2G No start on occasion, now most of the time

[voodoogsx]

So I've had an on and off no start (turn the key and the car cranks but doesn't start). I thought it was my fuel pump as I could usually fiddle with the wires for the relay (did relay mod) and it wouldn't have an issue for a while.

I recently had to replace my water pump as it was sounding like it was about to die and now I'm finding I'm running into the no start all of the time.
I started with checking for spark (it's there), fuel pump (it turns on, I can hear it going when I enable it in DSMLink, plus checked when turning the key and tested the original relay) I checked the injector signal wires and get signal there. I checked the signals for the cam and crank, all there.

So here is the one thing though that I noticed and is causing me some worry. https://www.dsmtuners.com/threads/how-to-diagnose-a-no-start.217951/
In the section "Verify ECU operation " it notes, "also watch the stock boost gauge, the needle should move halfway up the gauge." When you turn the key to ACC.

Now, before I even crank the car this tells me if the car will start.
If the needle doesn't move, the car won't start, if it does move to the halfway point, it will start.

In the no start case, the CEL stays on, but I can't seem to connect to DSMLink when thats the case. But when the boost gauge does move, I can connect.
secondly, if the needle moves up, I also hear a whirring/bzzzrrrrr sound for about a second or so, not sure what it is, it's always been there it sounds like something is being set, like something moving almost like a loud servo but more metallic.

I've looked to see what this gauge movement is suppose to mean, but all I get is everything in the world about how not to trust the boost gauge.

 

[Hickory-NC-2Ggsx]

Maybe remove the ecu and open it up to see if anything looks unusual or burnt.

 

[Justin DuBois]

Have you inspected the ECU main board?

 

[steve]

So here is the one thing though that I noticed and is causing me some worry. https://www.dsmtuners.com/threads/how-to-diagnose-a-no-start.217951/
In the section "Verify ECU operation " it notes, "also watch the stock boost gauge, the needle should move halfway up the gauge." When you turn the key to ACC.

Now, before I even crank the car this tells me if the car will start.
If the needle doesn't move, the car won't start, if it does move to the halfway point, it will start.

In the no start case, the CEL stays on, but I can't seem to connect to DSMLink when thats the case. But when the boost gauge does move, I can connect.
secondly, if the needle moves up, I also hear a whirring/bzzzrrrrr sound for about a second or so, not sure what it is, it's always been there it sounds like something is being set, like something moving almost like a loud servo but more metallic.

I've looked to see what this gauge movement is suppose to mean, but all I get is everything in the world about how not to trust the boost gauge.

First of all it's the RUN position of the ignition switch that is being asked for here. The position after ACC but before START. What this is checking is if the ECU powers up correctly or not. When the ECU powers up it runs some tests and setup code which for example, turns on the CEL for a bulb test, initialize the ISC/IAC position (Maybe the whirring/bzzzrrrrr you hear), reads the various sensors for Mass Airflow and outputs the correct PWM signal to make the boost gauge read Atmospheric pressure, and turns off the CEL is there aren't any outstanding fault codes.

If the ECU can't run correctly that long it won't finish these tasks. In some cases it cant run long enough to turn the CEL on, in others it doesn't run long enough to turn the CEL off.

As mentioned the ECU drives the factory boost gauge with a PWM signal based on it's idea of what the boost gauge should read given the RPM, temp, and airflow. It's a simulation which is why you shouldn't take it for gospel normally. It this case it's pretty accurate about if the ECU is working.

And in this case it sounds like it time to have it checked.

 

[bizie]

If you're getting fuel and have spark then you need to confirm compression and timing. If the ecu were malfunctioning the problem likely would not create spark. Usually no starts are very basic problems. My initial guess would be to check all the wiring for corrosion or bad connections. Areas like grounds and supplies. Check around the fuse/relay area under the hood.

Whats the battery reading at start and if it runs? You're sure you have 12.6 volts then start following wires.

 

[voodoogsx]

Maybe remove the ecu and open it up to see if anything looks unusual or burnt.

I'll pull the ECU today and check it out.

First of all it's the RUN position of the ignition switch that is being asked for here. The position after ACC but before START. What this is checking is if the ECU powers up correctly or not. When the ECU powers up it runs some tests and setup code which for example, turns on the CEL for a bulb test, initialize the ISC/IAC position (Maybe the whirring/bzzzrrrrr you hear), reads the various sensors for Mass Airflow and outputs the correct PWM signal to make the boost gauge read Atmospheric pressure, and turns off the CEL is there aren't any outstanding fault codes.

If the ECU can't run correctly that long it won't finish these tasks. In some cases it cant run long enough to turn the CEL on, in others it doesn't run long enough to turn the CEL off.

As mentioned the ECU drives the factory boost gauge with a PWM signal based on it's idea of what the boost gauge should read given the RPM, temp, and airflow. It's a simulation which is why you shouldn't take it for gospel normally. It this case it's pretty accurate about if the ECU is working.

And in this case it sounds like it time to have it checked.

Yeah, this is not sounding great, It's getting to the point where the CEL comes on, no ISC/IAC init and the CEL remains on (when I get the no start). There are no codes when I can get to the point where I can get DSMlink to connect.

I get that the boost gauge isn't the good source for boost, and was only using it as an example of an indication that something appears off as you also note....

Is there a list of what order of self tests the ECU goes through?

If you're getting fuel and have spark then you need to confirm compression and timing. If the ecu were malfunctioning the problem likely would not create spark. Usually no starts are very basic problems. My initial guess would be to check all the wiring for corrosion or bad connections. Areas like grounds and supplies. Check around the fuse/relay area under the hood.

Whats the battery reading at start and if it runs? You're sure you have 12.6 volts then start following wires.

When it starts the car runs great, thats what is throwing me off here. When I get the nostart condition, I still get spark, fuel (based on getting a signal through the injector plug). I'll check the suggestions you made before pulling the ECU later today.

EDIT: I just replaced the battery recently so it has plenty of juice to get going.

 

[steve]

Is there a list of what order of self tests the ECU goes through?

Not a good one that I know of for the 2G code.

Who socketed the ECU for DSMLink. Pictures of the circuit board?

 

[voodoogsx]

Not a good one that I know of for the 2G code.

Who socketed the ECU for DSMLink. Pictures of the circuit board?

It was sent to ECMtuning for socketing when I got it done...way back in 2007...

I can get pics in a little bit and plug it back in to confirm the ground lines. I didn't see anything wrong on the first go round, nothing loose. No burn marks or scorch marks/spark marks. Did some continuity checks with some of the caps ground side. All checked out.

 

[voodoogsx]

@steve Images of the ECU
C201 looks worse than it is. After replacing the cap(brown one that had broken though didn't seem to affect the car from running) a while back, there was some sticky residue that came from the surface of the board....I assume it was some protectant as both side have this coating that flakes off fairly easily at this point.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

 

[steve]

The coating is what they call a conformal coating used to keep moisture away but sadly is also holds the leaked electrolyte in allowing it to corrode the circuits.

There is quite a bit of corrosion damage around C201 that will have to be checked out and it's not soldered correctly on the bottom. Time to send it back to ECMTuning for cleanup.

 

[voodoogsx]

The coating is what they call a conformal coating used to keep moisture away but sadly is also holds the leaked electrolyte in allowing it to corrode the circuits.

There is quite a bit of corrosion damage around C201 that will have to be checked out and it's not soldered correctly on the bottom. Time to send it back to ECMTuning for cleanup.

The conformal coating as mentioned got a little gooey so it looked a bit bad(I didn't clean it up at the time) but I was able to remove it which cleaned it up a bit just a little bit ago.
The bottom side solder joint, I clipped the ends of the wire as they were initially extended at least 1cm beyond the component holes. I tend to cut them close to the board...I assume you are talking about the fact that they legs aren't protruding from the bottom...

The caps I'm more wondering about are the two brown ones that are still on the board. C202, and and the one by IC104. They aren't leaking but are the older ones I believe.

 

[steve]

No I'm talking about the cold incomplete soldering joint and the corrosion to the top of the board in the area surrounding the cap. It's too bad this one got messed up, looks like it was nice up to the point where that cap leaked.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

Sure the others should/could be replaced. Might want to consider upgrading to V3 while it's being repaired.

BTW, why do you have wires soldered to the O2 sensor inputs?

 

[chrysler kid]

No I'm talking about the cold incomplete soldering joint and the corrosion to the top of the board in the area surrounding the cap. It's too bad this one got messed up, looks like it was nice up to the point where that cap leaked.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

Sure the others should/could be replaced. Might want to consider upgrading to V3 while it's being repaired.

Steve is 100% correct.

You can try to soldier those connections on the back of the main board, but bottom line that capacitor is bad and will do it again

It is in your best interest to send it in for repair if you have not done any major soldiering before.

Dont cut anything. If you haven't soldiered before dont mess with it and send it in. With the corrosion damage they may need to run new trace leads off of the board if the board is too damaged

 

[voodoogsx]

Steve is 100% correct.

You can try to soldier those connections on the back of the main board, but bottom line that capacitor is bad and will do it again

It is in your best interest to send it in for repair if you have not done any major soldiering before.

Dont cut anything. If you haven't soldiered before dont mess with it and send it in. With the corrosion damage they may need to run new trace leads off of the board if the board is too damaged

The cap was leaking. I replaced it. The current one is not leaking. There was a lot of residue from the coating/flux mix.
I've done a bit of soldering though at the time I didn't have my full setup available. Some solder flux helped but the residue mixed with the coating wasn't cleaned up.

I cut the legs as they were long, If not I'd have to bend them or they would contact the inside of the case.
The board doesn't appear damaged(or not as much post cleanup), again, it was the coating which flaked off when I cleaned it after I took the pic due to the comment about it looking like the corrosion had done some extensive damage. I tested the traces from the legs to the via that it routes to and to ground with no issue.
If you guys want another pic I can get a macro of the area.

 

[steve]

If you guys want another pic I can get a macro of the area.

Not needed, it's not going to change the result.

 

[chrysler kid]

The cap was leaking. I replaced it. The current one is not leaking. There was a lot of residue from the coating/flux mix.
I've done a bit of soldering though at the time I didn't have my full setup available. Some solder flux helped but the residue mixed with the coating wasn't cleaned up.

I cut the legs as they were long, If not I'd have to bend them or they would contact the inside of the case.
The board doesn't appear damaged(or not as much post cleanup), again, it was the coating which flaked off when I cleaned it after I took the pic due to the comment about it looking like the corrosion had done some extensive damage. I tested the traces from the legs to the via that it routes to and to ground with no issue.
If you guys want another pic I can get a macro of the area.

Unless you are saying that its fixed now and the car runs perfectly, I dont really need to see more evidence. You can try to follow the circuit along the board for continuity, but if youre familiar with that process you could bypass the board with wire to complete the circuit.

Or just send it in to the experts