Ignition-only system

[rimcanyon]

I want to build an ignition - only control system (either coil-based or high voltage). This is for an older carbureted engine, pre-smog. The inputs would be tdc (hall sensor), vacuum and knock sensor. The distributor curve, vacuum advance and knock sensor retard would be settable.

Would it be possible to build this system using Megasquirt or its predecessor EFI332?

[Matt Cramer]

MS2 can definitely be run as ignition only; just don't connect the fuel outputs.

[trakkies]

I want to build an ignition - only control system (either coil-based or high voltage). This is for an older carbureted engine, pre-smog. The inputs would be tdc (hall sensor), vacuum and knock sensor. The distributor curve, vacuum advance and knock sensor retard would be settable.

Would it be possible to build this system using Megasquirt or its predecessor EFI332?

IIRC, you'll need to add a knock sensor module. MS can use the output from one of those, but doesn't include the electronics to read the actual sensors. Much the same as a wideband O2 sensor.

Are you keeping the distributor? You can use a trigger from a locked dizzy, use the MS internal coil driver and have fully mapped ignition. But you're left with all the intrinsic problems of a dizzy. Might be worth considering wasted spark and dispense with the dizzy.

[rimcanyon]

Matt & trakkies, that is what I was hoping to hear. This is for a 1964 Porsche 356, so keeping the stock distributor is part of the plan. Having a small footprint is also desirable. Best would be a package that would fit in the distributor. Next best would be something thin or small that could be attached to the firewall. Would microsquirt work for this application?

[trakkies]

You'd normally fit MS inside the car to keep it away from excessive heat.

I'd check the various Porsche forums. Chances are this has already been done.

[rimcanyon]

You'd normally fit MS inside the car to keep it away from excessive heat.

I'd check the various Porsche forums. Chances are this has already been done.

I posted a query about MS on the 356 Registry forum and did not get a single response from anyone familiar with MS.

[trakkies]

You'd normally fit MS inside the car to keep it away from excessive heat.

I'd check the various Porsche forums. Chances are this has already been done.

I posted a query about MS on the 356 Registry forum and did not get a single response from anyone familiar with MS.

I was more thinking of basics. MS can be used spark only - but many wouldn't think of it and go for an ignition only system. But the basics of locked dizzy, trigger etc are pretty well the same for all. MS is a good choice though - Tuner Studio is more user friendly software than some. And, of course, you could always go injection later.

[tvrgeek]

Today is the first day I decided to get serious and plan my MS system. I have been reading bits of the manual and several forums for awhile.

I too was thinking about starting ignition only on my 65 MG B. If for no other reason, I am more of a sparkie than wrench and the Weber is working OK, where my clone junk Lucas 45D has totally wrong advance curves. Step one would seem to be getting sensors in place. My first plan was to use a Davis DUI dizzy and as I will eventually use wet manifold TBI, I would not need a crank trigger. Alas, the DUI does not fit, so crank trigger coil-pack it is. I could lock down the Lucas clone, but the small cap limits the plug gap to .035 or so (as I proved on a Spit with MSD years ago). That's only 20KV or so before cross-fire. I am a fan of really serious spark energy.

The knock sensor seems to be the question. By basic calculations I found somewhere, my resonance is about 7K. But does that change with AFR? I would think the density would change, so the frequency would change. With a carb, the AFR changes a LOT and it is not predictable. A nice bandpass filter and a simple FET switch gate from the ignition with a timer doing reset on a flip-flop would seem to be a reasonable interface. It does not look like the "recommended" knock interface module is gated or does the ECU take care of that? ( only look after a spark) Any advice here? Is a sensor really worth it, or will my poly tube on a head bolt let me hear well enough I can derive and use fixed maps? I am looking for the 90% fix for 25% of the money/time/effort. I know better than to think I live in Perfect. When I nail it between 2K and 3K, the"tinkle" does not seem to change frequency, so maybe I am worried about a non-issue.

My precursor step was to install a Delco CS130 alternator, so I have reasonably clean negative ground power and enough of it.

Am I getting off on a decent path? I know I;m not a perfect confused person, as no one is perfect, but I am just starting down this path. I was a electronics lab rat in my past life, so that part is easy. I have played with little British cars for 30 years, so I know them pretty well too.

[feelingsupersonic]

Running Microsquirt on my 944 and love it. You can definitely do spark only with it, in fact you could do all 4 sequentially if you really wanted to.

Here's an idea though... what about using EDIS ignition, controlled by MS? All you would need to do is mount a 36-1 wheel. I've seen a lot of the VW Beetle guys do this conversion. In fact if you search youtube for "megasquirt beetle edis" you'll see a couple.

Basically you use a EDIS-4 coil pack which is small and can be mounted wherever, just connect spark plug leads like you did with the distributor. Then you connect it to the EDIS module. MS will talk to the EDIS module to control advance, and it sees the EDIS module as a distributor. This takes care of the crank signal. You'll need to get a MAP signal still.

[tvrgeek]

Yes, I have seen reference to this idea. I do like the idea of proven OEM parts. It looks like I need to bite the bullet and lift the engine to fit a 36-1 on the crank. I was trying to avoid that first pass and use a locked down dizzy, but I am a fan of high energy systems and the Lucas caps will cross-fire if you run a plug over about .035. Phasing gets difficult unless you have a 25D and can move the plate. That limits spark duration opportunities. I guess one of the big questions is Ford vs GM for generic parts. I really have no idea which is more versatile, easier to interface, robust, or cheaper. I am looking to put a good .045 plug gap in there and prove an old Taxi lump will actually fire almost every time. Fuel aside, the mechanical ignition leaves a lot of potential on the table.

First part came in. A Bosch knock sensor. Now to breadboard up a little filter board and get it mounted. Some have suggested the MG is too noisy to pull the signal out and that analog filters ( what I know pretty well) will never work. My DSP knowledge is theory only. I'm a systems architect, not a coder. I hope a little diff amp will be enough to feed my Focusrite A to D so I can use my acoustic stuff to acquire sensor signals.

[trakkies]

You'd not normally need to remove the engine to fit a trigger wheel. You'd remove the crank pulley and fit the wheel to that on the bench. As regards the sensor bracket, I'd ask on an MG forum to see if someone has already designed one. There is also a a company called Trigger Wheels in the UK who may be able to help with both.