General Purpose Low Level Ignition Drive

[Fred]

I was talking about the requirements for a good ignition drive solution with slacker.cam a few days ago, and he suggested fet drivers. Perfect. They are push pull, will pump out any voltage that you feed into them within reason and push sufficient current to go with it. One of these per channel with a switchable voltage input (5 or 12), a leading XOR gate, and a current limit resistor would give us total output flexibility for driving all sorts of ignitors and other external low current ignition devices with total reliability and a nice strong signal. Hooray.

The trick will be finding something common, cheap, package compatible with other manufacturers and not nearing end of life.

As a start, this looks promising:

Microchip TC4426

dual channel (require 3 for 6 channels)
upto 18v supply (sufficient)
more than sufficient current capability
small smd package for good board size
a dozen different suppliers
~1usd each

No idea if it is pin compatible with the other manufacturers, but I'll check later if no one else does.

If anyone can provide a more cost effective solution with similar benefits in discretes or some other type of integrated circuit, blow me away, please.

Fred.

[Fred]

My search:

http://octopart.com/partsearch/#search/ ... Bmax%5D=13

[jharvey]

I like the concept, and I'd bet a good push pull can be found. I'm concerned with the ones found in the above noted Octopart search. They don't appear friendly to driving inductive loads. Look at the reference schematics from this data sheet.

http://datasheet.octopart.com/MC34152DG ... 159693.pdf

Your build will typically have a known requirement of either top side drive, or low side drive, so I'd be tempted to encourage one to simply install the chip they need. I seem to recall top side drive OVP that was pin compatible with most MOSFET's. Hmmmm, Perhaps something like this could work.
http://www.st.com/internet/com/TECHNICA ... 151795.pdf

H bridge drivers are known for High and Low side drive and can typically handle inductive loads. I think this would be similar to the above noted devices, but more robust. To bad that one don't have good availability. I'd bet there is a good H bridge driver that is available.

[Fred]

Good info. It won't be driving inductive loads, though. Just ignitors, which are over sized darlingtons (or similar) and require quite a bit of input current to function optimally.

That driver you linked looks neat, but is MASSIVE overkill current wise (we need 50 - 100mA) and quite specialised (plus, no one has stock, or so it seems).

It *might* be nice to use the same driver to feed optional on board igbt or darlington style ignition drivers such as the bosch bip373, NGP15N41CLG, VB525SP etc, and just not recommend it.

Fred.

[jharvey]

Ah, it appears that chip was obsolete and NRND. I found a list of other potentail H bridge drivers at this link.

http://www.st.com/internet/com/SALES_AN ... pl0608.pdf

Which brought me to this one, uses the OVP found in many of ST's devices.

http://datasheet.octopart.com/VN771KTR- ... 163368.pdf

Lots of vendors, but they don't appear to have stock. Looks like ST has several Smart H bridge chips to choose from, just need to find one that's available in low qty, and not as a sample.

[Fred]

Most of the H bridge stuff (or at least some) is driven by an enable pin, and a polarity pin, not individual drives. So, you'd need to find one that doesn't work like that, too.

Fred.

[Dan]

Thought I might comment on this a little, (about time I started)

The Microchip TC42X series look like a good choice IMO.

I have found there are several options that are pin compatible which will give us inverting, non-inverting and, if needed, complimentary outputs, and have found them to be readily available where I am located.

They are dual channel and have a peak output current rated at 1.5A (which should be ample).

Datasheet here - http://www.farnell.com/datasheets/22020.pdf

The TC426 (inverting) and TC427 (non-inverting) average to be around $1.00AUD from my normal suppliers.

We could also use the TC428 (has one inverting and one non-inverting channel) and tie the inputs together, giving the end-user more flexibility in igniter choice? or simply state that they are to fit the component that best suits their requirements?

I can grab some for testing purposes if those whom are interested think it is a good idea?

EDIT - added more info.

[jharvey]

Cool chip. I like the high side and low side drive potential. I'm a bit concerned about the potential 10 to 20 ohm impedance. It notes the output could be a -, but typical appears to be some where around 6 to 13 ohms. I believe that should be fine if we drive an external lignitor. However you know someone will want to drive directly for some reason, and that many ohms will likely cause a LRC issues. Any chance there is one that can drive the external directly? How does it do about ESD? If driving an external device, it's likely going to get whacked by stray voltages, and will need to be fairly tough.

[Dan]

I agree with you regarding impedance, but are we catering for people wanting to drive directly?

If we are, I thought fred mentioned that we could drive on-board ignition drivers optionally with these types of devices? In which case, these chips would suffice wouldn't they? (am I missing something silly here? )

I also just re-read the datasheet and I cant find anything with regards to how it is ESD protected. It states that it is protected to 2kV, but doesnt say how it accomplishes this.

Do we think we could use these to drive external ignitors and provide a spot on the PCB for "optional" ignition driver IGBT's/darlington's as well to cater for those who want to drive direct so to speak?

EDIT: FWIW I like these "NGP15N41CLG".

[Fred]

I agree with you regarding impedance, but are we catering for people wanting to drive directly?

That depends on the particular PCB implementation. I believe that ignition should be run with external ignitors, so I'm likely to either not include IGBT/Darlington positions, or to include them as an experimental afterthought. The only thing that is clear is that 5v drive is not sufficient for many cases, ignitor wise, and that a decent amount of current is required to saturate them. (70 - 150mA)

If we are, I thought fred mentioned that we could drive on-board ignition drivers optionally with these types of devices? In which case, these chips would suffice wouldn't they? (am I missing something silly here? )

I believe you could, yes, and if fed with 5v instead of 12v they would provide a much stronger switch on current than a cpu pin ever could. Given the 12v option, they're ideal for driving IGBTs which often require a higher than 5v input. Correct me, ruthlessly if need be, if I'm wrong.

Do we think we could use these to drive external ignitors and provide a spot on the PCB for "optional" ignition driver IGBT's/darlington's as well to cater for those who want to drive direct so to speak?

Yes, absolutely, if the routing is done well, AND the user wires the box up well, I expect that it will function adequately.

About the chip in your previous post, the inverted/not chip selection is fine for diy stuff, but not suitable to manufacturing the device and telling someone "change this". The dual polarity one is an option, but adds complexity. With that you could bring both pins out, but that's 6 extra pins which really aren't necessary. You just gave me an idea, though, and it doesn't really belong in this thread, but..

We could take the feed to all of these drivers out of the connector and take a 5v and a sanctioned 12v out of the connector near by, then tell the user "connect this wire to this wire for 5v drive and this wire to this wire for 12v drive". Which makes the PCB be a purely manufacturable thing.

Which also gave me another related idea. We could do the same thing for the XOR polarity. Take one wire out that controls the XOR chips and take out a 5v and a ground dedicated to it and make the user configure it. In this case you could have a weak pull to whatever was default (ground or 5v) and have them short it to the other if required so the default "just works" for most with no extra config. (which would no be possible for the 5v/12v thing)

Reality is, most people will want to run ignitors, most of those will be happy with 12v and default ignition polarity, and thus the whole configurability thing is a bit of a red herring anyway. Probably what I'll do is use a zero ohm link to select the defaults at assembly time, and make it easy to rip that off and bridge two near by pads together.

Fred.