Fun day (not really). SO my friends truck with a zz4 and Edelbrock's MPFI conversion kit has been giving him major issues lately. I'm not going to bore anybody with details but here's something to be said about GM electronics; they're ROBUST!!! He's been way low on power and so today I helped him troublshoot the problems. Turns out the drivers side injectors weren't firing at all. I was able to grab the headers! Funny thing is that it still sounded like a v8. The number 1 cylinder injector was reporting 3.6 ohms, another on that side was 10. The rest were 16... I thought we fried the injector driver with this kind of load but it was fine . It's protected by shorts, I thought that was nice. We were going to through in standard injectors but they won't fit, so now we're contemplating going back to TBI. What a mess. Did anybody else know that your car will run on only 4 cylinders, I thought that was kind of odd. I would have expected some shaking but it was really smooth. Now here's a question for verybody; why did Edelbrock wire the injectors on both injector drivers? One channel is enough for 1.2 ohms of load and 4 of these pico injectors are 4 ohms. Any ideas? Do you think it's possible to edit a TBI ecm to fire each injector only once a revolution?

AFAIK the synchronous firings are triggered by the DRPs and the ecm just tells the hardware how long to fire the injectors for. I guess if you mess with the hardware you could but I sure as hell dont know anything about it really.

The reason they wired the injectors to both drivers is because of the peak to hold threshold of the driver IC. If you look at the data sheet for LM1949, it says that a voltage of 385mV across the sense resistor will trip the driver into low current hold mode. With a .1 Ohm sense resistor, which is typical in a TBI ECM, that means that at 3.85 Amps (current=volts/ohms) the driver will drop the current output to .94 Amps which you do not want with a saturated injector. By dividing the load between the two drivers (12V/4 Ohms=3A), they can be used to drive 4 saturated injectors but emulate a saturated driver since the driver never enters the low current hold mode because the current never reaches the trip threshold. As for making the ECM fire the injectors only once per revolution, that may depend on the ECM. I'm not sure but I am working on finding out if my PCM will do this. It is a 16197427 TBI truck PCM and has a CPI/PFI mode. I'll let you know when I get my PCM test bench built and test this. HTH

Makes sence. I had a feeling that might have been why but I didn't do the math.
Let me know about the mode in the pcm. I might be able to "borrow" the code. It would be neet using a c3 for port injection (looking forward to the Vortec Stealth Ram).

So, let me see if I'm understanding this correctly.

Are you saying if fires both banks of injectors at the same time, or alternating?

BTW, the Edel Pro Flow systems also use the Pico injectors.

They're firing 4 of the pico injectors like 1 TBI injector. So it's firing them twice a revolution unlike TPI batch fire which fires one a revolution. This means the pico's at idle are using ultra small pulse widths and are hard to control because of it. Then at higher rpms they're still firing synced so it's twice a revolution giving them hardly any time to close and re-open. I don't like the idea and neither does any auto manufacturer....ever! Trust me, it's a royal PITA. If I can figure out a way in source code of having the injectors fire only one a revolution (giving them more time to open and close) then we might be onto something. Oh yeah, and Edelbrock doesn't disable the async so it took even longer getting the transitioning correct. Obviously a big difference between having 2 HUGE injectors and 8 small ones.

Teh source coding part is relatively easy. Slap a two in front of the pulse width and make some other minor changes and it should be good to go. However, teh firings may be controlled entirely by the hardware like they are with our ecms. If tahts so, then the computer can only say how much fuel, not when.

I'm new to this, I've a '95 5.0 TBI (16197427) I was hoping to run something like the edelbrock TB or something similar on an intake with port injection bosses.

So what about running asynch all the time with larger injectors?

Can someone please point me to what specs I need the injectors to be? (if I were to run 8 instead of 2... saturated or P&H etc.)

I see the switch for CPI/PFI mode in GMPCM proggy, I have a running truck, and datamaster. Looking at the injector BPW vs. TPS, it looks like while off throttle or RPMs drop enough (I see MAP and RPM thresholds in GMPCM) the inj BPW kicks up a step, so this is when it enters asynch? (I'm supposing it's firing half the frequency so needs more duration?) Would there be any benefit for me to datalog it in CPI/PFI mode and see if the inj. BPW is any higher at the higher rpms? Or maybe we can wire some led's up to the injector harness and see if it's asynch or sync or is there even a reason to?

Someone set me straight! TIA

Mike

Or instead of redesigning the known universe, just plumb in a 749.

There's a recent thread about running a PCM as a tranny controller, that covers alot of what you want to know.

You can try it. But, you need to have the hardware, firmware, and software all on the same page. There's more in the thread I mentioned.

ha, wouldn't be the first time I tried to do something the hard way..

So a '749 computer will wire up relatively easily? (What years/models can that come from?)

I will do more searching so just ignore any redundant ?'s I'm asking. I must say the GM community has given me some great support, it seems like you guys respond immediately! (Experiences with moates, tunercat, gmpcm, TBIchips, some more I'm forgetting I'm sure)

Oh, and as for running the stock computer, some minor concerns were that I already spent the $$ on datamaster for the 7427 computer and I was thinking to keep the setup looking stock for emissions visual and in case they hook up some sort of scanner that could tell if it was the computer it was supposed to be?