3.2 Voltage at Injectors at key-off position
 

Greetings,

Quick question. on my 85 911 3.2, what voltage to ground (if any) should I measure at the injector harness/plug when the key is in the off position? And, should the voltage differ for each pin on the plug? I'm trying to figure out if its normal for the inject to retain some voltage after key-off, possibly retained by the DME.

I did a test and found the pin's on my injector harness has 12 volts even at key off position (after the car was running), but as soon as I unplugged the DME relay, it went to 0.

Thanks.

Just a quick addendum to my question above. I did this testing when the car stalled and would not start. I measured 12 volts to ground on both injector pins in the middle injector on the drivers side. I then pulled the DME relay and the voltage disappeared. When I reinstalled the DME relay, the car started.

Thanks.

I'd say the only way that could happen is if your DME relay was spot welded closed from high current draw.

http://forums.pelicanparts.com/uploa...1343073107.jpg

Electrical test for you.

What voltage would you expect at Point A and at Point B with each switch open and closed?

http://forums.pelicanparts.com/uploa...1344876392.jpg

Answer: With the switch open you will see 12 volts and with it closed you will see 0 volts in each case.

Here is the circuit that I think is the switch. http://forums.pelicanparts.com/3868545-post194.html

Rick - I'm helping Scott with this diagnosis. The DME relay was brand new when we did the test, literally 10 minutes of use. I haven't inspected it since but I doubt the contacts are anything less than pristine. Scott had about 6 DME relays that he has been swapping between and we looked inside several and the contacts were clean on all of them. So at this point I don't know where that voltage was flowing from - that is what we are going to go after next.

However, something I don't understand about the fuel injectors - it looks like they are wired together in parallel - how does the computer fire just one when necessary? What are the injectors themselves - coils of wire? I gather that is the gist of your second question, 12v on one injector harness pin is measuring the voltage at the junction off of DME pin 87. 12v on the other injector harness pin is measuring that same voltage but through an adjacent injector. If we disconnected all the injectors that are ganged together we could confirm whether the voltage on the second pin (wired to the CPU) is from the CPU or from the junction. That would be another useful test (I assumed it was from the CPU).

My working theory right now is that as long as that voltage is there, the CPU is getting power applied to pins that shouldn't be powered and that leaves the CPU in an odd state such that it can't start the engine. Pulling the DME relay killed the power and allowed the CPU to reset itself.

Russ

Even though there are two wires to the two sets of injectors they both go to the collector of T402 (represents the switch (single) in the simple diagram). All 6 fire at the same time. (this is 80's technology after all) They are all in parallel.

The injectors are coils of wire about 3 ohms and 2 milliHenrires. The DME uses the inductance of the injectors to control the current through them. The simple picture was supposed to show that if one injector had much much smaller inductance than the others the DME would just be basing its control on that one injector.

Do you have access to an oscilloscope you could look at the injector pulses when starting with it good and then bad?

Did you guys ever try pulling one injector connector at a time with it in a no start condition to see if it would try to start?

it fires the injectors in 2 pairs of 3's. (based on what i know of other EFI systems)

if you are concerned about why there was 12v at the injectors with the key off, i would look at what energizes the dme relay.

i think you are getting wraped up in this 12v at all or some injectors. i would pull the injectors and check to see if they fire and check for spark (1st). if you have at least One of these, you can say the CPS is good, if you dont have either, i would look at the CPS or the DME itself.

remember, fuel (pressure), spark (at the right time) and air.

Rick - hmmm... what I see in the big wiring diagram picture you posted is that injectors 6, 5, and 4 are wired to a junction positioned under injector 6 that is wired back to the DME Control Unit on the 2nd pin from the right side (can't read the number). Likewise injectors 3, 2, and 1 are wired to the DME Control Unit 3rd pin from the right. The other side of all 6 injectors are wired to the junction that connects to the DME relay. This junction has +12v from the battery on it when the DME relay closes.

So, my assumption is that when the DME CU "goes low" on that second or third pin, that effectively grounds that side of the injectors and causes them fire or activate, or whatever they do. When you say "two wires...go to the collector of T402" it sounds like you are saying all 6 injectors are wired to one transistor inside the DME CU - is that what you mean?

So then you say the DME CU uses the inductance of the injectors to control them. I'm probably at my limit of what I understand here but does each injector have a unique inductance that allows the DME CU to fire them individually? Bottom line, if I hook a scope up to the DME CU side of injectors 6/5/4 what kind of signal should I expect? Sine? Square? Some kind of high frequency analog signal tuned to the particular injector that is firing (e.g. the frequency of the signal is tuned to the target injector)? Obviously I will see the same signal on any of the 3 injectors but will I also see the same signal on the corresponding wire of injectors 3/2/1 even though they come from a separate pin on the DME CU?

I do have a digital storage oscilloscope but how would I hook it into the injector circuit? Is it best to try to do this under the seat at the DME CU or best to fashion a harness between the existing injector harness plug and injector (I'm thinking two short wires here) that lets me tap in right at the injector?

Thanks for your help and time,

Russ
Thanks, Russ

T77911S - If it fires two pairs of three how is it not firing gas into a cylinder on the wrong stroke?

I agree that finding out where the 12v came from when the key was off is essential. Unfortunately getting the car into that state requires running the engine at high RPM under load, e.g. flooring it. Engine runs great sitting in the driveway. Since we don't have a dynamometer, we do this by running the car down the street till it dies and then working on it at the side of the road... ugh.

The other thing we know is that, with the key off, pulling the DME relay made the 12v go away and the car started on the next attempt (after the same DME relay was re-inserted). Then we drove the car home and parked it being careful not to load the engine down. It worked fine.

You might be thinking "hmmm... high fuel load kills the engine - the fuel pump is drawing too much power through the DME relay" (at least we did) but we already addressed that chance by have the DME relay fuel pump wire power a relay that connects the fuel pump directly to the battery. Didn't make a difference.

I don't think spark is the problem since the DME relay seems to be in the loop with the problem and spark isn't anywhere near that. But we should probably check for spark the next time when we are cranking it and it's not starting.

Scott - have you tested for spark in that condition?

Russ

There is a constant 12v to the injectors in parallel. The dme fires them using the harness points 14 for one bank, and 15 for the other, which are dme controlled grounds. If you measured the voltage from the positive lead to chassis ground you should read 12v. If you test it from the gray/brown-gray/white you will get a lower reading from the losses of passing through the windings of the solenoids of the injectors.

You can download the 911 DME schematic from here DME Schematic Download
you can see on the left side of the analog board pins 14 and 15 are tied together.

Since you have a scope I would just back probe the connectors on the injectors and look at the grey/white wire or the grey/brown wire and put the probe ground to engine ground.

I haven't done this but users ischmitz and scarceller have and I'm pretty sure they will see this.

I think you will see voltage waveforms something like as shown in the bottom trace (inverse of course) of figure 4 http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00000086.pdf THAT IS NOT THE IC but I'm guessing it is of similar vintage. If one injector is shorted I'm guessing the timing will be radically different from the good trace.

Russ, et al.,

During the test's done in the past, when the car died, we did have spark and we did have fuel. To add a couple more tidbits of information, we've done a lot of work on this car, but with no luck. It seems to die at around 5k rpm's under full load, and the DME relay is hot to the touch. Two weeks ago, when it died, a test showed we had 5 volts at the fuel injectors (during the no start period), but Russ and I did not experience that this past Saturday.

Scott

The injectors all fire simultaneouly. It just varies the pulses to control the f/a mixture. Just like the cis, continuous injection fires continuously, it is not a direct injection system. The fuel air is there waiting for the valve to open to suck it in.
With the key off there should be no voltage to the injectors.
http://forums.pelicanparts.com/uploa...1344886168.jpg

One more thing, I have a running post on this issue for over 10 months now. As you'll see, I've replaced, repaired and testing quite a number of areas since this stalling problem started.

http://forums.pelicanparts.com/porsche-911-technical-forum/638355-3-2-engine-stalling-stumped.html

Rick, the links on that page are dead - can you email the files? Send them to rpsinet-auto at yahoo.com.

However, I understand now that this is not direct injection but rather just mixture control - the mixture sits in the intake manifold waiting for the cylinder to pull it in. So that explains why all injectors would be tied together.

I didn't measure any currents in the injector circuits because when mine didn't work I found the problem in the last place I looked. :) DC current probes cost lots of money. I'm thinking this is how they work from the other thread
Even more electronics gibberish:

• Inductors add in parallel just like resistors. 6 - 3 mH in parallel = .0005 Henries
• The voltage across an inductor is proportional to the rate of change of the current through it. di/dt =V/L If you want X amps to flow wait T amount of time when you apply 12 volts. This time gives the DME time to react to change the drive on and off to obtain the desired current.
• The inductance in the coil changes as the slug moves to open the injector. The field will move the slug in a direction that increases the coil's inductance

This is why they call it a peak and hold driver:

• All the DME does is provide a low impedance to charge the coil(s) to a set current and move the slug. Remember it measures the current.
• After a time has elapsed where it thinks the slug has moved and the injectors are open it backs off and provides a hold current.
• To turn them off it dissipates the energy in the coil and moved slug in a zenner diode.

Rick - Great info. I just read through Scott's other thread (something I had not done before) and see that you had posted lots of info over there. My apologies for you having to double post. I'm going to move this discussion back over there for continuity. I'm putting another post up on that thread.

Russ

This will work DME - Schematic Download