With all the recent DIYPNP installs going on (and the fact it was as cold as Alaska here last night) I was looking at a way to send a +12V signal to the starting coile of the relay (therefore deleting the need to wire into the FP relay)... I had a PNP transistor leftover in the parts bowl fromt he MS build (can't remember where from, but anyways) using this circuit i was able to switch (on the bench) the gnd sending signal to a +12V. This proved working after the DB37 plug... I was wondering (for ease of installation until I move the whole shebang into a mazda case) would there be an intermediate position on the ms1 board where I can intercept the FP pin and send a +12V out of it instead of the gnd? Cheers Nuno

Attached Images

image3.gif (1.8 KB, 75 views)

heck you could just use the 12v power wire....

no wait... My MS sends a gnd signal to the FP relay grounding its coil...
Is it meant to switch to gnd? or does it send 12V?

Now I'm a little confused :(

I hope I didn't build it wrong... It's been going about a year!!!

On the '90-'93. the fuel pump relay does not come to the ECU per se., however you can backfeed +12 into the ST.SIGN wire at 1C (violet). This is how the MSPNP does it, and it's why you have to pull the ST.SIGN fuse.

On my car, I just used the stock MS FP driver (which is a closure to ground) and ran a wire across the dash to the light green wire of the Circuit Opening Relay- the same wire that the AFM normally closes to ground.

My only concern would be that the 2N3906 is only rated for 200ma collector current. I have no idea what the COR draws. Other than that, it looks like your circuit should work.

__________________

Thanks Joe, I have a wire under the dash at the moment as well.

My issue is that I may have to pass an inspection here locally (got a warning from the local police, I'm just hoping the car's not referred).

So for simplicity's sake, I wanted the stock harness (since it has the ability to do so) to switch the pump (that way I can mount the MS in a Mazda case). The 2N3906 is not the transistor i used, that is just a pic of the circuit (I should have clarified that)... I used a left over NPN transistor from the DIY kit, it is rated at 1000mA and switched a standard 30A relay fine on the bench.

What I am stuck on however, is intercepting the closure to ground signal on the MS board before it gets to the DB37 socket. I will take a closer look tonight maybe I can pull the flyback diode on the existing track...

Cheers
Nuno

You'd have to remove Q2 from the board and build your own circuit that wires into the three holes there.

Here's the stock circuit. The CPU puts +5 on "FP-1", through R16, to turn on Q2. R40/Q19 are just overcurrent safety.



I can't really see a way to flip this around using the stock holes, but I do have an altogether different idea. It involves an AQV212 photorelay. Here's an edit:




The IF of the AQV212 is min 1ma, max 3ma. The output pins on the MS1 CPU are rated for 2ma at VDD-.8, and a healthy 10ma at VDD-1.5. So as drawn, the CPU should be easily able to drive the AQV212 on, providing you with a nice, isolated 12 volt switch, capable of passing 800ma. I probably should have drawn a current limit on the output, but I'm feeling lazy right now.

__________________

Well silly me i just went and assembled it on the proto area without looking at the stock circuit (for the life of me i couldn't find it)...

What i will do is remove Q2 and relocate it to the proto area, leaving D4 as is acting as a protection diode.

Effectivelly having the circuit below (I already have the wires flying out near the Q2 position.

The transistor used was Q22 (from memory) used in the VR circuit which we do not use.

Hopefully it works as it did on the breadboard.

Attached Images

f_FPm_mod.GIF (8.8 KB, 50 views)