Coil-on-plug drive signal

[Paul Deslandes]

Anyone know what the drive signal should be to drive a coil-on-plug coil unit with built in amplifiers. I've a set off a Honda Civic R K20 engine, all four of which are suspect, and I want to test them off the engine. I assume its a ~3ms +ve pulse but don't know whether its a full 12Volts or something else.

 

Do they spark off the leading or trailing edge of the input pulse?

 

Ta.

 

Paul

 

 

[techbod]

I don't know how much this helps (I've not tried this with a COP setup yet) but I have done similar with a coilpack and external ignition amplifier.

 

Using a 555 in astable mode (with suitably chosen resistor values to give a reasonable range and on/off ratio), it directly drove the ignition amplifier, which in turn drove the coil pack and so fired the plugs. As I'm using ford parts, a 4ms pulse was reasonable and it was an inverted signal. If you hunt around on the megasquirt forum you maybe able to find out if the Honda is inverted/non - inverted.

 

(The only thing I can find so far is here, which implies that at least some civic engines are non-inverted.)

 

I originally assumed the amplifier required a 12v signal, but it turned out it would turn on with only a few volts, can't remember exactly how many.

 

I would expect to be able to fire a COP setup in a similar fashion

 

As a matter of interest, what engine are you trying to use them on?

[Paul Deslandes]

The engine is a Honda K20Z4 out of a 2007 Civic R that's on throttle bodies and installed into a 7. We're having real problems getting it to run properly and swapped the CoPs to a wasted spark yesterday and it ran like a dream. So its the CoPs or what's driving them. I've wired one up on the bench and dabbed a 1k resistor between the signal input and +12V and the plug sparks, so something's working. The coils allegedly have built in amplifiers but I'm suspicious that the drive signal is wrong in some way for these particular CoPs.

 

I've been rummaging around my boxes of bits for a triple 5 timer this afternoon to no avail so I think it needs a trip to Maplins for one tomorrow so I can built a more repeatable tester.

 

Info on Honda components is really hard to find.

 

 

[techbod]

've been rummaging around my boxes of bits for a triple 5 timer this afternoon to no avail so I think it needs a trip to Maplins for one tomorrow so I can built a more repeatable tester.

 

That's what I did in the end, too may problems with bread board, based version, so bought this one, left out the relay (contact rating to low for this application plus it was unhappy at the frequencies it was being driven at) and changed the VRs to give a better period.

 

Hunting round the web, most people seem to think the k20 COP has a built in amplifier, though I can't find a definitive answer. This thread has some conjecture on what voltage and which edge is used to fire, but again it just looks like someone's guess...

 

You're right, the Honda stuff looks hard to find

 

I'm guessing it that the ecu, engine and COP/ignition bits all came together?

 

Edited by - techbod on 18 Nov 2010 00:58:00

[Paul Deslandes]

The engine came with a Hondata ecu that was horrible so we now have a DTA which runs the engine on wasted spark a treat and is relatively simple to set up. There just seems to be an issue with the original COPs. It could be something odd with the loom but I know its connected up correctly, although it could be a funny earth problem or crosstalk, but its only audio frequencies for goodness sake! We know the ecu is seeing the correct pulses from the crank and cam sensors and there's no noise on them.

 

Having read that thread you kindly provided I'm wondering whether these COPs have built-in dwell which might be considerably retarding the spark. I really need to build this test rig and borrow a scope to see what they're doing. They might just all be faulty of course but initial tests suggest that they're okay.

 

The engine also came with a set of Honda injectors on which I haven't been able to find any data whatsoever. In the end we put one on an Asnu tester and measured its flow rate at 480cc/min. I suspect they're made by Denso for Honda and are a funny length being halfway between the short Picos and full length Bosch. Fortunately they fit the Jenvey TB injector hole and fuel rail so they're fine. The fuel rail just needed non-standard spacers.

 

Ho hum. Still, what else is there for a retired bloke to do in the winter?

 

Thanks for your input.

 

Paul

 

 

[techbod]

I'm sure you've spotted this so apologies for re stating it... I wonder if its a pullup issue?

 

Glancing through the manual for the s80 (picked one at random, I don't know which you have got) it says;

 

"External Coil Amplifiers

blah blah blah... Certain amplifiers need a pull up resistor between 12v and the ECU coil output. A 1K Ohm resistor is fitted internally in the ECU. Others have them fitted internally. If in doubt contact the coil manufacturer"

 

I wonder if you need an external pull up? From your original post and the fact you were using a 1k resistor I suspect you've tried this, but just in case.

 

You are right a scope will tell you what happening in no time at all. You are welcome to use mine but I imagine I'm a fair distance away...

 

 

Edited by - techbod on 18 Nov 2010 11:12:00

[Paul Deslandes]

No, I hadn't spotted that and it wasn't mentioned by Alan when I was at DTA the other day. Unfortunately he's on holiday for a week but I'll give it a go as its easy enough to do and won't do any harm if its not needed. The DTA (an S80Pro) uses the same output to drive wasted spark and COPs. In WS mode, presumably the ECU output tx switches to ground to energise the coil and then lets go after dwell time. In COP mode the signal is reversed and the o/p tx is 'on' and switches 'off' for the dwell time. If there's no pull-up the internal coil amp tx won't turn on properly. Thinking about it, there can't be a pull-up resistor in the COP otherwise the amp would be hard-on without the ECU connected and would burn up the coil, which it isn't and doesn't.

 

Thanks again - I'll keep you posted. I'm near Guildford in Surrey by the way.

 

Paul

 

 

[techbod]

Again, not sure if you've been down this route but in case...

 

Have you looked at the s80 wiring diagram? Page 7 has some specific k20a information, including mention of a k20 map and connecting the VTEC valve to aux 3?

 

In the west midlands so a long way for a scope!

 

[Paul Deslandes]

Yes I had seen that bit and its all connected accordingly. The inlet valve timing and VTec side is all working correctly, its just the ignition that isn't. I'm hoping to get out into the garage a bit later and try the pull-up resistors but SWMBO has got me doing something for her at the moment I have a feeling that this may be the answer but time will tell. I remain optimistic.

 

I used to have an old Telequipment scope but chucked it out years ago. I'm thinking of getting a Picoscope which works with a laptop and looks like a really useful piece of kit. Thanks for the offer of the loan of yours.

 

Paul

 

 

[Pooh_R]

I can highly recommend the picoscope. I've been using one for nearly 10 years. Pico's tech support is great, but I bought mine through autonerdz in the states to get their tech support free, which has proved invaluable.

 

Pooh

[Paul Deslandes]

Well, in case anyone's interested, the DTA does have internal 1k pull up resistors and fires as said earlier. I did try with some additional external 1k resistors but it didn't make any difference. I've also made a separate coil/ignition loom to isolate any possibility of interference from the main loom wiring.

 

I built an astable tester and have found that some of the coils are quite sensitive to supply voltage. Varying the voltage to the coil, but not the drive voltage, one will operate down to ~9.5V and others stop at somewhere between about 10 and 11.5V. This probably explains why the engine sometimes tries to fire when we stop cranking the starter; the battery voltage jumps up when the starter load is taken off. They're all relatively insensitive to drive voltage and will happily tick away with a +5V input pulse.

 

I really need to borrow a set of Civic R CoPs to try out, just to confirm that its the coils that are the problem. No idea why they should have failed like this though. Anyone near Guildford got a spare set I could borrow?

 

One last test to do before giving up is to reduce the plug gaps a bit to see if that reduces the voltage sensitivity.

 

I tried the conventional wasted spark coil again yesterday and it went first time.

 

Paul