VR input in parallel with coil?

[robs]

I'm installing a newly assembled MSII/V3 board to an '85 Peugeot 505. The car has Bosch LE2 Jetronic with a gradually failing AFM and a MS seemed a good project to get it running well again.

I am trying to be systematic moving to MS, starting with hooking it up as a data logger (more or less) before handing over responsibilities for injection and, finally, ignition. I started with an O2 sensor (which the car didn't have before) and that's hooked up and looks to be working fine.

Next was the ignition trigger. I want the stock ignition to continue to control spark for now but, because I'll be handing MS ignition control later, I hooked it up to the reluctor output at the Bosch 227 100 123 ignition module, which I left connected so it would trigger the spark as before. I used 2-conductor shielded cable, with +ve to pin 24, -ve to pin 2, shield floating, VR input jumpers.

If the MS is plugged in, the car won't start. If I start the car with the MS unplugged, then plug it in, it's fine. I get a good tach signal and one more gauge working in MegaTune. I guess plugging in MS has degraded the low revs signal that last little bit so it won't fire. Is what I'm trying to do here unreasonable? I'm no electronics wiz, but I thought MS would have very high impedance on its VR input and plugged or unplugged should make no difference.

A second question. If I use the pin-36 high current ignition output, is one pin really enough to sink the coil current?

Thanks for any suggestions.

[robs]

Well, rather than understand the problem I just changed plan.

I have gone with ignition first, but still letting the mechanicals determine advance (setting the entire advance table to 0). Car's running now with the Bosch module unplugged and MS making sparks. I can press on with everything else. Still planning to leave the ignition advance to last, but the MS makes a dandy ignition module.

[Mike_Robert]

One thing that would help with potential noise on the VR cable is that you should ground the shield at one end and one end only, preferably the MS end, and not leave it floating. That almost certainly had nothing to do with your original problem, thought I'd mention it. My guess would be that your hypothesis is correct; both devices in parallel loaded the VR output down a little too much.

-Mike

[robs]

Hello again. More ignition problems.

Having got it all running with VR tach input, direct coil output, and all squirting controlled by the MSII, the car has a pretty bad miss, especially at idle. The miss disappears if I switch back to the original ignition module doing the sparking (have to swap the wires while the engine is running; not #1 for convenience as a long-term fix).

Some possibly important information:

• I hooked up the original ignition module to a CRO and it runs roughly 4.8ms dwell (the miss remains if I bring the dwell up to this, or even higher). Perhaps the BIP373 doesn't expect to be hooked up to such a low performance coil.
• While running with the original module, but looking at the MS coil output on the CRO, there don't appear to be any misses, so the problem seems to be in making the spark, not the VR input or timing.
• I have a fairly heavy wire running from the coil -ve to the loom output from the MS, but the "PIN36" wire in the wiring bundle I got from DIYAutoTune was pretty thin. As I said earlier, I wonder if this is intended to sink a whole coil's current.
• I did apply a reasonable amount of heatsink compound when assembling the board

I can see two approaches. I can try a more modern coil, but I'm not sure -- this might make matters worse, or I can just use the MS output to control an ignition module. I'm leaning towards this second option. Perhaps I can even get it to control the original Bosch module.

I'd really appreciate it if there are any suggestions on other avenues to explore, or if you have any advice on better coils to try or on anything else to get it sparking cleanly.

[grippo]

Driving an external ignition module is the best way to go - it gets a lot of noise off the board and if it is the one on the car you know it will work. But first make sure there is no VR miss by verifying from a datalog that the trigger+/- output doesn't change and rpm never even momentarily jumps to 0.

[robs]

Thanks for this. I've checked several logs and the only time I see Trigger change value is occasionally during cranking, and there are no unexpected zero RPMs. I think the occasional trigger changes may be due to cross/back/misfires while cranking (killing cranking speed pretty quickly). They aren't good, but perhaps they stem from the same cause as the misses while running. Cure it and I hope it'll cure them.

How should I drive the Bosch 123 (not 124) module? It expects VR as input and triggers on the steep falling part of the wave. Feeding pin 36 (with a pull up), I'd need to trigger on the rising part of the resulting square wave. Would I get that by setting "Spark Output" to "going low" (or the opposite of whatever it is now) in Megatune? Anything else?

[robs]

Once again, following up to my own question.

I have done as I said, added a pull up for pin 36, and used this in place of the VR input to the Bosch 123 module. Started it up and it ran, but was ratty as get out at higher RPMs. Then it dawned on me that I had forgotten to invert Spark Output as per my previous post (module had got warm, but coil hadn't). Turned it off, fixed it in Megatune and it ran much better.

But it was still different from running on the module fed direct from the reluctor. It had about 5 degrees extra advance.

So, to the questions:

1. Why did it run at all with the Spark Output the wrong way up? Sparking at the beginning of the dwell period I guess (hence way advanced at higher revs)
2. More importantly, where did the extra advance come from when it was the right way up? Do you think this advance will be constant throughout the range, or will it increase (or decrease) at higher speeds?
3. Before I started cutting wires at the module, I thought I'd retry my setup as per my first question here for a fallback position. I added a 1K resistor in line on the MS VR input to see if that would leave a bit more signal for the module to trigger from. No dice. It would still only start if MS did the sparking. Should I have tried a bigger resistance, or was I barking up the wrong tree altogether?

The second question is the only really important one, but I'd like to understand as much as I can with all this. My working guess is that it's due to some built in delay in the Bosch module, but I've never done any of this stuff before, so would like some reassurance.

Other points that might be relevant:

• I still have vacuum and centrifugal advance in the distributor, with the entire MS advance table set to the initial advance.
• The max dwell is set to 4.8ms in line with what the module itself does (according to CRO).

Thanks for any advice.

Have fun,

Rob.