The patient is '84 Porsche 928 4.7L 5 speed (sorry guys ) getting a GM 1227749 transplant. The '749 has been modified with Fairchild HUFA76429P3 injector drivers to run the 8 low Z injectors. Injector pulses are MIA. It starts if we spray starter fluid in the throttle body, but Stephen objects to sitting up there and spraying while I go for a test drive :-)

All the sensors seem to work OK (TPS, MAP, MAT, CTS), fuel pump is working and I have 40 ish lbs of fuel pressure. The ignition circuit is function with one anomaly, I have no voltage on the bypass signal. I doubt that's keeping the injectors from firing (but need to get that fixed too). REF signal is pulsing nicely at the ECU connector. I verified all the supply voltages and grounds. I swapped in my spare (stock) '749 with the same results.

What could be keeping the injectors from firing? Anyone have any thoughts?

Thanks in advance,

Bill
'84 928s 5 speed (FrankenPorsche)
http://mywebpages.comcast.net/p-928/749.html

 

If the $58 mask then no VATS, so no problem there.

When the ignition is turned on (no crank) does the SES light turn on, blink off then turn on solid? Is the ignition feed to the ECM on during crank? If the ECM controls the fuel pump does the relay activate during crank?

It may also be the injector wiring, is the ECM wired as SY/TY or as Sunbird? If Sunbird (for PnH) may need to reduce the I-sense resistors in the ECM.

Can disconnect all the injector connectors and place a small light bulb (1156/1157) in place of an injector. Then see of the bulb blinks durng crank.

RBob.

 

I need to hook up the SES tonight, but when I checked with a 'scope last night with the key on it was low (SES on). ECM controls the fuel pump, it's wired up like a Sunbird. Fuel pump runs for 2 seconds when the key goes on, but does not run while cranking. I've hooked a scope on the injector pin at the ECM to verify no pulses.

Can you elaborate on the I-sense resistors? I changed the injector drivers (on one of the 2 ECMs I have) but didn't make any other internal mods.

Thanks,

Bill
'84 928s 5 speed (FrankenPorsche)
http://mywebpages.comcast.net/p-928/749.html

 

Fuel pump: no run during crank may mean that the ECM is not seeing the proper DRP's. It may also mean that during crank power is being lost somewhere.

The ECM will re-enable the fuel pump relay once it sees DRP's. If the ECM is not seeing DRP's then no injector pulses either.

OK, Sunbird wiring is with PnH's in place. The reason I bring up the I-sense resistors is that by running 4 injectors in parallel the current requirements have doubled. This has the affect of halving the PnH currents allowed by the ECM.

Start with the fuel pump relay/DRP issue first.

RBob.

 

Arrrg.... lost ground tonight. Stephen knocked off a vacuum line while messing with the throttle linkage and a bunch of gas poured out. Traced it back to one of the regulators. Swapped it out with a spare and now we're holding pressure better. Got back to the injector problem, found that the fuel pump is in fact running while cranking. SES lights, blinks off, then lights solid when the key is turned on. No error codes reported through FreeScan. Still no start. Gave it another shot of ether and it wouldn't fire tonight. Checked spark and we no longer have spark. One step forward, two steps back.

Bill

 

FWIW, check to see if you have spark coming out of the coil, then out of the distributor. A dead rotor can play a heck of a game of peek a boo.

Be sure to have well done wiring on the power side of the injectors. With P+Hs they can draw some serious amperage, compared to a saturated set up.

Be systematic. ie first spark, then fuel.

Use a Noid light on both sides of the injector harness. Just to be sure it's just not one side that's fubar'd.

 

Hi Grumpy, haven't heard from you in a while. Good to hear you're still giving them h3ll! How's things?

I made a little progress here today. Found I have an intermittant somewhere in the ignition circuit. Thought I'd nailed it at one point when I wiggled the trigger wire on the coil and got spark, but then couldn't reproduce it. It's close by here somewhere, I'll get it yet.

When I have spark I also have the injectors firing. Got it to pop on a few cyls, that was good to hear. I'm thinking I need more fuel to get it going, the SyTy is set up with 27 lb injectors and the 928 is running 19 lbers. I'm going to need bigger injectors before I'm done, but now need to get it running, first track event of the season is next Friday and I'm going to take a SERIOUS razzing if I have to go in Stephen's 944!

As a point of reference, Stephen had his 944 in the garage while I was working on the 928 today. The 944 engine is 1/2 of the 928: 4 cyl, 2.5L, same Bosch controls, 19 lb injectors. The head even bolts onto the 928 (but you need two, of course!). The 944 was putting out 3 ms injector pulses while cranking, vs the 1 ms I'm seeing on the 749/BBZB.

I burned a bunch of eproms today but didn't manage to change much. Stock BBZB is giving me about 1.3 ms pulses while cranking. Moved the injector constant from 27 to 19 to 10 to 5 with no change (yep, just a decoration). Scaled the open and close VE vs RPM vs MAP table by 1.5 with no difference. Fattening up the VE table should ahve given me wider inj pulses, shouldn't it? I'm not quite sure how to get it, but think I need more gas.

 

Use the BPC vs EGR table. Only need the first entry for 0% EGR.

IIRC the $58 has a crank AFR table.

RBob.

 

Nothing I'm doing here is making any difference. Changing the BPW vs EGR or the cranking AFR table made no difference. It's almost like I'm in limp mode, but the fans aren't running. When I pull out the eprom and turn it on to force limp mode the fans kick on. Pulse width is the same, about 1.3 ms regardless of limp or any eprom changes I've made.

Would I be in limp mode without the fans running? What would put me in limp mode? I don't show malf any codes.

Bill

 

Well, it surely won't start with a PW of 1.3 msec, hmmm. Limp mode should have the PW rather high, 10-20 msec depending upon coolant. And since you can't use a scan tool while in limp mode I'd say you are scoping the PW.

Check the TPS value, maybe it is hooked up in reverse and the ECM is in clear flood mode. That will do it every time.

Another thing to check is the 100 KPA, lowest RPM row of the three VE tables. Those locations are also used during crank.

Also check the CTS sensor along with the others. Look for reasonable values. A scan tool is best for this.

RBob.

 

Yes, I'm using a scope and a scan tool. Scan tool says TPS is normally at 0% and rises to 80% when I floor it. Temp sensors read nominal, 16-20 degrees C. MAP reads just under 1 bar when I turn the key on. VE tables have standard SyTy values, around 43 kpa for the low rpm values, 30ish for the rpm adder. Everythink I can think of looking at seems normal, except the pulse width is always 1.3ms.

Oh yes, and the ign bypass signal is still wierd. The ignition module should be supplying a voltage to the ECU to say it's not in bypass mode, shouldn't it? Or is it the other way around, the ECU supplies the voltage, and the ignition module than just uses the reference pulses for the spark timing. Another issue, but I doubt it's causing my inj pulse problem.

Bill

 

The scan tool also reports a 1.3 msec PW?

The ECM pulls/allows the BYPASS (to go) low any time the 'engine not running' is true. The ignition module in turn pulls the EST signal from the ECM low. If the ECM doesn't see the EST low while it is holding/allowing the BYPASS low, then DTC 42 occurs and the EST logic is locked out. (There is also a pull down on the BYPASS line in the ignition module).

Once the ECM sees the engine RPM above a set value (typ 400 RPM) it will pull the BTPASS line high. The ignition module in turn will release the pull down on the EST signal. The ECM is then in control of the timing.

RBob.

 

FreeScan is not reporting injector pulse width. The info on bypass helps a lot, it is probably hooked up ok, just don't have it running yet so I can't see anything happening there.

I'm making some pretty good progress here. Grumpy helped me rule out a software problem yesterday (thanks again Grumpy!), so I started focusing on the hardware. Found the sense jumpers wrong on my test bench. Once I fixed that it started working properly. I can clearly see the fixed pulse and the variable hold portions of the injector cycle, and can vary the pulse width as expected. I think what I was seeing on the scope before was the initial high current pulse and no 'hold'. That would explain why I couldn't see any changes.

Out in the car I can now see the pulse portion drop nicely to ground, the hold portion of the cycle seems to have a problem. It's only dropping a volt or so, not enough to hold the injector open. When I pull 6 of the injectors off suddenly the 'hold' portion is dropping 2-3 volts and it starts to try to run on 2 cyls. Pretty obvious I have a drive problem, 2 injectors works but 8 doesn't.

You mentioned changing sense resistor values before, Is that what I'm looking at now?

Bill

 

At this rate, I'd say back way up and start from 0.

Pull the plugs, disable the spark, and see exactly what you have.

Plug the stock ecm, in, and see if you have a flashing node at the injectors, when you crank the engine. Triple check the SES operation.

Then install the plugs with the fuel disabled. And make sure you really do have a good bright blue spark. Then hook up a timing light, and see if the timing is close.

That should cover all the bases as far as just getting it to cough and burp.

 

I think I'm way beyond the basics. Timing's pretty close, I have spark, but not enough fuel. It wants to run when it has fuel, spray ether in and it fires right up for a sec or 2 .

With 2 injectors hooked up it fires real well on those 2 cyls, but just can't drive all 8 low Z injectors. That story is pretty clear from what I see on the o'scope. Peak part of the injector cycle goes way down near ground in both cases. The hold part of the cycle drops a couple of volts with 2 just injectors connected, and it pops on those 2 cyls (we got gas!). When I hook up all 8 injectors the peak part of the cycle is way down near ground, but the hold part is only a volt or so and it won't pop (no gas).

Bill

 

Yes, the PnH drivers are not set up for the 8 injectors. The stock Sunbird PnH setup has two injectors in parallel with each driver. This ends up being 1.2 ohms per driver. The driver peaks at 4 A and holds at 1 A (2/.5 per injector).

With 4 injectors per driver, what happens? The current requirement is doubled, but the driver is doing the same 4/1 A as before (Wired with only 4 injectors per driver?). Now instead of 2A peak/injector, there is only 1 A peak. Also check the injector resistance, they should be 2.4 ohms each. If lower, then will need to modifiy the next statement accordingly.

Open up the ECM, find the 2 current sense resistors and parallel them with another resistor of the same value (0.1 ohm, 1%, 2W).

RBob.

 

New resistors are in and it runs on the bench. We'll see tomorrow if it runs in the car. We might make Lime Rock on Friday after all :-)

Thanks again guys,

Bill

 

Bwahahahahaha...it's alive, IT'S ALIVE!!!

Bill
'84 928s 5 speed (FrankenPorsche)

 

Hey-hey, way to go. Congrat's.

RBob.

 

Kewl.

If you look at the RPM, TPS AE corrections, they are HUGE.
What can happen is as you get things close, you might see the top end going lean, after a given amount of time. What can happen is running on the AE shot so long that you think your tuning PE but your actually still getting fuel from AE.

You basically just need enough AE so that at any given point in time, you can get to PE.

I'd suggest raising the PE TPS enables a little, and then doing alot of work on the VE table, without using any AE or PE. Then work on the PE qualifiers, and PE AFR. and only then start in on the AE stuff.