specsalot

Perry,

Didn't mean to lose you. Unfortunately we're on a very short foot race when it comes to the computers. You've either got wiring issues or computer issues. I suppose its possible to have issues with both at the same time, but that would be really brutal. It's probably fair to believe that wiring issues could lead to computer issues. But it's more likely that voltage surges or stray voltages to ground are the most common cause of computer issues.

If I am remembering correctly you are getting fuel but not spark? My belief is that to get spark, you must have communication of valid signal across all 3 DME / KLR communication paths:

DME 21 to KCP 9 - This signal should cause your tach to bounce when cranking the engine. I don't think you're seeing any tach bounce.

DME 31 to KCP 24 - I'm not positive about how the KCP uses this, but I think it's a reference clock associated with the KCP's listening for engine knock. If the KCP hears combustion knock, one of the actions it can take is to retard ignition timing.

KLR 16 to DME 32 - This is the revised ignition signal (retarded as required) that actually fires your coil via DME circuits.

The FSM says you can run without the KCP by jumping KCP plug pin 9 to 16. You've done this without success. Most likely this didn't work because there is no square wave passing from DME 21 to KCP 9

One thing worth trying if you haven't already done so is to verify continuity between each of these pairs of connector pins. Unplug the computers and check each pair for continuity. Also check each pair against ground to be sure none of these paths is shorted to ground. This would at least take broken / grounded wiring out of the equation.

I wish I could provide better insights.

specsalot

#20 has +12V because the first stage contacts of the DME Relay (Terminals 30 to 87) are closed. This supplies B+ voltage to the second stage DME Relay coil. The return path of that coil (Terminal 85b) is DME pin 20. Measuring with the DME disconnected will always show B+ because there is no current flowing.

If you are getting fuel injection, this circuit of the DME is operating correctly. Based on my own experience if the speed sensor signal is being received the second stage of the DME relay will be closing. I think people have reported both issues (spark - no fuel; fuel - no spark) in these forums.

I still haven't solved my boost issue despite replacing what was a failed hose. I have more testing to do as far as leaks goes. Car makes a funny sound like some kind of leak. Turbo can't produce more than 0.3-0.4 bars above atmospheric. I was able to pull blink codes and got 3-1 (low boost) - I though - No kidding - Didn't need a diode in the test connection to tell me that..... So I'm still fighting my own fight.

PerryB

Specs,

Yes, the 12 volts at DME harness pin #20 comes through the first stage of the DME relay. The DME computer is supposed to ground this circuit so that the second stage of the DME relay will supply power to the fuel pump. Currently (no pun intended) it isn't doing this. Prior to this, it was grounding the circuit and the fuel pump would run when the key was turned on. I can get the fuel pump to run by grounding this circuit. Again, as a new symptom this points to a failing DME computer.

I have tested for continuity between the DME and the KLR. There were no short or shorts to ground.

The boost issue that you're experiencing could also be a problem with your KLR--again, there's no way to directly test this computer.

Are you running the stock wastegate? How have you pressurized the intake to check for leaks? Have you put a vacuum gauge somewhere on an intake port? Can you pinch off the pressure line to the wategate then drive the car to see if the overboost protection comes on?

I guess that for both of us there needs to be some more conclusive way to test for faults. Your comment about the blink code was hilarious! Let me know if I can help out in any way.

Perry

Van

It shouldn't do this - that's not how the circuit is designed. The DME needs to see an RPM signal (or the start signal) to initiate the fuel pump side of the DME relay.

specsalot

Thats what I was thinking - Thanks for stepping up. This is why my thoughts are that the crank and speed sensor inputs are not being recieved in the DME. Could be cracked solder joints, etc...

Both inputs are combined in one IC package. If this bit the dust its game over for having a start. Specialized was not very specific in their descriptions. I dont think they ever are. My Initial DME failure was cracked solder joints, and some failed voltage comparators (IIRC). So some issues were pathway issues, others were component issues. Second go round, I sent DME and KLR. They said something in the KLR smoked components in the DME. Had to have both fixed to have a running car.

In Perry's case, both components have been repaired and retested. He also substituted a spare DME. Hard to concieve of two back to back DME failures.

Tiger03447

Gentlemen: Perhaps I missed this on an earlier post..but is the firing order correct from the distributor? If it's too far off you can crank until the cows come home and nothing will happen..Are you getting spark at all? Will it fire on ether? You DID put the rotor back on the distributor shaft...right? and the coil wire is hooked up? Sorry for being captain obvious here..but sometimes it's simple stuff that will drive you right around the bend...and send you on a wild goose chase..good luck with the car...I've got my engine totally out and am scared about all the hookups etc...Thanks, Tiger

specsalot

Perry,

Per Vans comments, I think your DME may be OK with respect to not grouding Pin 20 unless the engine is cranking.

I did a number of careful test drives with the waste gate line clamped off. Car went into limp mode with blink code 3-1. I even did some "key off" events at stop lights to clear the fault and drive again. No joy.

The ISV hose was definately a source of boost leak, but the car still makes a pronounced loud exhaust / hiss sound when I hit the accelerator. Only a very faint attempt at turbo spool up.

I had previously stroked the waste gate using a pressure source. It seemed to close well. During the test drives, I also disconnected the actuation line for the recirculation valve on the chance that it was getting vacuum to open the recirc valve. Still no joy.

I had the car bottom out on a very irregular piece of pavement (going slow). It may have loosened something up internal to the CAT. The car got a nice jounce when it bottomed out.

My no start mess killed my car budget, so I'm working on the cheap trying to hit low hanging fruit. I'm going to do compression tests to rule out burnt valves or a blown head gasket.

If the KLR was pulling timing out, the only way I could see this would be with a dual trace o-scope looking at the square waves (trigger and final KLR return to the DME). If I wind up in that situation I will be a very unhappy camper.

Thanks for your support. If you want to ship me your electronics (DME/KLR) I'll plug and play them again to see if they will fire my car.

specsalot

My golden rule is when the car breaks the first thing I ask myself is "what was the last thing I did to the car". Good questions but I'm sure Perry covered these bases some time ago.

Its no coincidence that Perry got a no start after a clutch job. Same thing happened to me in 2010. Disturbing these old wiring harnesses and connectors is a recipe for trouble with these cars. But it's really better to think in different terms. Namely that Bosches 24 year old wiring and connectors have probably held up better than most others over the same time frame.

Have courage hooking our stuff back up. When I did my clutch - first thing I did was buy a bundle of paper tags. Every connector and vacuum line was tagged at each end when removed. Taking lots of pictures helps as well. Hopefully you've documented your deconstruction.

PerryB

Specs, Van, Tiger,

Thanks for your input.

Tiger, don't worry about stating the "obvious." It might be exactly what I need to fix this problem. To answer you question, yes, the firing order of the ignition wire is correct and the rotor is solidly affixed to the cam. The problem is that the coil doesn't fire at all--no spark under any normal conditions. Everything seems to be fine except that the main computer doesn't fire the coil. Why this fail to happen is the subject of all this misery and confusion.

Van, Specs, okay, I'll agree that the DME will only run the fuel pump during cranking, so this "symptom" is a normal condition.

Specs, there is a picture on this forum (I think that it is this forum) where someone has a low boost problem and it turned out to be a collapsed inner wall of the downtube between the turbo exit and the catalyst. Unfortunately it will most likely take a visual inspection to confirm or rule out this type of diagnosis with either the catalyst or downtube or any other part of your exhaust.

Good Luck!

Perry

specsalot

Perry,

I've seen that pic. Made an attempt to look up in there but all I could muster was an inspection mirror. Yep - a flex cam is what it will take. I'm going to tackle / rule out basics first. When I road tripped to Tallahassee last spring, the car ran well. Round town driving in Tallahassee, the boost started flagging a bit. By the time I got back home (300 mile RT) the car was in limp mode and making a loud hiss rather than the nice sound of a turbo spooling up. Strange symptoms.

Compressor side of the turbo looks good and spins freely. There is a barely perceptible bit of radial slop same for axial movement. No signs of any kind of metalic contact. So for now I'm ruling out the turbo per se. I have to update my "sticky wastegate" thread and see what kind of response I'll get.

divil

Wow this thread is incredible! I just read the whole thing, but it's a lot to take in so I apologize if I have missed something...

PerryB, as I read through it I began to think quite some time ago that the wires that take the ignition signal from the DME to the KLR and back must be shorted. I've seen you state several times that none of them are shorted to ground. But I only saw that you *think* none of them are shorted to any of the positive wires. I think this is more likely than a DME fault.

Sorry if I missed this, but I didn't see if you mentioned what voltage is on KLR #9 and #16 with the DME *disconnected*. Because if the 12v is only on those 2 KLR pins when the DME is connected, I don't think that necessarily proves it's the DME's fault. It might just mean that the short is with some wire(s) that only get power when the DME is connected. If the KLR pins have no voltage with the DME disconnected,
I would check every DME pin for continuity with those KLR pins. (EDIT: I wrote that in a very backwards way the first time lol...tired. )

I have to qualify all this by saying I know less than probably everyone else who has commented here, so this could be total nonsense. But I can't get past the fact that your DME/KLR will run fine in the repair shop's car. Remember, they *did* claim to fix an issue where the KLR was damaging the DME. My guess is that is what stopped them from running in Specsalot's car earlier (unless I have mixed up the order stuff happened in on this thread, for which I might be forgiven ).

PerryB

Divil,

Thanks for your input. I have tried to isolate the wiring between the DME and the KLR during my tests. There is perfect continuity between the two female connectors in each harness connector. There is no continuity between these female pins and ground or the B+ terminal. I only ever get the 12 volts at the two KLR pins #9 and #16 when the DME is installed and energized. These pins never have 12 volts at any other time. It seems pretty clear to me (which doesn't make it correct) that the DME is sending 12 volts to this pins when it shouldn't be.

While Specialized ECU Repair (the guys that rebuilt my DME and KLR) say that they have installed my computers on a running car and had them work, I have no independent proof that this in fact happened. I know that this is a accusation that I cannot prove and I'm not really accusing them of fraud, but it is just one possible explanation for this disaster that there is still something wrong with my computers, despite Specialized's assurance that everything is okay with their repairs.

Please continue to make any suggestions to this thread for things that I can test.

Also, I will be in the San Francisco Bay Area next week. If anyone knows of a shop that can test my computers, or someone that is willing to install my computers in their car in this area I will greatly appreciate it.

Perry

divil

The part in bold is the thing I was getting at - my point was that maybe there is more than 1 place the 12v could be coming from, not just the wire that powers the DME. For example, the DME completes the circuit for the fuel pump, and (I think) the O2 sensor heater wire via the DME relay. So there is surely a wire or 2 somewhere in the harness that is not the main power supply, but nevertheless sees 12v only when the DME is powered up. Could your KLR pin 9 be shorted to such a wire? Again I'm sorry if I'm going over something that you've already checked. You have my admiration, not only for sticking with the troubleshooting, but also for your patience in keeping the thread gong!

PerryB

Divil,

No, you are correct. I'm just not sure how to isolate those wires.

In other words, how can I test to see where the 12 volts is coming from? I know that it comes from the DME, but which output?

Maybe I could take apart the harness and remove pins individually to see which ones get a 12 volt output? Is that what you're getting at?

I could try that tonight, if I have time after work.

Please don't hesitate to ask these questions! I hope that they will lead me to a solution.

Perry

divil

I'm honestly not sure what the best way to go about checking it is...I just wanted to make sure it wasn't being overlooked in principle. When I wrote that first post I was thinking, maybe the DME outputs 12v on certain pins, by design, for things like the AFM, O2 sensor heater etc. So I thought it would just be a matter of checking continuity between KLR #9 and all pins on the DME connector that might output that (or just all pins). But now I don't think it's that simple. For instance, the AFM only gets 5v, so that can't be it. The O2 heater power is not directly powered from the DME, the ground side is - but maybe that could still be the cause somehow?

Maybe taking out DME pins until your mystery KLR #9 voltage goes away is the way to go - but I would be cautious about powering the DME with pins removed in case that could damage anything.

EDIT: Actually, if you started by removing the DME pin that KLR #9 is supposed to be connected to (21 I think?), that might be a shortcut to eliminating my theory. If the 12v is still present at KLR #9, I think that exonerates the DME.