Logic spark wiring

[racingmini_mtl]

As I mentioned, using the TC4427 will work with any "logic-level" coil and even those using 12V. And a single DIP-8 chip will drive 2 channels. So you would need 2 chips and 4 current limiting resistors (if needed) for 4 spark outputs. That seems simple and cheap enough for a universal solution.

Jean

[jsmcortina]

As I mentioned, using the TC4427 will work with any "logic-level" coil and even those using 12V.

We did consider that for the MS3-Pro, but went with a discrete solution. With SMD assembly it was cheaper and should be just as effective. For DIY through-hole that could be a different matter. DIPs are dead easy to install for sure.

James

[prof315]

As I mentioned, using the TC4427 will work with any "logic-level" coil and even those using 12V. And a single DIP-8 chip will drive 2 channels. So you would need 2 chips and 4 current limiting resistors (if needed) for 4 spark outputs. That seems simple and cheap enough for a universal solution.

Jean

Works for me, if someone will whip up a schematic and component list for this, I'll be a test dummy with my next build.

[tutuur]

I've only just converted to wasted spark wired k20 cops and aside from some misfiring issues (probably down to oil in the plug holes) it works pretty good.

I'm by far not as clued up on electronics as you guys but i do like to know what exactly is the problem now. What is the reason and consequence of the coils firing occasionally with the going low setup?

[dontz125]

The problem is not that the coils fire "occasionally", but rather that they fire EVERY time you turn the key on. If there is any fresh mix in a cyl, particularly in one with slightly open valves, you get a loud and annoying BANG. With the 'going low' arrangement, the processor pin has to be HIGH to block the coil from charging; if the coils and chip are powered up at the same time, that leaves a window of a few milliseconds where the chip is unable to control the coils. How long does it take to charge the coils? A few milliseconds ... A situation where the coils are powered while the processor is disabled - corrupted firmware, damaged components - leaves the coils charging out of control. This will quickly damage the coils.

The solution is twofold; the first part is the 'going high' arrangement, where the pin goes high to PERMIT charging, rather than to block it. If the processor is disabled, the coils are inert, not charging. The second step is to power the coils via the fuel pump relay. As the relay is not activated until the chip is ready, the coils are not powered until the CPU is able to control them. In the event of a processor failure, the relay deactivates and the coils are left unpowered.

[jsmcortina]

Powering coils via the fuel pump relay is particularly important and is strongly recommended on all installs.

Another big reason for this discussion topic is standardisation - if the documented circuit is changed so that "going high" is the setting for logic coil outputs then it matches up with the setting used for high-current outputs with an internal BIP. That would mean most installs would then use the "going high" setting, therefore there would be less chance of new users getting the setting backwards and damaging things.

James

[dontz125]

From what I'm reading in the data sheet, no gate resistor is needed, and since I would presume the coil input pins are designed to handle the current of 5v/[input resistance], no output resistors should normally be needed either.

[schematic deleted - see revised version further down]

Very clean, very simple - at $1.64 each at Mouser, pretty cheap, too. Thoughts? Comments?

[racingmini_mtl]

It might still be a good idea to use something like a 10 Ohm resistor at the output just in case there is a short. That low value should not have an impact on the coil operation but it could just save traces and components. At worse, the resistor would act as a fuse.

On my board I do use a 10R resistor and a polyfuse to protect everything but that is likely not necessary for a DIY circuit. And it does work with the VAG coils.

Jean

[ol boy]

I like Jean's solution. I've used a Schmitt trigger as an output buffer for one project. Some Rc airplane ignition boxes are the going low type.

Thanks Ryan

[prof315]

I will be ordering some of those and trying them on my next build. I do a lot of MS2 builds for Nissan KAs and SRs with LS coils (and a fair number for Miatas with stock coils too!) this will make things way simpler.

[dontz125]

Here's the revised schematic incorporating Jean's comments. I made the limiting resistors 15Ohms, so that the voltage regulator wouldn't be redlined for however long it took for the particular resistor to blow.

Logic Ignition.jpg

Edit - Corrected ALED number; it is of course D16, not D1.

[prof315]

Awesome, I ordered parts this morning and the first test will be on a Honda D16 using Toyota COPs.

[racingmini_mtl]

I'm likely worrying for nothing but it would be good to have someone test the setup with VAG coils. I do know that it works with a 10R resistor and the voltage drop difference with a 15R will be small but I don't know how high the voltage requirements are for the different coils (for the VAG coils, if they do behave strickly as a 120 Ohm resistor as per Don's diagram then going from 10R to 15R would mean going from 4.6V to 4.4V no taking into account the voltage drop from the chip).

As for Toyota COPs, I think that also has been tested with my board so it should work here too.

Jean

[prof315]

I should be able to come up with some VAG coils for testing. I'll just scab the wiring for them in on the Honda. In fact I can test Toyota pencil, GM pencil, GM LS, Honda pencil and possibly one or 2 other logic coils as well.

[elaw]

It's been a while, but I recall a while back when I did some bench testing of my VAG coils (p/n 077 905 115Q) they were *very* voltage-sensitive. They'd start to behave strangely if I fed them any less than about 4.7 volts, which seemed pretty odd to me if they're really meant to run from a nominal +5 supply. It's almost as if they were designed to be driven from a higher voltage source, but I've never had access to an OE ECU to check.

[dontz125]

If the voltage drop becomes an issue, then perhaps the solution is a dedicated regulator. The LP2950 is rated for 100mA and is self-limiting to 160-220mA; no limiting resistors are needed.If this secondary regulator taps off the S12C pad, it can take advantage of the protection diodes and input cap of the main regulator. A single output cap is the only other component required.

This is slightly more complicated than the previous diagram, but not by much.

Logic Ignition - VAG.jpg

[prof315]

Parts just came in today I should be testing within a week

[jsmcortina]

Did anyone test yet?
I ordered parts too, but have been away.

James

[prof315]

Just about ready, I should be testing this weekend.

[wmax351]

What about on a microsquirt module? I'll be using a generic Bosch 0 227 100 211 4 channel driver, with 3 coils, for 3 cylinder wasted spark.