driving vag coils, why not use a gate driver instead of bjt?

[weeblebiker]

So i've been doing research into some 4kv igbts and superfets at work and paralleling my ms coil on plug project. My work project is driving a pulse coil from around 10kv to around 300kv in a 1liter size canister I think with better resonance control I came drive a lower voltage primary with a igbt or superfet instead of the 9 kV spark gap and get the desired output
My impression is the vag coils are driven with a MOSFET, so why is the suggested circuit a PNP bjt running at 5v and not a gate driver at battery voltage for the gate driver vcc? Why not use a low side ultra fast MOSFET driver with inverted output like the ixys ixdi404 or a maxim integrated max4427 and skip the having to set the ms output to inverted? Ignition coils have better output the quicker the primary current is switched, so why are we not using a component designed to do just that?

[piledriver]

We look forward to boards for testing.

OTOH that would probably make for a better direct driver for the current BOSCH units, which are arguably underdriven by 5v..

Have any evidence the logic level driver circuit in use for the VAG etc coils are too slow and a faster circuit would help anything?

[weeblebiker]

I will test the vag coils after I get them from under the tree. Will build up a proto board with gate drivers and capaciter bypasses for vcc before then. Evidence is the higher than logic level current needed to drive the vag coil. All automotive grade ICs are required to work at 4.5v to work when cold cranking the starter motor, doesn't means they work well there . some of the new coil driver MOSFETs and igtb (and almost all "logic level" gate igbt and mosfets) have the gate driver built in like a Darlington. The cutting edge coil and capacitive load driver ICs has everything built into the t220 package (gate drive,over current, over temp, short, auto shut down and restart, with a diagnostic pin out AND the igbt or mosfet). I suspect the vag cops use regular old (2001) power MOSFETs. BTW on new cars with the latest design coils the obd2 will tell you which coil is faulting and what the fault is. the coil will have a few more pins.

[jsmcortina]

OTOH that would probably make for a better direct driver for the current BOSCH units, which are arguably underdriven by 5v..

I've tested them and they shutdown if driven over 6V or so. I found that dirving the BIP373 with the logic drive as we do now worked far better than what I was trying to do.

Have any evidence the logic level driver circuit in use for the VAG etc coils are too slow and a faster circuit would help anything?

I performed testing on the "low resistance" VAG COPs and found that my 5V NPN/PNP circuit was working well. This is not to say that a FET driver wouldn't work though. The discrete design was cheaper than comparable FET drivers I looked at.

James

[weeblebiker]

I performed testing on the "low resistance" VAG COPs and found that my 5V NPN/PNP circuit was working well. This is not to say that a FET driver wouldn't work though. The discrete design was cheaper than comparable FET drivers I looked at.

James[/quote]
The ones Im looking at are 2 channel ICs at under $4 ea. the ixys IC is the same one used to drive the v2.2 board injector fets. There are 2 channel gate drivers in the 25-35 ns delay range for around $1.50 ea.

your 5v/NPN/PNP circuit, is it a totem pole layout like the ic gate drivers or a series? the only other concern of using discrete components vs. ICs is variability and stacking of on/off delay times.

[jsmcortina]

your 5v/NPN/PNP circuit, is it a totem pole layout like the ic gate drivers or a series? the only other concern of using discrete components vs. ICs is variability and stacking of on/off delay times.

NPN with pullup driving an PNP with pulldown.

James

[weeblebiker]

We look forward to boards for testing.

OTOH that would probably make for a better direct driver for the current BOSCH units, which are arguably underdriven by 5v..

Have any evidence the logic level driver circuit in use for the VAG etc coils are too slow and a faster circuit would help anything?

The bP373 is a "triple stage darington" according to the datasheet, it already has a "gate drive" stage built in.

[weeblebiker]

since almost all mosfets will accept a gate voltage of +/-20v the easiest thing will be to raise the input voltage until the "igniter" shuts down / blows up below 15 or so volts
hmm, may be better if I can find a used one to try that on

James,
Do you happen to still have access to a newer low impedance vag coil and be able to carry out the experiment?

[weeblebiker]

I've tested them and they shutdown if driven over 6V or so.
James

I totally missed this before. looks like the vag cop mosfet/igbt are current driven more than voltage driven.

[Matt Cramer]

I totally missed this before. looks like the vag cop mosfet/igbt are current driven more than voltage driven.

I suspect they may be BJT darlingtons, like the BIP373s, and not an IGBT or MOSFET at all.

[weeblebiker]

yah the voltage and current isn't that great.
this is supposedly the ecu driver chip datasheet from the vw 1.8t ecu that was designed for the vag coils.
http://www.bosch-semiconductors.de/medi ... t_Info.pdf

it shows a 5v current source to the coil gate with a control signal going high and npn grounding the gate with the control signal going low. basically a gate driver.
edit additional:
and since it is a current signal, not a voltage signal it is more likely a bjt than a power mosfet