touch the tip of the iron to whatever part you are soldering the temperature starts to drop a bit. A larger large heater element will get the temperature back up there quickly. A smaller element may need a higher starting point to deal with the temperature sag. It also depends on where the sensing element is located in the iron which may result in a small or large difference between the actual tip and the measured temperature.

I constantly clean the tip on a damp sponge and re tin the tip with solder so that I get good heat transfer to the part I am soldering. You want to be able to heat the part up quickly and get solder flow and then get out of there. Particularly with semi conductors, you don't want to be doing a slow cook on the part as you wait for the solder to start flowing.

I am assuming when Bobxyz suggests 700, that is 700 F. That is around 370 C and I typically use 350 C to 370 C on my 60W station when soldering with 63/37 solder. I use 22 gauge solder for circuit board work. The finer gauge reduces the amount of solder that has to be heated and reduces the risk of forming giant solder blobs of bridging pads like on the small TO92 transistor pads. If you use a lead free solder you will have to crank up the soldering iron temperature.
 
Last edited:
I've been working on this, so far no change..

Baking soda scrub - really cleaned things up.
Re-flowed/re-soldered every connection on the board. With new solder.
Posted my problem on the MS support forum - no answers yet.
Purchased flux remover - will use it in the morning.
Next step will be to remove and test various components.
 
As a simple test, you could check that the JS10 daughterboard pin is connected to IGN out -- measure the resistance from daughterboard pin #17 to the IGN pin on the DB-37 connector, pin #36. There's a chance that the trace on the main PCB underneath the CPU board is damaged. If so, just run your yellow wire to the end of pin 17 instead of the JS10 hole.
 
I just saw your .msq on the ms-extra forums. You're set to "fuel only" mode, i.e. no spark. You need to go back to "toothed wheel" mode, or whatever you're trying to run, and turn off the cam input (currently set to JS10).
 
I have good continuity between the daughter card #17 and DB37 #36.
Cam input is greyed out, so it is unavailable. The info note in the cam input says it is only for single cyl or 2 cyl odd fire.
Thanks for the suggestions.
TR
 
Set the MS to toothed wheel mode, no cam, and try the ignition output test mode, or set it to basic trigger mode and try the rpm pot.
 
Hope you're doing well now, with no lingering issues.

I had my MS setup on the bench while looking at something else and tried your .msq file to see what would happen.

You've run across an area of "unspecified behavior" in MS. [I'd call it a bug.] When configured for fuel-only mode (no spark), the JS10 IGN pin will pulse erratically as rpm is changed. Maybe .......blink.....................................blink..blink.................................................................................................................................................................................................................................................................................blink..........................................................................................................................................blink.................. etc.

If you temporarily change to "basic trigger" (used with distributor points), the JS10 IGN pin works correctly. Note: with the stim, the LED turns off when the coil is charging. With a 3ms dwell time at 1500rpm, the LED is off for 3ms and on for 17ms, which dims the LED just a little bit.

You can go back to your previous setup with Toothed Wheel / Single Wheel With Missing Tooth (Cam Input should be grayed out) and try the Output Test Modes - Coil Testing. With the stim, change the test mode dwell time to 25ms to make it easier to see the LED blink. Change the dwell back to 3ms before leaving the Test Modes menu just so that you don't fry a coil if you later want to use the testmode in the car (with the coil center cable fixed ~1/8" from a bolt head).

Give it a try and see if the MS appears to be working OK with the stim.
 
I think I have it working again.
The problem is, I don't know why.
It checks out on the stim, I'll try to get it reinstalled today and see what happens.
I'm not quite sure when this all started, but I do recall that I upgraded to the newest release in the not too distant past. I plugged it into my tablet with the Android version and that is when it started working.
Therefore I think you are correct in calling it a "glitch"
Keep your fingers crossed while I do a bit of wrenching and I'll be back with my results.

Thanks again for your help.
TR
 
I've got it up and running. It isn't optimum, but it works.
I switched around to the opto isolator circuit and am running fuel only with the tach signal coming from the coil negative.
Not nearly as sophisticated as the 36-1 and ECU control of the ignition. I'm not really sure that the ignition circuit is in tact.
Thanks for all of your help and suggestions.
TR
 
Back
Top