Important question (for me):

How much additional current can I sink off the EPROM Vcc supply? I was thinking like 100mA maybe at the most, just wanted to check with those knowledgable on the Delco systems to see how much excess might be available there. And about the ground pin at the EPROM, is it tied common?

This question refers mainly to the 1227165/730 systems. It requires a knowledge of the relative power available via on-board regulator compared to that consumed.

Any help from a hard-core guru is much appreciated.

Many thanks,
-Craig

 

There was real thread on this on one of the DIY-EFI lists. Kinda brutal actually.

In all fairness, I just wouldn't.
GM spent a ton of money getting things right. I wouldn't get in there and expect any success LONG TERM. I've over done things on the bench to distruction, and it seems that long term is the problem.

 

I don't know what your plan is, but I think Grumpy is right about not overtaxing the Vcc supply. Whatever you were planning to draw from that supply, you could easily use an external linear regulator(LM7805), from any +12v source with some additional capacitors for filtering, and supply almost anything you want. The regulator can handle almost an amp(probably more with really good heatsinking), and if you are already adding some additional hardware, it's basically only soldering a couple of extra pins(its a 3 pin device). You can find datasheets on this and other regulators that can handle whatever power you need, from national semiconductor, Texas Instruments, Linear Tech, etc, stick with linear regulators, they are much simpler than switching ones, they are less efficient, but with an alternator as a power source, a couple extra watts won't mean a thing. I can give you more details, web links if you want.

 

Yeah, I was keeping a 78L05 in the hip pocket just in case. Just would have rather had no additional power supply connection needed.

All I'm hoping to run off the Vcc is like a 60mA quad S/R latch. And maybe a parallel array of five 47k pull-downs.

Maybe I'll go the safe route with an external supply and make an option available for internal with associated risk...

 

I've seen some 7805 stuff where they were too worried about spikes and actually got to generate a ringing on the supply side. Just a 1Mfd. and a .1 is more then enough, be sure to use a reverse biased diode, and then a blocking diode (so it can't be hooked up wrong).

Just a FWIW.