Motronic Injector Signal Testing

[vreference]

As far as I can tell the injectors are receiving no pulsed ground. I've connected my scope up to one of the injector pairs and I don't see much of anything but noise. Voltage appears to come up while cranking. What frequency should I be looking for? Maybe it's there and I'm just looking in the wrong place.

Strong signal out of reference sender, Good looking sine out of Speed - both well over 2v. CHT sensor giving good measurement. Pretty sure I'm getting a solid spark on TDC (looks to me like It's only firing on TDC of cylinder 1? Does that make sense?) 12v at injectors, injectors click when grounding 15/14 at the ECU. I'm looking at the list of things that can cause no-fuel and I just don't see it. (fuel pump is running, flow test passed... although to my mind it wouldn't matter anyway with what I believe is still no signal to the injectors)

I've checked all sensors (would be nice if I had just missed something) many times over, had a second DME in it and both have gone out to SteveW - says they are both good. The car has been in the garage since 2008.

[JJ 911SC]

Check the plug at the fire wall?

[vreference]

Yeah, since I appear to be getting good connections at either end (DME and Injectors), it seems like the connection there is good.

[Tippy]

Courtesy of "lorenfb":

Waveforms

[vreference]

Originally Posted by Tippy

Courtesy of "lorenfb":

Waveforms

I could be wrong on this, but my understanding is that the low-impedance Motronic injectors don't have discrete firing pulses like a traditional part (and the only graph they have). I've heard that they all receive the same relatively high frequency signal... which I'm guessing might be easier to miss than the ~2hz spikes I'd expect from most cars during cranking. ...I don't know, I've never seen it.

[Tippy]

Originally Posted by vreference

I could be wrong on this, but my understanding is that the low-impedance Motronic injectors don't have discrete firing pulses like a traditional part (and the only graph they have). I've heard that they all receive the same relatively high frequency signal... which I'm guessing might be easier to miss than the ~2hz spikes I'd expect from most cars during cranking. ...I don't know, I've never seen it.

Over my head. I'd ask "lorenfb", "ischmitz", or "Steve Wong".

[vreference]

Well, kind of grasping at straws here. Decided to look at the crank position pickups one last time. Showing exactly 2v sine on speed sensor and a little over 4v on the Reference. Technically supposed to be 2v on reference per Bentely - I can hear the box making a discharge kind of sound every time the reference comes around and I want to say this is associated with the signal to the coil. I guess my speed sensor is a little low.

[rusnak]

This may sound idiotic, but what is the battery voltage? What is the voltage at the fuel injector harness? I know you say 12 volts, but what exactly?

And have you tried just squirting some starting fluid into the intake while the engine is cranking over? That's probably the surest way to check for spark and no fuel.

[vreference]

I'm using an older but pretty good analog meter, unfortunately 12 volts is sort of in between ranges to really read tenths so it's not ideal for this. Considering the connection is strong enough to energize all injectors (they click) I'm pretty sure they are getting good power. I want to say the entire motronic system can operate well under 12v and considering it's a brand new Redtop and it turns over easily I don't think the battery is suspect.

It will fire intermittently with a bit of gasoline poured in the intake. Normally, the air coming out of the exhaust is just oil tinged air. When I energized the injectors last night, the next time I ran the starter was the first time I'd smelled gasoline out of the exhaust in a while.

I re-adjusted the crank pickups (found the cardboard still taped to the speed sensor from the last time, I had expected it to be peeled off - did it sans tape this time and removed the cardboard) and I'm now getting over 4v on both reference and speed. Since I can hear the DME firing the coil, I want to say it's changed; I can now hear multiple discharges per rotation but it seems too intermittent. I've only looked at spark on #1 which was firing 100% reliably.

It just occurred to me that I can use more than one input on the scope so I'm going to go look at signal to the coil compared to reference sensor output I guess. although really with absolutely no injector involvement, I'm not sure how much that will tell me. Far as I can tell, still nothing to the injectors. I found some really high frequency noise, but it just looks like noise at 1ms/ - I would guess from the starter.

[rusnak]

well then I just don't know. I was going to suggest the factory DME Test Plan, but I think yuou might be on to some sort of new technique/ breakthrough

[vreference]

Originally Posted by rusnak

factory DME Test Plan

I don't know what that is.

I've literally run every test listed in the ECM Harness Electrical test chart.

I haven't really spent THAT much time with the injectors but... I've heard tell of them shorting internally and causing problems. ... I guess I'll be testing them individually soon but I'm pretty sure I tested both banks for resistance a couple years ago when this started and they came out in range.

I actually did put the scope on the harness with all of the injectors unplugged and still got no ground... but then it turned out I either forgot one or didn't unplug it fully or it just moved back mostly into it's old position so I'm not 100% on that test, I'll have to redo it sooner or later I guess. ...assuming no-load would tell me anything.

[steely]

Just out of curiosity, did you say you are trying to read the injector pulses in circuit?
I am stumped about your reading good ref signals too, since the DME will prevent the injectors from firing if it doesn't get good ones. *On edit - I think it kills power to the fuel pump instead.

[rusnak]

Just do this. Remove the air cleaner, and spray some starting fluid into the intake while someone else cranks the starter. It might take a few tries and you might need to use more starting fluid than you think you do.

After it starts, remove each injector connector one at a time until you find one that does NOT result in the engine running slower. That injector is suspect.

You might also have stuck injector(s) (one or more) if the car has sat for awhile. If no bad injectors are found, then you might only have a massive vacuum leak somewhere.

[Tippy]

I was told that a overcharging voltage regulator can burn out the injector drivers in the DME. I hope not, but throwing that out there.

[vreference]

Had a friend over messing with it just now and noticed a very odd behavior that might mean something to someone. While moving the throttle (I believe any time the idle switch is not activated), the ECU's signal to the coil goes completely bananas. You can hear the regular beat of the ECU firing turn into just static and the signal to the coil reflected this as I watched it on the scope. There was some debate over whether the injectors might have been firing during this epilepsy as well but I never put anything on them to try and look for signal. The gas smell out of the exhaust may still be from fuel I put in earlier.

When the throttle is at idle, I get great spark on #1 (it misses maybe 1 and 10, likely the loose test spark plug's connection to ground), I pulled #2 and #6 and found no spark. I didn't pull #4 to confirm my suspicion but I expect I'll find reliable spark there too. It only appears to fire the coil when the reference sensor comes around. Am I correct in thinking that the reference sensor is only signaled once per revolution?

Originally Posted by steely

Just out of curiosity, did you say you are trying to read the injector pulses in circuit?
I am stumped about your reading good ref signals too, since the DME will prevent the injectors from firing if it doesn't get good ones. *On edit - I think it kills power to the fuel pump instead.

I tried to read the reference signal with the computer attached a few minutes ago and that did not work. I didn't really get a signal to show up on the scope and the computer stopped firing the coil.

Now getting a read from the reference or speed sensors with them not connected to the ECU is easy since they generate a voltage on their own. (They aren't hall-effect). I'm just connecting probes to the two appropriate pins on the ECU's connector and turning the engine over. The third wire is a ground for the shielding. Unfortunately it looks like getting a reading from them while the computer is attached is harder.

Originally Posted by rusnak

Just do this. Remove the air cleaner, and spray some starting fluid into the intake while someone else cranks the starter. It might take a few tries and you might need to use more starting fluid than you think you do.

After it starts, remove each injector connector one at a time until you find one that does NOT result in the engine running slower. That injector is suspect.

You might also have stuck injector(s) (one or more) if the car has sat for awhile. If no bad injectors are found, then you might only have a massive vacuum leak somewhere.

Spark is not so much intermittent as it is only when the reference sensor comes around. So it will occasionally hit on a couple cylinders if you add fuel manually (1 and 4 unless I miss my guess.)

As far as an overcharge is concerned, I've heard that as well but both computers that have been in the car went out to Steve W after the fact. Including the one that ran the car for ~15 seconds upon first start and has never done more than hit on a couple cylinder sense. He said both computers were in working order when he got them and shored up the usual cracked joint on the ignition side. so I expect that this is not an overcharging issue.

[steely]

My question was directed to the injector gnd signal you said was missing in your first post, not the reference/speed signals. May be a moot question if you are getting spark now.
Yes, reference is once per revolution - and yes, you could 'swamp' it's signal when measuring it in ckt.
In the 1st para of post 15, you would not get injector fire if the engine isn't turning.
If the DME is good, I'd suspect SPD & REF signal connections and/or your injector connections at the DME card edge connector. How's that look?

[vreference]

Originally Posted by steely

My question was directed to the injector gnd signal you said was missing in your first post, not the reference/speed signals. May be a moot question if you are getting spark now.
Yes, reference is once per revolution - and yes, you could 'swamp' it's signal when measuring it in ckt.
In the 1st para of post 15, you would not get injector fire if the engine isn't turning.
If the DME is good, I'd suspect SPD & REF signal connections and/or your injector connections at the DME card edge connector. How's that look?

Speed and reference signals measured right at the DME connector look very good. very strong sine on speed, well over 4v. Reference also strong and regular over 4v.

I was measuring the injector ground multiple ways. All while cranking, Probing 12+ and ground of one injector with it connected to the injector. Probing ground and a different 12v source. and probing one injector lead 12+ and ground with all injectors disconnected.

All of the sockets on the DME connector look uniform and springy to me. Before I sent the DME out I was working with it open, straight from the pins rising off the board and was able to trigger the injectors then as well so I'd say the DME has a solid connection.

Something tells me I won't be able to ignore the DME output to the coil going completely haywire when the throttle is actuated. Whether it's the lack of signal to the idle switch, WOT being triggered, the output from the AFM changing or something else, I don't know yet.

[Lorenfb]

"I could be wrong on this, but my understanding is that the low-impedance Motronic injectors don't have discrete firing pulses like a traditional part (and the only graph they have)."

Yes, you're are wrong. The injector signal is as shown here: Waveforms.
Once the injector driver reaches it pre-determined current limit, the trailing edge of
the signal rings until the driver is switched off.

The 3.2 DME ECM is very simple to troubleshoot and doesn't require the use of a scope.
All that's needed is a VOM and a test light. Many can troubleshoot it in about 20 - 30
minutes. Most usually assume the problem is the DME ECM and waste time and money
sending it out and having it rebuilt when some rebuilders indicate a problem or a potential
problem.

[vreference]

Originally Posted by Lorenfb

Yes, you're are wrong. The injector signal is as shown here: Waveforms.

Well that certainly would make it easier to spot. But not for me I guess. Largely wishful thinking on my part. Noid light, LED test probe, scope... no discernible signal. ...also no fuel coming out of them unless I ground them.

I've checked all grounds leading to DME, power to DME -> ICV humming with key on, Reference sensor, (was told AFM can't cause no-start) Under load the signal looks gorgeous, Faulty signal to coil, CHT reading 5K @ ~32f, No signal to fuel injectors although connections test good.

That rounds out the no start checklist for things that seem reasonable to check when you have erratic spark and no fuel. (CO/Idle Speed & Fuel Pressure don't seem as relevant here). There's a long pinout /testvalue list in the Bentely manual, I've been down it too.

[hcoles]

Aren't the Motronic injectors operated with PWM by way of a ground connection? I can't find the modulation frequency using Google. I also don't remember the voltage involved as mentioned it may not be 12.