No output from reference sensor
 

I've got an 86 944 turbo that won't start. Using a number of valuable post in these forums it seems I'm not getting any output from my reference sensor. I've got an analog volt meter that when I set to AC 15 volts I can read the output from the speed sensor OK. I've replaced the reference sensor (although I don't think the old one was bad anyway) and I still can't read any output. If I remove the sensor and tap it against something metal the meter will see the signal.

So my question is does anybody know what could cause this?

It would seem like if something happened to the gap the speed sensor wouldn't be signaling anything either. I've not tried to reset the gap because I've no data to indicate it would have changed

Seems to me that if the sensor produces a signal when it nears a ferrous metal, then something has to be wrong with the clearance or the union of the leads in the engine bay.

I've never had to remove the sensors but I have had problems with the union that is several inches from the sensor. Do the sensors get actuated by the same cog?

I second Lawrence's opinion. If you're confident that gap is ok, and you get a signal when it's connected to a meter, then the likely cause is a bad connection between the sensor and the DME. I would try putting the sensor back on the car, then looking at the output on the meter while cranking the car. If you can still see the signal on the meter, then it's probably the connection.

I'm not sure what activates the sensor. I can't seem to find my Haynes manual and I've not had any luck finding any pictures of the mechanism on the flywheel that the reference sensor picks up.

I wouldn't say I'm confident about the gap just that I have no evidence anything would have changed it.

I'm measuring the sensor output at the sensor. I can read a signal from the speed sensor which has it gap set in parallel to the reference sensor which isn't sending a signal. This while cranking the engine.

If I switch the speed and reference sensors I have the same problem. Speed sensor produces a signal, reference sensor doesn't. This while cranking the engine. Actually, I have three sensors and they all produce a signal when they're used as the speed sensor.

The "pickup" is a set screw (4mm iirc) raised above the surface of the flywheel. The speed and reference sensors pickups are on two different planes so your test result is certainly very possible.

On non turbo cars with rubber centered clutch discs, my guess would be that a piece of rubber was blocking the sensor tip but since you have a turbo this is not possible.

Short of removing the sensors and/or mounting bracket to inspect the set screw presence or height (8mm iirc), maybe you can remover the starter and inspect from there.

I've never tried it.

If you want a picture of the sensor pickup set screws, email me at 944t at bellsouth dot net.

TS

ref sen
 

I think even if your ref sensor is good you won't see it on a VOM because the duty cycle i.e repetition rate is much less than the speed sensor. So there is not enough signal to show on the VOM (it works by RMS (root mean square) ) so it has to have a signal that is large enough in amplitude and repetition rate, which in this case depends on the speed of the car cranking, and is not enough. Long story short you need a scope or you could use a sound card (mic input) you will need to check the spec on your sound card. the two signals are posted below.http://forums.pelicanparts.com/uploa...1161445823.gif http://forums.pelicanparts.com/uploa...1161445845.gif

VR sensors can be tested with multimeters. Connect the multimeter to the two leftmost pins according to the pic on clarks garage (the ones that get 600-1600 ohms). Then you need a wrench or pliers or something, bring the wrench close to the tip of the VR sensor then jerk it away. Youll get a pulse on the multimeter. You have to move the wrench closer and further fairly quickly to get a good pulse. Nothing will show if you just touch the two together and leave them there.

http://clarks-garage.com/shop-manual/ign-02.htm

Edit: Yes ive done this myself on 951 VR sensors.

I agree that the pulse will make the VOM register, but what you need to do if you want to be certain is to look at the input at the DME of the sensor, and unfortunately no substitute for a scope. My sensor was showing a signal at the connector next to the fire wall engine side, and it was showing a signal at the DME, but it was just a little weaker 1.volt peak to peak (some of the time ) drove me nuts :) I finally just replaced the connectors at the fire wall for both sensors. No more problem.http://forums.pelicanparts.com/uploa...1161480667.jpg

Update so far...

I bought an oscilloscope and have verified the reference sensor is actually OK. I've even checked the signals at the plug to the DME and both speed sensor and reference sensor are OK at the DME plug.

However, I have no no spark. Here's what I've verified (or had verified)

1.) DME relay works. I applied voltage to both sides and checked that if completes the corresponding circuit
2.) Coil works - completing grounding circuit manually causes spark
3.) DME rebuilt and tested. I had previously found a burnt trace and actually sent the first replacement back and had the DME reverified.
4.) Ignition switch works. I have power at the coil with the key 'on'
5.) Started solenoid signal OK. I have power at terminal 4 of the DME plug when cranking.
6.) Distributor cap was removed and visually inspected.
7.) Ignition wires are relatively new and tested.

I'm not sure what to check next.... any ideas?

Check that there is a ignition input to the KLR on pin 9 and if so then is there an out put at pin 16 KLR. Let me know what you find out.
ps. worst case there is a limp home mode to test your system pull the KLR and run a jumper from the pin 9 klr to pin 16 klr this should let you start up and go into a low power "get home mode".

When you check for ignition input on pin 9 of the KLR I assume you mean I should I should unplug the KLR and use one probe on pin 9 and ground the other.

What am I looking for and when?

I only had a few minutes this morning. When I unplugged the KLR and put one probe in pin 9 and grounded the other I was getting a 60Hz sin wave of about 3.5 volts. And the key wasn't even in the ignition.

The only other thing I did was jump pin 9 to pin 16 with the KLR unplugged and still got no spark.

How do I test output of pin 16? What am I looking for?

Thanks for the suggestions!

Sorry I should be more clear..
Check the input to the KLR from DME (ignition signal) Should be a square wave if the car was running. If it is not running as is your case the the frequency will depend on the crank speed. The 60hz sine wave sounds like an AC hum being picked up by a bad ground or if you have a battery charger hooked up. If you see no signal in to the KLR then of-course you will get nothing out on pin 16. Try looking at the pin 9 without the KLR hooked up. (just in case the input of the KLR is shorted) see pic. try that and let me know. My understanding is :
1) you have the right signal from the speed & reference sensor. (amplitude and frequency).
2) The DME relay (fuel pump) comes on for a few secs when you turn on the ignition.
3) no signal on pin 1 of the DME connector.

http://forums.pelicanparts.com/uploa...1161803931.gif

Also see if your tach needle bounces when you crank the car...

OK more data and clarifications...

1.) The tac bounces while cranking
2.) I've not verified anything about the fuel pump but I have pulled the DME relay and applied voltage to both circuits and verified that it completes the corresponding circuits using the instructions from clarks-garage
3.) According to my oscilloscope the signals from the reference and speed sensors fine when tested at the plug to the DME
4.) I don't know how/what to test for at pin 1 of the DME

I put the scope on pin 9 of the KLR when unplugged and got a flat 9 (ish) volt signal when cranking (after removing the trickle charger)

The KLR may be bad but jumping pins 9 to 16 which, according to SK, should allow it to start did not work so it would seem like there is something else wrong or am I missing something here?

"1.) The tac bounces while cranking
2.) I've not verified anything about the fuel pump but I have pulled the DME relay and applied voltage to both circuits and verified that it completes the corresponding circuits using the instructions from clarks-garage
3.) According to my oscilloscope the signals from the reference and speed sensors fine when tested at the plug to the DME
4.) I don't know how/what to test for at pin 1 of the DME

I put the scope on pin 9 of the KLR when unplugged and got a flat 9 (ish) volt signal when cranking (after removing the trickle charger)"

Pin 1 of the dme is just the output to the ign coil. THis is very weird because if your tac is bouncing then pin 9 on the klr should have a signal, same connector. Also if nothing on pin 9 klr then jumping it won't do any good, as there is no signal in so there will be no signal to jump to pin 16. I am confused, if tac is getting a signal but pin 9 klr shows flatish 9v some thing is no computing, I know you have done this already but humor me...
Remove the klr connector, connect the scope probe to pin 9 and crank the car (trickle charger disconnected) should be a square wave of 9 to 12 volts. If not then connect the probe to pin 21 of the dme and do the same test (remeber not to connect the klr). This will eliminate the connection from the dme to klr. If no signal then the tac is getting some funky stuf from somewhere else...
good luck and patience :)

Sorry also connect the scope to fuse 34 in the fuse box. Set it to trigger on a positive going signal. turn the ignition on there should be a positive going pulse which will go away in a few hundred mili secs.
Let me know what you find out.

User error on the scope. Here is what I got from pin 9 on the KLR.


http://forums.pelicanparts.com/uploa...1161902404.jpg

Still no start when I jumper 9 to 16.

Should I still do the fuse 34 thing?

Ok now we are geting some where...
put the scope on pin 16 KLR with the KLR in and tell me what the scope does. (nice scope btw)
if there is a signal similar to the one pin 9 then check pin 1 dme same way. the fuse 34 is just to confirm that the fuel pump is comming on momentarily. let me know I will leave my computer online so if you post it will let me know. Lets see if we can knock this out tonite..
luck

SK,

$99 USB scope should've got it sooner.

I looked at one of your other threads and saw how to test pin 16 with the KLR plugged in. I was missing the part about removing it and taking it apart.

Anyway, that and dinner were a bit of a delay. The output on pin 16 was flat 9 volts ish

so you have the square wave in to the klr at pin 9 (with the klr installed) but flat 9v at pin 16 ?

yes

ok now lets try the limp mode pull the klr off and jump the pin 9 to 16. you will have to use a jumper of some type (piece of wire with pins on the end ?) when you do this put the scope on pin 9 see if the squre wave is still there or did it get pulled down (gnd) or up to 9+volt. If it still shows then go look at the dme pin one and see if the ignition signal shows there..

I'm not sure what to make of this. This is pins 9 and 16 jumpered with the KLR unplugged. There is no trickle charger attached.

http://forums.pelicanparts.com/uploa...1161917606.jpg

looks like the dme may be pulling the line down. I don't figure you have any way of only disconecting one pin on the dme? because we need to look at the klr output with out the dme loading it. What we are trying to achive is to have the signal go from the dme to klr's pin 9 but not hav the pin 16 of the klr conected to any thing...let me think for a sec if you have a sugestion fire away but we need to unload the klr output line to check if its a klr output drive problem or a dme input pull down problem

BTW the DME is newly rebuilt and I'll void my warranty if I open it. Let's think about it overnight ;-) I'll go catch the last few innings of the game

sorry you have already eliminated the klr by jumping it looks like the dme is pulling the line down...

ok later

Just thinking some more with the help of a few adult bevrages :) Looking at the schematic of the DME (wish I had a KLR one hint to any kindly soul in possesion of one) You can ohm the dme pin 32 to gnd it should be a couple of 100 k or more, if it is low like less than 1k then there is a short or partial short in the dme which could be the reason your trace looks pulled down, but some where back you said it was 9ish volts ?? btw how long did the new/rebuilt dme work before it went dead IF its dead :)

When the car didn't start when I got the 1st rebuilt DME I sent it back right away. When I got the current DME the guys at the place I got it from swore they bench tested it. I've no evidence it ever worked other than their word. I don't know exactly how they tested it.

If we're pretty sure the DME is still the most likely culprit I'm happy to send it back (again). It's still under warranty as long as I don't open it.

lets make sure we eliminate every thing else first. Do the resistance check from my last post first. I have superimposed the two traces (as the time base is the same) looks like the same signal but pulled down. There is only one connection to the klr out which is the dme ign in (pin 32). So if we ohm the circuit at pin 32 dme with the klr out but the dme in it will tell us if that line is pulled down. Some things to consider
1) first test with dme in if resistance is high then the problem is else where.
2) if resistance is low then unplug the dme and do test again to separate out the dme and the wire leading to it. if dme measures low R then it is (probably) the culprit.
3) if dme R is high then its the wire(best out come :) )
If case 1. above then we are going to have to scratch our heads some more.
Luck

http://forums.pelicanparts.com/uploa...1161959851.jpg

OK, I think I understand what I need to test (and why) but I need a little help.

1.) How do I measure the resistance from pin 32 on the DME to ground without opening up the DME and with it still plugged in?
2.) To measure the resistance from pin 32 to ground without the DME plugged in I need to know which pin is ground or if the case is ground.

Does anyone have the pin descriptions for the DME like the one in this thread for the KLR?

Thanks

you can test it at the klr pin 16 with the klr out gnd. is pin 10 or 11 on the klr also pin out for the dme...http://forums.pelicanparts.com/uploa...1161979939.gif

I'm getting an open circuit from pin 16 to pin 11 on the KLR plug

with the dme pluged in?..and the out put from the klr is still squashed? this is weired...to reetarate ( I can't spell right now)

input to klr is a square wave.
output is around zero.
if you jump the klr 9 to 16 still zero.
dme pluged in klr out pin 16 to gnd (pin 11 or 10 ) is open.

1)check if you have good gnd. With the klr out the gnd is pin 20 so check 16 to 20 (sorry)
if also open then... heck I don't know, does not seem possible. Some where we are mis comunucating.
the signal is present at pin 9 and but not at 16 even with the klr out and a jumper, but the the resistance is open ?
not possible signal into open should still be there.. (no current sink). maybe we shopuld try doing this real time like over the pelican parts chat room ...

............slow down guys.

This system is relatively easy to troubleshoot. I am assuming it is JUST a no spark issue. You have confirmed injectors firing, fuel pump running etc. etc etc

Step 5. If you wanted to test if the KLR is bad all you had to do was jumper 9 to 16 as has been said on the KLR connector and the car should still run (without boost-control or knock-detection & ignition retard). No other tests needed. If it worked replace KLR.

Your problem is elsewhere.

I love the scope pics by the way...........

1. I am assuming the spark test was done at the center wire coming out of the coil. If you have spark look at distributor/wires.

2. I cant remember if you tested the ignition switch? I saw a coil test earlier. That you grounded it and had spark?

3. If the reference and speed sensor outputs while cranking at the DME harness are correct. You have a scope...........you should post some nice pics of them for future reference. Then we can assume the sensors, and wiring harness between sensors and dme are good.

4. Wiring from DME to coil. Pin 1 to coil. Make sure its not broken.

"............slow down guys.

This system is relatively easy to troubleshoot. I am assuming it is JUST a no spark issue. You have confirmed injectors firing, fuel pump running etc. etc etc

Step 5. If you wanted to test if the KLR is bad all you had to do was jumper 9 to 16 as has been said on the KLR connector and the car should still run (without boost-control or knock-detection & ignition retard). No other tests needed. If it worked replace KLR.

Your problem is elsewhere.

I love the scope pics by the way...........

1. I am assuming the spark test was done at the center wire coming out of the coil. If you have spark look at distributor/wires.

2. I cant remember if you tested the ignition switch? I saw a coil test earlier. That you grounded it and had spark?

3. If the reference and speed sensor outputs while cranking at the DME harness are correct. You have a scope...........you should post some nice pics of them for future reference. Then we can assume the sensors, and wiring harness between sensors and dme are good.

4. Wiring from DME to coil. Pin 1 to coil. Make sure its not broken."


I love easy...but if you look at the thread in this post it boils down to there is a signal out from the dme as seen on the pin 9 of the klr so:
1) ref sensor ok
2) dme internal signal condition to convert it to a squre wave and out put to the klr ok
3) when you jump the pin 9 to 16 of the klr the signal gets squashed (looking at his scope traces)
4) input of the dme/wire from the klr as checked by resistence between pin 16 and gnd pin 20 is very high.
So whats pulling the signal down ?
enlighten me please ??

sorry the post also contains ref and speed sensor pics I posted earlier

I'm reading 14K from pin 16 to pin 20 on the KLR plug with the KLR out.
When I unplug the DME this becomes open.

I'm happy to open up another line of communication... chat room , IM or whatever let me know. Also, I'm very greatful for all the help so far.

mcmahonl ok right now I am on kitchen duties breakfast :). let me finish and I will post again and you should get an e-mail alert. Then we can use the board chat room.
In the mean time 14k seems to be low, but it should only be a 650 micro amps draw at 9 volts (but thats dc). can you check if the car when the ignition sw is on and
1) Is there 10 to 12 volts steady on pin 16 of the klr (with klr in)
2) if not do you get it if the dme is unpluged.
3) if not try this test unplug the klr leave the dme in get a 9 volt battery or any other source of dc at that level put a 1 to 3 k resistance in line with it. feed the positive to the pin 16 f the klr plug (no klr) and gnd to pin 20. See if there is a drop i.e does the dme draw the voltage down. (trying to eliminate the ac coupled current draw if any)
All this means is that I am trying to find out if the dme draws more current at ac (car cranking and producing a square wave) versus straight dc. Let me know.