jfdturbo

I'm debugging my 951S with 25k miles on it. For some reason the KLR box
seems to have failed and is not passing the ignition signal through. If I jump pins 9 and 16 in the connector with KLR out of car it will start. Wondering if anyone has a schematic for it.

I'm thinking it is probably a simple capacitor that went bad over time. The car
spends all of its time on a lift in my garage, never gets driven in
moisture, so it has had a pretty easy life so far.

I can’t even get the LED line to drop to ground, so my microprocessor may not be running. Does anyone know whether the LED line should drop low if the microprocessor is operating correctly and you simply switch the ignition on?

Any help would be appreciated.

Thank you,

John

Tom M'Guinn

I'd start by checking the solder joints on the mother board in the KLR where the main connector mounts to the board. You have to pull both boards out to get to it, but those solder joints are prone to cracking since they carry both an electrical signal and bear the stress of mounting the connector. Here is a thread with some fuzzy pictures...

https://rennlist.com/forums/944-turb...dle-issue.html

I'm not aware of schematics floating around for the KLR, but most guys end up swapping it out to see if that's the cause or not. In your case, you probably already know the KLR is bad, as long as the harness signal ohms out between the KLR and DME. If it's more complicated than cracked solder joints, there is a place in Florida -- ecudoctors -- that will rebuilt it in a flash.

Not sure what you mean when you refer to the LED signal?

Welcome to Rennlist.

refresh951

Good to see you here John. Welcome.

jfdturbo

Thanks Tom and refresh for your replies.

The LED that I referred to is the diagnostic port LED that flashes codes to debug the KLR system. I was thinking that if the CPU were running the code properly that I might see the LED on all the time, but that line is high, so no voltage between it and the positive terminal in that diagnostic port. Thus, LED will not light up. When I first turn the key to "on", I get a little spike when that LED line goes low, but then immediatly goes high. So, I'm thinking that the CPU is running the code, but I really don't know.

Knowing if the CPU is running would help me narrow down the fault. Right now I'm suspecting the transistor that feeds the ignition signal out of pin 16 and back to the DME. That transistor takes an input at its base from the CPU port and probably acts as an isolator for the line out to the DME.

I will go back with a magnifier and look at all of the solder joints - great suggestion Tom - thanks.

John

jfdturbo

After very close inspection with binocular magnifiers, all the solder joints look good on both boards. I think it must be something else.

I am still looking for a definite answer about the LED. If the car is off and you just switch to on without starting, should the LED in the KLR diagnostic socket light up or stay off? Mine stays off and I'm thinking that might be bad.

If anyone knows the proper operation of this diagnostic port - please reply. Thank you.

John

Tom M'Guinn

I'll see of I can test a known good KLR and report back. I powered up a core box I had sitting around from eBay, which was claimed to be good, and the LED line did indeed sink to ground (i.e., turn on an LED) but I have no confidence that KLR is working correctly.

Tom M'Guinn

Ok, I put an LED in the diagnostic port of my working car to confirm. When the ignition is on but the engine is not running, the LED is on solid. When you start the car and let it idle, the LED goes off. If you pull the TPS connector off while idling, you start getting blink codes (4 - 1). Whether it would act the same with the ignition bypassed, I don't know.

jfdturbo

Hey Tom - thank you so much for checking that out!! Based on what you said, I hooked up the KLR to a 12 volt supply and started to measure voltages around the board to figure out why my LED would not light.

It turns out that my Intel processor has no power (measuring voltage between pins 20 and 40).

So, my next step is to figure out why the power supply section is not working. It should take the 12 to 15 volts from the battery and regulate it to 5V. I noticed that Bosch didn't use a typical 5V regulator, but instead they have a circuit with discrete components and a power transistor at the output. I'm in the midst of trying to figure out the circuit they use and narrow it down to which component(s) failed.

I'll keep you posted.

Thanks again!

Rogue_Ant

The circuit is an early switching power supply design...

Tom M'Guinn

Makes you wonder what would happen if you supplied your own regulated voltage to the rail?

Rogue_Ant

That should work fine.

jfdturbo

Making progress!! I found that the power transistor that is the final stage of the +5VDC power supply within the KLR box was bad.

I found that I could replace it with an NTE332 power transistor. Great thing about the resources on the web. The original ON946 transistor crosses to a BD202, which crosses to an NTE 332. When I pulled it out I found that the collector to base measured at 0.3 ohms. It had definitely shorted.

I did some tracing to find the attached schematic. The transistor in question is T300.

Previously Vcc measured at zero volts, and last night I just replaced the bad T300 with a new NTE 332 from RadioShack and I now get 4.97 volts at Vcc. I measured this between pins 20 and 40 of the Intel microprocessor.

Hopefully in the next couple of days when it warms up I can get the KLR box back in the car and see if it runs!

I did notice that when I power up the box the LED line drops low, giving power to the LED that can be used in the KLR diagnostic port, so that is a good sign. Previously that line stayed high which would not light the LED.

Tom M'Guinn

Great progress! Any idea what cause the transistor to fail?

jfdturbo

I'm not sure what caused the failure. Hopefully it is not from something drawing too much power. Often power transistors are the things that fail because they are passing a significant amount of current and are some of the more highly stressed components in a circuit.

I'm hoping that once I get it back in the car, that if something else is wrong the microprocessor will send a blink code that will clue me in as to what it might be.