Wasted Spark Question

[adamv61243]

Hi,

Forgive me if this was discussed earlier, but I haven't had any luck finding this exact situation using search. We currently run an `03 GSXR 600 with Wiseco 13.5CR pistons using the MicroSquirt (uS). The engine is a 4 cyl COP design and it is running relatively well using a 12-1 missing tooth wheel (as described in previous posts, we had a large number of issues trying to get it to run with the stock 8-0 crank + cam sync). We do have an issue where the uS doesn't always catch cranking VR pulses, but this may be a bad cranking tolerance on our end (we've checked the sensor signal and its spacing)... otherwise the engine runs great once it's started.

My concern is that we've been running the engine for some time with each IGN out driving two coils in series (not parallel) for wasted spark. This series coil configuration appears to be a hold over from back in 2008 when we used the Performance Electronics ECU, and is described in that ECU's manual on page 42: http://www.pe-ltd.com/Downloads%5CPE-EC ... ual_vG.pdf . I am aware that the uS wiring documentation clearly states that a given uS IGN output can only drive one coil and cannot handle the amperage from more coils in parallel... but the uS manual doesn't discuss the series configuration, which I believe (and please correct me if I'm wrong) only increases the dwell time needed and not the amperage. Still, there may not be enough time at the higher end (~12000rpm) to effectively charge the coils, correct? I am also aware that the uS manual states that more coils are only possible if connected to an external, logic-level ignitor after reconfiguring the uS jumpers for TTL out.

Does anyone know why this okay for the PE ECU and isn't for the uS? I realize it is a different ECU, but I am still curious because it seems to be working for us with the uS. While not a rigorous test, we've observed no discernible difference between the spark produced from either coil when we crank the engine and watch the plugs spark while holding them against the block for a ground. I should point out that the previous engine tuner empirically determined that a dwell of ~4ms works with this setup, and we have never broken a coil or a uS.

I do not have inductance specifications, but it may be of help to know that our 12V coils are ~1.25 ohms each, meeting Suzuki's ignition coil/plug cap resistance specifications of Primary Coil: 0.8 - 2.0 ohms (pos. tap to neg. tap) and Secondary Coil: 8 - 15K ohms (plug cap to neg. tap).

-Adam

[24c]

I have never seen this done before, but I looked at the PE pdf, and the first thing that surprised me was their ECU case
Back on topic, this is what you are doing,

PE COPs wiring

Screen shot 2010-09-04 at 06.27.16.png (24.96 KiB) Viewed 1753 times

and they only use this set up with COPs. Quite a neat idea, because the COPs resistance is usually significantly lower than a twin ended wasted spark coil, so unlike the parallel set up, you shouldn't overload the MicroSquirt built in ignition drivers. On my four cylinder motorcycle engines, I currently use a couple of twin ended inductive coils and a Bosch type igniter with COPs, basically as per MegaManual recommendations, but after seeing this post, I am going to reconfigure my test rig.

I will still use the Bosch type igniter, but pair two channels into a series COPs arrangement, and evaluate the spark quality back to back, and log the amperage. If it looks OK and within limits, I'll test it on a MicroSquirt. Interesting post

[24c]

I've had a quick look, and decided to measure the amps through the Microsquirt, as it's easier with the way stuff is wired on the rig.

Upto 1400 RPM using four R1 inductive COPs, I get 2.4-2.8A going through the MicroSquirt. I would expect this to increase as the RPMs rise, and also if I reduce it to two, I'd expect less. Believe it or not I do, at around 1.2A being drawn, but if I just connect all the coils via the igniter box and to a good battery, the four stick coils pull 8.48A maximum, and start to hum! I'll have to see if this limited by the ignitor box, or it is the maximum they pull.

Next I'll try series, as per your set up, and see what happens.

[adamv61243]

What bothers me is that the dwell needed to be 4 ms to get the engine to run properly with them in series. I don't remember if that was for the top end or not, but a quick calc says:

(12000 rpm)(1/60 hz/rpm)(1 normal spark / 2 crank rotations / cylinder)(4 cylinders)(2 wasted sparks / 1 normal spark) (1 spark / 2 possible IGN banks) = 467 hz

and 1/467 = 2.5 ms between wasted spark firings.

I don't think this is enough time to truly charge the coils in this configuration... but I don't have their inductance to calculate it.

[Matt Cramer]

Driving coils in series isn't ideal - they each have half the voltage, and you get double the flyback spike. The Microsquirt might be able to handle certain coils this way, but I'd prefer swapping out the coils or using an external module.

The resemblance is pretty uncanny, but there's only so many shapes you can use for an extruded case.