I posted this in my local forum, but wanted to share for the FI people too as FI people seem to use the AEM ☺.

I noticed a misfire and diagnosed it to be the AEM ECU as the problem.
Being curious, and not afraid to take things apart, I did just that to find the root cause of this problem inside the ecu.

car runs fine now that I did the following:


AEM 1052 (non-u) the U will vary a little in the board layout, but the process is the same


when you take out the circuit boards it looks like this


after separating the boards, you see this


closer up of the connector board


the main board


close up of the mosfets


I have removed the jumper here to show the Q5 mosfet, where you can read the part number 12N06lE


you can see the mosfets labeled Q5, Q3, Q2, Q1, Q4


on the back side of the connector board, you see the legs of the mosfets


take your multimeter on the diode setting, put the positive on the gate and negative on the source
and charge the mosfet with the voltage provided by the meter
this is q4 (the bad one) the reading here should be open between the gate and source after a few seconds
as you see here, its not


then put the meter on the drain and source, this should discharge after a few seconds
as you see here, its not


compare the above (bad q4) with the below (good q1), same measurement points




since Q5 is not used as a coil driver, and is instead (on this ecu for the s2000) used for the coolant temp
display in the dash, (which doesnt work correctly from AEM) you can replace Q4 (the bad mosfet) with Q5 (coolant mosfet)
then add a new Q5 with RFD12N06RLESM since the RFD12N06RLE is now obsolete. I have ordered this part, but not received it yet, and will update this post when i get it to confirm results.
The new mosfet is the same as the origional, but in a different package. This just means the leads coming off the componet still align with the original holes, but need to be bent straight.


Update, May 18th
I finally got around to ordering the replacements. I got a pack of 4 for 6.95. So now I have even more spares IF this happens again.


With the replacement... can you tell which one ☺?



I delayed on getting the RFD12N06RLESM as i was going to look for a more suitable mosfet to drive the coolant gauge properly.... kinda like the modifry, but without the modifry unit), but alas I was lazy and just did as I described above. Maybe i will look for a coolant mofset replacement one day.

 

That's awesome work. I'd be hesitant to take the AEM apart though, my skill is lacking in the circuit board area. I do have a misfire though but I believe I had the misfire prior to getting the EMS.

 

Wow, great job!

I used to notice what I believed to be a intermintent miss fire with my box but I could never figure out, what were your symptoms?

 

x2!!!

 

I believe i was having this same problem with my vers. 1

 

Old thread, I know... Wish the pictures would show as I've had an intermittent misfire show up and am starting to wonder about the ECU. I have a Haltech 1500 here in the shop I may try to see if the ECU is the case.

 

AEM V2?

 

Yes.

 

So I assume you've tried different coil packs and the usual misfire fixes. I'm eager to know what would happen if you swap that Haltech in since I have a weird "issue" with my V2 (or at least that's what I think). It will not fire an AP2 coil (new OEM 2004/5 AP2/K20, used or Denso 673-2301 from Amazon/Rockauto) in cylinder 4. It will only fire AP1 coils in that cylinder and fire them perfectly. It will however, fire either the AP1, AP2/K20 or Denso 673-2301 coils in cylinders 1 to 3...WHAT??? This phenomenon happened after I inadvertently had fake NGK spark plugs in there (brand new from Amazon but I had no clue they were fake) and the car cutout by 7 grand WOT. It would cutout then run on 3 cylinders. Limped home changed spark plugs back to original NGK Iridiums and it never happened again with the caveat being cylinder 4 now only works with AP1 coils...

I don't have a spare standalone like you do to test on my end though Please try it and let me know how it goes.

 

Yes, new coil packs, new plugs, and got rid of the PnP injector harnesses. My original thought on this problem was injectors as the car sits a lot, and I am still thinking it's the injectors. I turned off O2 feedback and forced a rich condition and it handled the misfire better, though still there at times.

The coils came in today and I wasn't upset about replacing them as they are 19 years old (39k miles). I will keep the originals as spares. The plugs were great and the injector clips did not fix it.





I have 2 new ID1050's here and have a buddy sending me 2 more. I'll swap the set in when I get them on Wednesday and send the older set to another buddy for a proper cleaning.

The coils that I got had the same part number molded into the side and aside from the sticker on top, look exactly the same.



As for the ECU swap, I decided against it. The Haltech would require a bit more work as I can't configure the C16 pin for my WBO2 or configure my charcoal canister circuit for my second fuel pump, in addition to having to redo everything in my D connector. The AEM v2 has been in for 10 years and pretty well configured how I need it.