Resolved Botched socketed EPROM ECU?

[bcjjones]

Just got an eprom ecu from a tuners memeber. Planned on having it socketed to install an Ostrich 2.0, got it in the mail an realized it was already socketed. Which the seller I guess didnt realize.

I am currently using TunerPro just until I get my IAT and MAP sensors to run speed density. TunerPro is just too much a learning curve for me. I downloaded what I beleive to be the correct stock .xdf and .bin files. I have them both uploaded onto Tuner pro and Ostrich but when I plug the ECU in and start the car it runs like crap. Bogs down real bad, doesnt want to drive without bucking horribly, if I put my foot on the gas it wants to rapidly jump from around 800-1500 rpms. Never wants to go faster. So either I uploaded the files wrong or the ecu is messed up. Heres a picture

 

[stacked1G]

id say that looks unhealthy...but thats just my .02

id say that looks unhealthy...but thats just my .02

 

[rts91tsi]

It looks like whoever did the socket never removed the conformal coating before soldering, or a very good job cleaning the board afterwards. Clean it up with some rubbing alcohol, and post a pic afterwards.

A few of the solder joints look very questionable as well. The solder joint should cover the pad 360*, have a shinny appearance (Dull indicates a cold joint with poor conductivity), and have a slightly concave shape. Some of the joints appear to be cold or contaminated, and some with an inadequate amount of solder.

If you have a DMM, trace the connections, and continuity test them to ensure they ring out and have a good connection.

As far as uploading/using Tunerpro, just make sure you're using a 32k bin with the 32k XDF and you should be fine.

 

[4g63t]

You can always email tom at ecmlink and send this way to get it repaired. That’s what I have done in the past and will be doing with another one I got a hold of that does not run my car and looks like the above pic.

 

[CrawlinZX14]

The ecu doesn't look great, but the problem your describing sounds like a bad MAF. Is everything wired up correctly?

 

[bcjjones]

Ill try cleaning it and post another pic. I thought there wasnt enough solder by looking at it. Doesnt go all around the point, which I was assuming might effect its contacts. Ill see what cleaning it does. Also I am using a 32k xdf and bin but maybe something isnt correct with the bin? MAF scalar or something?

How quick is the turn around when sending it off to ecmlink? Also what was the charge?

I also thought it had something to do with the maf like you said but when I put my non eprom back un, the car runs fine. Unless like I said above there is something wrong with the MAF settings in the bin im using?

 

[rts91tsi]

Ill try cleaning it and post another pic. I thought there wasnt enough solder by looking at it. Doesnt go all around the point, which I was assuming might effect its contacts. Ill see what cleaning it does. Also I am using a 32k xdf and bin but maybe something isnt correct with the bin? MAF scalar or something?

How quick is the turn around when sending it off to ecmlink? Also what was the charge?

I also thought it had something to do with the maf like you said but when I put my non eprom back un, the car runs fine. Unless like I said above there is something wrong with the MAF settings in the bin im using?

Cleaning the PCB around the socket and re-posting a pic was only suggested so that the members here can further assist you in the troubleshooting/damage assessment of your ECU.

Regarding the inadequate amount of solder joint statement, there is no might about it, and you will not have a proper connection with the pads of the PCB and the pins of the socket.

Your issue is more than likely your ECU's socket, and not the bin and XDF for Tunerpro.

Hopefully you didn't pay more than $135 for the Eprom ECU, when you could have easily just paid ECMTuning convert your non-Eprom ECU to Eprom.

 

[snowborder714]

How quick is the turn around when sending it off to ecmlink? Also what was the charge?

Send an e-mail to [email protected] for more info. They probably won't be able to give you an exact charge as it will depend on if the traces need fixed or not. Extra work = more money.

 

[dexterholland04]

Definitely clean the flux/mess from the bottom there. If you don't have the right chemicals, a little soap and some water with a somewhat stiff brush goes a long way. Just be sure that you dry the board really well before installing it again. Either sit it in front of a fan, or use compressed air (from a large air compressor...not one of those stupid cans of air) to dry it off. From the looks of your picture (its hard to tell with your camera flash), that there might be some discontinuity in a few of your traces on the board. If that is the case, then send it to ECM Tuning to get it repaired...I've repaired broken traces on boards before and if you have no clue on how to do it, then send it to someone who can.

 

[shanebro]

Just take it to a techie he will fix you up. If you do it on your own you might end up totalling it.

 

[daren_p]

Just take it to a techie he will fix you up. If you do it on your own you might end up totalling it.

Alot of the so called "professionals" out there are also hacks. I'll second (or third or forth) send it to Tom/Dave at ECMTuning. They know these ecu's inside & out. They will fix up any issues no problem & verify proper ecu operation. Also have them replace the caps while they have it there as it is common for 2g ecu caps to be leaking now. There prices are very resonable for what they do as well.

 

[4g63t]

Ill try cleaning it and post another pic. I thought there wasnt enough solder by looking at it. Doesnt go all around the point, which I was assuming might effect its contacts. Ill see what cleaning it does. Also I am using a 32k xdf and bin but maybe something isnt correct with the bin? MAF scalar or something?

How quick is the turn around when sending it off to ecmlink? Also what was the charge?

I also thought it had something to do with the maf like you said but when I put my non eprom back un, the car runs fine. Unless like I said above there is something wrong with the MAF settings in the bin im using?

As far as turnaround time think the first ecu i sent them took two weeks from me dropping it off at the PO to having it back at my door.
You pay the return shipping and put in the box a copy of your order# and ship it off. Super simple.

ECMTuning, Inc.

 

[bcjjones]

I went ahead and cleaned it up with some rubbing alcohol. It seems to clean up really well, and everything seems to be fine. I'm not ecu expert but none of the traces look to be damaged, like I said though im no professional. There was just about no solder on the contacts so I went ahead and soldered a little, no professional at this either but have some experience. First picture is the socket and the other is where im assuming a trace was repaired? I cleaned that rust looking stuff up to and figured that was just the protective cover doing its job.

I also did a log with the eprom/ostrich and with my stocker and both picked up a maf reading. The car runs fine with the stocker. Still the same symptoms with the eprom/ostrich. But the maf reading with the eprom/ostrich was very different even at idle then with my stocker, so what do yall think that maybe the .bin is incorrect in translating the maf reading? Thats just me wishful thinking. I have the stocker for parts if anyone has suggestions on how to use that

Well its something with the ecu, I tested it with another socketes eprom and the car ran fine with the ostrich and same bin. Im gunna try to get a refund. Wish me luck

 

[allthrottle]

Board traces can be pulled up at the eprom/socket side too, and can be difficult to notice without pulling the socket out or checking continuity. If you continue having problems with that ecu, I'd either send it in to the guys at ECMlink for repair, or cut your losses with it, buy another pcm and have it sent to ECMlink for a socket/cap service and checkover.

The wire "fix" in the 2nd picture, is it touching any other pin? In the past, I've had issues with pulled up traces on the socket side, copper wire trace repairs touching other pins, and insufficient socket soldering. I was able to diagnose most of them just by checking continuity between the pin and it's destination.

 

[bcjjones]

After reading the above post I went to look at the socket side of the ecu and noticed whats in the picture. The socket isnt pushed all the way down on the ecu. Some of the contacts are, but most are not. I looked at the other socketed eprom and all the contacts are pushed down snug on the ecu. Could that possibly be the problem?

 

[rts91tsi]

I see cold and frosty joints on the pins of the socket, which will have poor conductivity, which I said earlier.

The jumper wire for the repair trace looks like a piece of #24 AWG phone wire with red insulation, which is not an uncommon repair. However, that is the first time I've ever seen the use of a polyester film capacitor with the electrolytic capacitor.

Considering someone added the film cap, it makes me wonder if it was added as a filter because EMI issues, or for added power.

Regardless, with the shape that ECU appears to be in, you should really send it to ECMTuning. As said before, they will be able to test it for proper function, as well as give it a proper repair.

 

[steve]

One interesting point about these connections is that it either works or it doesn't. If the ECU isn't crashing then it's not the EPROM socket. The ECU reads from the EPROM not only for tuning data but for the actual software it's running so it's just as likely re misread an instruction as some tuning parameter.

The location of the film cap makes me guess that C14 is missing from the top of the board.

The jumper implies a trace between those two points is open. That's usually due to leakage damage and also likely that more traces have failed since they fixed that one since it's clear they didn't remove the parts from the damaged area and fix the issue on the top side, just patched it up on the bottom.

 

[rts91tsi]

IDK steve, it looks to me like C14 is there from the side profile shot of the socket. Any insight as to why the film cap was added?

OP, another thing you could try is to clean up the orange/yellow corrosion around the connections for the header pins and repair trace, as there could be some bridging occurring. I doubt it will fix it, but it can't hurt to give it a shot.

 

[steve]

IDK steve, it looks to me like C14 is there from the side profile shot of the socket. Any insight as to why the film cap was added?

If you can see C14 in any of those pictures you have better eyes then I do.
Remember it's a SMD capacitor across the 5v rail and ground, next to D118, between the two filter capacitors C108 and C106.

I usually see a film caps added where that one is because the repairer doesn't have any SMD caps to replace C14 with or the board is so badly damaged they can't replace it.

 

[rts91tsi]

If you can see C14 in any of those pictures you have better eyes then I do.
Remember it's a SMD capacitor across the 5v rail and ground, next to D118, between the two filter capacitors C108 and C106.

I usually see a file caps added where that one is because the repairer doesn't have any SMD caps to replace C14 with or the board is so badly damaged they can't replace it.

It's more than likely that you're correct, as always, considering the years of experience with these ECUs and their repairs.

I didn't have a ECU in front of me to reference at the time, but after looking at a board, and comparing it, I see what you're saying. It would be almost impossible to tell if C14 was actually there, considering the amount of shadow being cast by C108.

Thank you for the explanation!

*EDIT* Wouldn't having the film capacitor soldered on to C108 affect the overall charge for that branch of the circuit, since it is actually in parallel like that, instead of in series as it was designed?

Since charge=Q, C=capacitance, and V=voltage, Q=CV...

C total would be depend on these formulas:

Parallel: C=C1+C2
Series: C= 1/(1/C1+1/C2)

 

[steve]

*EDIT* Wouldn't having the film capacitor soldered on to C108 affect the overall charge for that branch of the circuit, since it is actually in parallel like that, instead of in series as it was designed?

Look again, in both cases C14 is in parallel with C108.

Also keep in mind that C14 is ~0.01uf. C108 is ~100uf. C14 is there for RF decoupling on the main 5v bus.

 

[rts91tsi]

Ah, I see it now!

The parts board I grabbed for a quick reference was badly tarnished and I didn't see the other node.

You rock steve!

 

[steve]

After rereading this it may just be your bin file that bad. Did the ECU come with an original EPROM chip?

If I sold something and the person that bought it started working on it, it's not returnable. The time to return things is before you try fixing them on your own. Once you do you own it.