OT - Any good, adventurous ECM Shops?

rdrast

Lifetime Supporting Member
rdrast
Join Date
Apr 2003
Location
South Carolina Lowcountry
Posts
5,544
Desperately seeking a facility that could repair a defective Mitsubishi manufactured Engine Control Module.

The OEM model is: 6B5-8591A-10-00, the case is (was) stamped with 6B5-10. I have absolutely no information on it, nor can I locate any.

This was a fully potted board (with what seemed to be black silicone rubber and lots of white gravel), but I have removed it from the case, and removed most of the potting.

The symptom is no spark for two cylinders (4 cylinder FI engine). Statically, I have measured that two SMD MOSFET's are a bit different from the other pair, but haven't figured out yet even if they are the outputs to the spark coils, or the fuel injectors.

I'd even be happy with a partial schematic of the beast at this point.

Thanks
 
Hey,
I am not sure if these guys can do Mitsubishi but it's worth a try. I found them on ebay, their store name is "AUTO COMPUTER EXCHANGE" Toll Free 1-888-664-8787. Hope they can help,
Regards, Jason
 
jawolthuis said:
Hey,
I am not sure if these guys can do Mitsubishi but it's worth a try. I found them on ebay, their store name is "AUTO COMPUTER EXCHANGE" Toll Free 1-888-664-8787. Hope they can help,
Regards, Jason
Click to expand...

Thanks, gave them a call, not encouraging /sigh.

And panic - I'll post one when I get back home, I don't have it here at work with me right now. (been spending nights outside at home picking off the potting compound lol)
 
I assume you've made absolutely sure it's the ECM, and not connectors, etc.
The only static measurement you can do on a MOSFET is the intrinsic diode between Source and Drain. If that measures as a normal silicon diode junction (~.400-.700), the MOSFET is probably good.
The output is most likely sinking. Measure the output leads to see if there is 12 volts coming from the coils. Temporarily ground them to see if you get a spark. Spark will occur when you open the circuit. Try the grounding method first through a low Ohm resistor (10 Ohms or so) just to make sure you're not shorting a 12 volt supply directly to ground.
With the coil wires identified, you should be able to trace them back to the driver device. Visual tracing, or random probing with the Ohmmeter.
 
Last edited:
keithkyll said:
I assume you've made absolutely sure it's the ECM, and not connectors, etc.
The only static measurement you can do on a MOSFET is the intrinsic diode between Source and Drain. If that measures as a normal silicon diode junction (~.400-.700), the MOSFET is probably good.
The output is most likely sinking. Measure the output leads to see if there is 12 volts coming from the coils. Temporarily ground them to see if you get a spark. Spark will occur when you open the circuit. Try the grounding method first through a low Ohm resistor (10 Ohms or so) just to make sure you're not shorting a 12 volt supply directly to ground.
With the coil wires identified, you should be able to trace them back to the driver device. Visual tracing, or random probing with the Ohmmeter.
Click to expand...

Easy isolation, I actually have two engines. Swapped the ECM's, problem swapped. The spark outputs seem to be 0V, directly to 4 individual spark coils (Looks like the +12 is common). 1 & 4 have 0 output, 2 & 3 are fine.

I was going to go the "Follow the harness" route, but my back was starting to hurt from sitting in the engine box.
 
Coils are common +12, 'looking for a ground' to activate. That means with the connector unplugged from the ECM, there will be 12 volts coming from each coil. Identify the 4 (or more) wires with +12 on them. Disconnect coils one at a time, and see which +12 signal just dropped. To do it remotely, use a test light to do the grounding, then unplug each coil until the light goes out. (You won't catch this old man crawling around tracing a harness. I will go to extremes to avoid that!)
If discrete MOSFETs are for the injectors, then a high current IC must be the coil driver. Look for some more goodies that are heatsinked/surface-mount-soldered-cooling.
 
Pictures attached. Still can't determine which connects where.

The large transistors across the top are MN638S darlington power transistors, the smaller two, paired on front and back, are 2SK3377 MOSFETS (datasheet attached).

My shop manual for these engines seems to be "Close, but not Quite" when it comes to the wiring diagrams.

Other information from the external case, is base part number: 6B5-10, then "F8T94872 4920"

ECM_TOP.jpg ECM_BTM.jpg
 

Attachments

  • 2SK3377.pdf
    47.1 KB · Views: 26
Now I'm getting desperate. I'm to the point of trying to work up my own ECM out of a Rabbit uP, but I'm woefully inexperienced on how to control an engine that rev's up to 11,000 RPM. Anyone have real experience in engine control? I'm willing to take this to PM's or emails...
 
You probably already know this but the darlington transistors MN638S are the ones that fire the ingition/spark. Here is the data sheet I found regarding these items. It is a list of electronic components used on autos. There is also a web page but I am having trouble posting it that says how to test these devices. Hope this helps. Actually I am having trouble posting the pdf file so if you want since the file is larger than what is allowed send me an email and I will send it to you that way.
 
The center lead is the Collector. Right lead is Emitter. Check for shorts C-E. If so, find out why you lost 2 devices. Shorted coils or harness.
It may be tough to get MN638S. I found an ST device that should work as a replacement, and be much harder to blow up. STD37N05TZ. Your device is 380V, +/- 50V, 6A. The ST is 370V, 5A, but designed for automotive coil applications. It should work fine. Replace all 4 to keep parameters matched. Can't find it at distributors. You'll need to contact ST for samples. Another option is BU931T. I don't like it because it's in a standard TO220 case, and gain is low. You'd need to use a Dremmel to cut off the top, and shape the leads for surface mount.
 
Last edited:
panic mode said:
i just googled 6B5-8591A-10-00 and found few places that sell it. seam to be popular ECM for small boats.
the price was a shocker (~1000-1500 USD). This one seam to be the cheapest at $915:
http://www.boats.net/parts/detail/yamaha/Y-6B5-8591A-10-00.html
Click to expand...

Lol...Considering my signature line, ya think I didn't go that route? :p I've called just about every dealer and parts supplier in north america, nobody stocks them. I do have one on order through boats.net, but they aren't expecting any in until sometime in July :(
 

Similar Topics

R
OPC based SCADA, need to read data from engine ECM
I have a client asking for an inductive automation IGNITION SCADA system for a new lab in which they will test various engines that use a variety...
Replies
4
Views
2,438
BobB
BobB
Back
Top Bottom