General Purpose Low Level Ignition Drive

[Fred]

Ignition drive has been partially tested using two different chips and is showing MASSIVE promise, consider that done! Dan has tested by shorting to ground and didn't damage his chip. Em_knaps has been driving around with this running his engine. More testing is required, but I can't see there being any problems with this approach. The TL427 device even states something like "do xyz when driving resistive loads". Perfect.

[Fred]

I now personally have some of the MC33152V devices to test, though these aren't being used for ravage, they are being used, and I'd like to try them and see if they are up to the task or not.

[Fred]

I also now have both SMD and TH versions of the TC4427 part to abuse/test. I look forward to seeing your results so that I can duplicate them :-)

[TonyS]

Hey Fred,
I know there was a discussion about this a while back, but can you summarize why such a "beefy" part is needed?
Is there a list detailing what the source voltage and current requirements of the different ignition drives one might have to interface to?
I think this information may come in handy (Ravage - ignition drive voltage source jumper position).
Also, please note that given the 200 Ohm output resistor (at least on Ravage), drive current will be considerably less than the TC4427 is capable of.
Thanks,
Huff

[Fred]

Beefy, two reasons:

1) None of these spec a continuous output current. They only ever spec a peak current and a capacitor discharge time. They're all designed for FET driving, not resistive driving. We're not driving a FET with them, rather the base of a darlington, so our load is different.
2) The beefier they are, the lower their equivalent to RDS is, and the less heat they'll put out when loaded down.

The spec is 5 to 14.4V (nominal 12V) and up to 200mA, at least 100mA continuous.

I'm well aware of the current potential for the 200 Ohm resistors, and it's only barely enough with 5V but they can't be any lower value in 12V or the resistor wattage would be huge given a short condition.

No line driver chips seemed capable of doing this. The FET driver solution is in service on a vehicle and getting some bench testing too.

Fred.

[Fred]

I fired up one of these TC4427A chips tonight, and it's not nearly as beefy as I'd hoped, though still more than adequate for the job. I loaded it with an 8 ohm speaker from 5V and connected one side to power and the other side to the driver. It was noticeably muffled. In contrast my transistor 5V output with 10 ohm current limit resistors was MUCH louder despite the fact that the data sheet says 8 ohm on resistance. I'll have to run a scope over it at some point soon and try out another similar chip to compare. On the brighter side, it only got slightly warm driving a 50% duty 8 ohm load at 5V with no current limit resistor, so that's promising. I'll try some other configurations too, such as 12V.

[Fred]

A bit of circumstantial evidence to verify my experimentation:

Code: Select all

IC = 7  A IB =  70 mA
IC = 8  A IB = 100 mA
IC = 10 A IB = 250 mA

These are from a typical ignition darlington device and it lines up with my findings perfectly. i saw no benefit over about 150mA on the coils/ignitors that I tried. And it is darlington's which they use, with some protection stuff in some cases.

FWIW.

Fred.

[Fred]

Bump to myself and anyone else planning to test these!

[Peter]

I don't know if it makes any difference, but mine are designated as TC4427EPA in an 8 dip package. I have these driving some BIP373s with 220Ohm resistors on the gates. The coil I have shows an internal resistance of 0.7 Ohms. The ignitors and ecu are grounded with 12 gauge wire to the same place on the engine block. The mosfet drivers have to be running on the 12V/battery input, because the engine won't start if they're running on a 5V supply. Using the 5V supply gives me a nice square wave on my scope, but it never acts like it's even trying to start. So what I really don't get is how the MS people are getting away with driving these triple darlingtons strait from the mcu pins. :-/

[Dan]

something doesn't sound right here, let me think about this tonight and get back to you.