TBI conversion on a Ford 400 - need help getting it to run

[Dr_Grip]

Hi guys,
complete and total newbie here. I am trying to get a homebrew GM TBI setup running on my Ford 400 and it seems like I hit a wall by now. I really appreciate some help from you guys.

So, what has been done so far?
-I got a TBI system from a 1990 5.7 K-Series truck (7747 ECM).
-Mounted it on the engine's Edelbrock Performer intake
-A Ford DuraSpark II dizzy is connected to a brand-new HEI module (getting a perfect spark at the right time since putting the wires the right way 'round :D )
-Fuel pump's an Airtex E8094
-New coolant temp sensor, IAC, O2 and MAP sensor are fitted
-I did not fit a VSS
-Tested the TPS, it's good.
-I used TunerPro to set the BPW to 161.5 to match the displacement and disabled the EGR (based on the ARHT bin)


Sooo, what happens so far is that the engine starts up nice after a bit of cranking, idle's a bit high at 800-ish (but as I did not set base idle yet and the park/neutral switch isn't connected, that's not too much of a problem, I think).

When I try to build up revs slowly, the engine will hold any given speed (1500, 2000, 2500rpm, etc) only for a few seconds before losing revs (massively leaning out?). Furthermore, suddenly stepping on the throttle leads to violent backfires through the intake. The ECU is not throwing any codes. Any ideas what might cause this?

Once running, fuel pressure sits at .9 bar/13 psi, just as it should - but it takes about 10 seconds of cranking to build up fuel pressure at startup. Is that normal or does it indicate a problem with the fuel pressure regulator (most likely unrelated to issue above, but who knows)?

As I said, I am a total newbie when it comes to EFI, so I am at a loss here. Any help is very much appreciated.

[mmigacz]

Do you have the ability to datalog? Posting the log file, and bin would help to determin what is going on.

[Dr_Grip]

Do you have the ability to datalog? Posting the log file, and bin would help to determin what is going on.

I attached the bin file. I'll have to solder myself a log cable and get a logfile, I think I'll be able to do that during the course of the week (the workshop's about 20 minutes from my home so it's tough to get some real wrenching time in during the week).

[fastacton]

As far as fuel pressure goes, the regulator is normally pretty reliable provided the pump can keep up and you have a free flowing return line (more important with a high volume pump). You should get full pressure as soon as you turn the key as long as the pump is running (pressure bleeds off as soon as the pump stops). Have you watched the fuel pressure when you're actuating the throttle? It should stay right about 13 psi as long as the pump is capable. If it's dropping, then that could be your problem, but datalogs will help.

[Dr_Grip]

As far as fuel pressure goes, the regulator is normally pretty reliable provided the pump can keep up and you have a free flowing return line (more important with a high volume pump). You should get full pressure as soon as you turn the key as long as the pump is running (pressure bleeds off as soon as the pump stops).

As I said, it takes the pump a few seconds to get up to full pressure (the short pump activity burst when turning the ignition on won't do it). I got an in-line gauge installed and you can watch the needle creep up very... very... slooooowly. But...

Have you watched the fuel pressure when you're actuating the throttle? It should stay right about 13 psi as long as the pump is capable. If it's dropping, then that could be your problem, but datalogs will help.

We never noticed any drop once the pressure was where it should be. But I'll double-check and try to get some logs... It seems like soldering a serial cable is a piece of cake - now I just got to find an old laptop with a serial port :)

[fastacton]

If the pump can't get it up to pressure immediately, then that might be your problem. At least with the pumps I use, I'm used to an immediate jump to full pressure, and I'd be concerned that any pump that couldn't do that would be able to give good fuel pressure much past idle. What part of Germany are you in? I spent a few years there in the Ansbach and Schweinfurt areas.

[Dr_Grip]

If the pump can't get it up to pressure immediately, then that might be your problem. At least with the pumps I use, I'm used to an immediate jump to full pressure, and I'd be concerned that any pump that couldn't do that would be able to give good fuel pressure much past idle.

The Airtex pump is designed for the TBI systems, but maybe I got a bad one. I'll watch the pressure gauge closely and report back asap. How much pressure can the regulator take without getting damaged? I might as well think bigger in case I need a replacement pump...

What part of Germany are you in? I spent a few years there in the Ansbach and Schweinfurt areas.

I live in Berlin now, but I went to college in Würzburg, so I know the Schweinfurt area. A nice part of the country...

[Dr_Grip]

Reporting back - I fixed the fuel pressure problem. The car now runs mostly fine. There's only one and a half issue bugging me.

The half-issue is unstable idle in closed loop, I've seen tons of threads about this here so I guess I can tackle that on my own.

The biggie is the spark advance: I am trying to re-create my stock dizzy's advance map as a first step and the engine starts having ignition problems as soon as the advance crosses 30°. It's clearly rpm-independent: Wherever I set the advance map to cross 30°, the ignition starts having problems, I even blew my muffler with a violent backfire!
First, I thought I might have an arcing problem so I phased my dizzy correctly. Didn't change a thing. So I connected an engine tester and discovered something really weird:


Yep, somehow, each plug suddenly starts firing twice.

I then moved the P/N wires further away from the coil to rule out any interference from the discharge. Checked the signal with an oscilloscope, it's clean. The oscilloscope also tells me that the coil in fact is triggered two times per plug by the HEI module.
How's that possible? Bad HEI module? Bad ECU? I really need some pointers here...