Bosch ign module on VN,VP V8 driving other ign box

[Circlotron]

Made a home brew ignition box for the 304 VP and for the time being I am triggering it from the output of the original Bosch box that comes as standard. I have a 100 ohm pullup resistor at the input inside my ign. Now the Bosch box would normally switch 5 or 6 amps or so but now it is only seeing the 100 ohm resistor i.e. 140mA. I imagine it would widen the dwell right out in an attempt to make the current what it used to be (impossible of course) and in doing so it is causing me some trouble.

What happens is that at low and medium underbonnet temperatures the thing idles just fine, but as it gets fully warmed up it will drop a cylinder once every 10 or 15 seconds or so. Only when it is in gear (auto) but when in neutral/park the idle is slightly higher and it doesn't miss a beat. Driving along is okay too. Behaves the same on petrol or gas.

Is it possible that the Bosch box is widening the dwell so much that it occasionally blots out the now narrow output pulse? Only seems to happen if it goes below 500 rpm and if the box is hot. It is analog after all and maybe a few RC time constants are drifting around a bit.

Bosch box by itself goes okay and so does my ignition on the bench. I'll see if I can find time to put a scope on it on the weekend. I expect any one that has put an MSD or similar on their Commodore and triggered it by the white "points" wire would have possibly had the same problem. They have a 47 ohm pullup resistor inside.

[VL400]

Once the ECU is running in EST mode the dwell is totally controlled by the ECU - that is whatever is sent over the EST line appears (inverted) on the output of the original ign module. With a standard calibration it wont drop out of EST mode unless it stalls. At idle you can get REALLY long dwell times so I doubt they are joining up but it can happen at high RPM with really fast accel.

I think you will need to get a scope on it and see what is happening when it misfires. Maybe get a hot air gun on your box when testing on the bench and see if it starts dropping events with long dwell times.

[Circlotron]

Put the [analog] scope on tonight, after the sun had gone down a bit.
The Bosch box output into my 100 ohm resistor at idle pulls low for a constant 5mS then a variable high time for the remainder of the cycle. Above maybe 2000 rpm they both shorten somewhat - I didn't look to closely at that. Anyhoo, it seems that the rising edge of the Bosch box that normally turns the coil off is not as clean as it could be. There =might= be a double pulse separated by about 100uS. Very hard to tell because I just couldn't get the scope to trigger on the falling edge. Actually, maybe the *falling* edge is dirty...

I could leave it idling as long as I liked in neutral and it would not miss a beat; only when in drive and then maybe once every 30 seconds or so. This was with the bonnet up and the Bosch module was not getting hot the slightest so the heat theory is out.

What is the changeover speed from fire-at-hall-effect-edge to fire-after-ECU-delay? Maybe I am going through the transition rpm and it is giving the waveform a blip?

[VL400]

The module to EST mode switch only happens once on a factory calibration, its after the RPM is over 450RPM. It does not drop out again unless there is a EST/bypass related fault.

Does the pulse look any different in neutral vs drive? Can you change the ECU calibration to make it idle lower in neutral to see if it is just RPM related?

[Circlotron]

Pulse looked the same in both neutral and drive. At least, not obviously different.
I =am= able change the calibration but I only have one memcal with the eprom soldered in. I do have an adapter I made so the memcal will go straight into the programmer zif socket. Have UV eraser too.

What I might do is block off the IAC hose and wind down the idle stop screw on the butterfly and see what happens.

[blownvp]

do you have a dual trace scope.if so compare input to your module to negative side of ign coil.be careful will be 100 to 300 volts positive when coil module switches off.if you only have single trace put on negative side of coil.factory ignition module has a high voltage npn transistor to switch coil like a bux80 or equivalent.could be your input circuit doesnt like the longer squarewave at the lower revs.factory one would also have a low value emitter resistor for current monitoring which shouldnt affect your setup because of the low current.the transistor should turn fully on.usually they just wind the base drive down if the coil current goes to high.if you dont mind me asking what component are you using to switch the coil.with the rising edge being dirty could be the zener diode or capacitor they have across the collector emitter of the npn transistor to limit the back emf spike when the coil switches off.are all your chassis connections for the 2 modules at a common point on the engine.ive played around with several different ignition setups over the years.hope you get it sorted.cheers darren.

[Circlotron]

G'day Darren, thanks for your interest.
I do have a dual trace cro but channel 1 carked it several years ago
My module will tolerate any duty cycle at the input - only looks at the rising edge of the wire from the Bosch module.
The boxes have widely separated earths but in this case it doesn't matter - my module's input is through a H11L1 schmitt trigger optocoupler so the input is completely floating. I =do= understand the significance of what you are saying here though.
I'm using an Infineon SPW47N60 600V 47A 70 milliohm mosfet to switch the coil. Has about 500mv drop at 7 amps.

Yesterday I got to work and left it idling i gear for several minutes and it didn't miss a beat. Coming home late last night and it acted up again.

What sort of ignitions have you played with?

[blownvp]

both cdi and tai.built both in kit form.modified a nissan sr20 distributor with built in coil and ignitor to work with an 808 used a mc3334p ignition ic for the interface.jaycar ignition module and timing kit.might be getting a positive spike or noise from something falsely triggering your input.shielded cable?filter on input?high enough gate source voltage to turn mosfet on properly?connect your scope to negative side of coil to see whats happening.

[Circlotron]

Gate-source voltage = battery voltage but limited by 15v zener.
Coil neg looks cool - 300V half sine ~20uS spike then 2.5mS arc duration. Takes 5mS to charge to 7 amps. 1:100 coil.

Yeah, come to think of it, all the wires going to and from the box are bundled together and the trigger wire is not shielded. I wouldn't have thought it was an issue noise-wise seeing it is feeding only 100 ohms inside the box but it is a wirewound resistor and they can get a bit inductive at high frequencies.

The box is kinda behind the glovebox and the wires go through the firewall via that grommet that is nearby, then across the firewall to the coil which is mounted on the front of the brake master cylinder. It just occurred to me that the wires pass behind the distributor and all it's noisy wires, so that might be something worth looking at too.

[VL400]

Are you using a driver for the gate of the MOSFET?

Can you just use two 200ohm 2W carbon films paralleled up instead of the wirewound? Its a long shot as the inductance would be pretty low in a wirewound, can easily check if it is causing any undesirable voltages by just probing the input to your circuit with the scope.