No start, so spark: ECM faulty

Status
Not open for further replies.
Joined
Apr 28, 2015
Messages
723
Location
Reunion Island
Hi,

I regularly service a friends bike, MY2013 Suzuki Inazuma GW250 (L3, P21).

A few days ago, after 102000 km of loyal service, it refused to start. The dealer mode feature gave C24 and C25 codes, probably meaning that ECM can't sense ignition coil spark feedback. Long story short, after unplugging ECM, spraying WD40 on the pins (but they were clean) and plugging it back, it started immediately and on the morning after it ran well on cold start. For 15 minutes, and then the engine cuts out definitely with same codes. And never sparked again. Ever.

Well, it already had some misfires when starting cold on the morning before the first no-start issue, and we recently replaced spark plugs (worn), spark caps, and HT leads (corrosion inside spark plugs caps). It started better after those replacements, still some light misfires occurred on cold start. All this was before the former no-start issue. The ignition coils were replaced because one went cut at approx 75000km.

I used service manual and electrical wiring diagram and checked every wire related to ignition, continuity (especially CKP wiring and coils primary negative to ECM), and short-to-grounds, CKP resistance value, coils resistances values, +12V presence where (especially ECM, coils primary positive) and when there should be, fuses of course...

Everything tested was in spec. Still, no start, no spark, codes C24 and C25. I swapped a coil and spark plug with known-good units. No spark. The thing I don't have any means on hand to test is if there's some signal on CKP when cranking, and if ECM really tries to trigger coils or no. I assume CKP is good because resistance value is normal, and when I unplug it it throws a code for CKP (C12) trying to start.

I think ECM is bad (internal component dead, I assume it uses transistors, or a solder went bad and when we moved it first time it made contact again).

Do you think I should replace ECM as part of diagnostic or is there a simple procedure to test if there's signal to coil?

And when looking at OEM parts catalog, original reference 32920-48H20 was superseded by 32920-48H21, without further comment, so it comforts me of some possible design flaw on the original *48H20 that equipped the bike on factory.
 
Try tightening the pins up so the fit tighter together first, a mini screwdriver or small awl will do. Don't go hog wild just a little usually does the trick.
 
Wiring diagram.

FA602740-890A-4BC2-AB62-A2A160A41A17.jpeg
 
Originally Posted by Trav
Try tightening the pins up so the fit tighter together first, a mini screwdriver or small awl will do. Don't go hog wild just a little usually does the trick.


Do anyone knows how to gain access to those little AMPSEAL orange connectors to tighten the pins?

Here's a pic of the ECU pins. 18 + 26 pins connectors.

0D996D57-8B56-4805-B6B9-D547F86D088E.jpeg
 
Last edited:
Superflan, all you want to do is to VERY carefully spread the pins outward toward the sides of the housing.just a bit outward, to make the plug a bit harder to push into the socket. No special tool is needed Often, E 28 ECUs can be temporarily revived by tapping on the case to restore a bad connection internally. The typical failure mode for the ECU was the power transistor legs were loose in their soldered joints. A dab of flux and a touch of a hot soldering iron to re-flow. At the demonstration, 2 out of 3 dead ECUs were revived. #3 had the smoke leak out judging from the smell. If the ECU isn't potted ,then this maybe possible. I can't think of a worse application for soldered joints than a bike.
grin2.gif
 
i would verify the signals at the ecm(backprobe connectors).
reseating connectors can mean loose female sockets or cracked solder joints from pins to the board.
this is pretty common.
and agreed that the vibration of a dirtbike would be hard on the solder joints.
if its not potted have a good look at all of them.
 
Thanks for all the inputs.

From judging others Suzuki/Denso ECM, they are potted.

I'll try the trick from Andyd (bending pins slightly to increase contact pressure), if it won't start we'll order new used known good ECM.

I'll post results here
smile.gif
 
Status
Not open for further replies.
Back
Top