MoTeC setup

[scott.mohekey]

Does any body know how we would set up the REF and SYNC of a MoTeC 100 series ECU?

There are three application notes on their website in regards to Mitsubishis, but none of them mention the v6 VR4.

http://www.motec.com.au/filedownload.php/t19.pdf?docid=1288
http://www.motec.com.au/filedownload.php/t17.pdf?docid=1286
http://www.motec.com.au/filedownload.php/t13.pdf?docid=1282

And no, this doesn't mean I have one. I'm just doing all my research up front.

[Mitsiman]

The V6 on the VR4 is a very different setup with a 3 pulse on the crank, and four varying pulses on the camsahft wheel. To my knowledge it has not been logged by anyone to date.

We have successfully used the Haltech range of ECU, and make a plug and play system for the manuel model vehicles with a baseline tune as well.

But you hve to modify the cam sensor from four down to 1 signal. Very easy to do takes around 3/4 hour by removing rocker cover and plastic cover and cutting with tin snips from behind.

[scott.mohekey]

When you say varying pulses, do you mean that each of the four pulses is of a different length? Do you know if they're timed on the rise or the fall of the pulse?

[Turbo_Steve]

I seem to recall the signals are identical to those of the FTO? Certainly the crank is (and I'm fairly sure you don't need a cam signal for the Motec unless you want to).

[jinglis]

there is a guy running a motec m400 on a fto powered colt? maybe get onthe colt forums and ask as the fto and vr4 engines are very similar. i'll try and remember who it was

[jinglis]

k the guy on the colt forums is called tintin(martin) and he runs a m800 and its was mapped by ben at eurospec. hope that helps

[Turbo_Steve]

Lol! "beeeeeennnnnn!!!" :d

[scott.mohekey]

Yeah.. where he at!?

[Goku]

I'm guessing the UK.

Try the guys down at NZEFI Scott, they should be able to help you somewhat.

[Eurospec]

I'm here! In the UK!

Right- Haltechs work if you modify the cam angle sensor as mitsu man mentioned.

Martin (tin tin) runs his colt off the crank angle only (thus injection is batch, not sequential- not that you would notice.) He has an M800. On motec you dont have to run cam angle as steve says.

Personally for a vr4 i would go haltech or vipec. Either way there is no manufacturer produced pnp, you can make it pnp by building a patch loom though. Our 6A13 FTO will run haltech if i ever ever finish it!

Cheers,

Ben.

[Turbo_Steve]

I didn't know nobody did a PnP setup: thought the FTO had it covered :(

That said, if you're stuck on Motec, you could look at:
a) using an Evo timing wheel, though you'd have to work out the timings. This isn't that hard, though, really. And, of course, you'd have to make the thing fit, and potentially make a crank sensor change (maybe).


b) Probably easier simply to make your own timing wheel. Motec supports a number of "regular" timing wheels, as well as "3-2-1" wheels and "missing tooth" wheels.
The only drawback I've managed to find with "missing tooth" is that the engine has to turn a full 360degress before the ECU will even try firing it.
This is bad for a road car, in that you lose that sensation of "touch the key and it fires" and you have to sit there cranking like a diesel for maybe 3 secs.
This does have the advantage that your oil pressure is that much higher when it fires.

I've only ever used the Haltec EMS (never the Miniceptor or any of that) and it really is very good. It's not quite as much of a doddle to look after as the Motec, and definitely doesn't have as many toys that can be plugged in, as many logging options etc, but it does all the essential stuff the Motec does, and should be able to run a fair few accessories.

I also rate the AEM EMS, which has some pretty good software (although it seems to work in some weird units / use weird graphs sometimes). It's really pretty good, though, and again allows 2D control over a number of accessories.

I'd say the Haltech and AEM are pretty good competition for each other, depending on the price.
I mildly prefer AEMs software. Neither of them will do cool software dashboards like the Motec stuff LOL.

[Eurospec]

Thats a very good point about the crank- you do get the delay with a lot of stuff as it looks for the sync that isnt there, shrugs its shoulders and fires it anyway!

Steve- you tried a vipec mate? They are awesome value for the pnp apps. Not mapped one myself yet, but i've seen a few now and they are NICE. Very autronicy feeling as you might expect.

Cheers,

Ben.

[Turbo_Steve]

Ohh..no!? Hadn't even heard of a Vipec.....it's off to google I go, I go........thanks Ben!

[scott.mohekey]

Ok, I need some clarification here I think.

The MoTec is talking about REF and SYNC, and you guys have mentioned that with a missing tooth crank angle sensor you don't have sync, does this map to the sync that motec is talking about?

What are REF and SYNC exactly?

One of the things I really like about the MoTeC is its full support of the CAN protocol.

[Turbo_Steve]

Effectively, REF is Cam, Sync is Crank.
Usually, REF triggers injectors, Sync triggers ignition.
REF can be generated from SYNC by using a simple timing differential (i.e. you can tell it to add a couple of ms to the SYNC signal and use it as REF for the injectors).

The missing tooth wheel (i.e. making your own crank timing wheel) just replaces the factory wheel, and then you simply program the Motec to use x number of teeth per cycle, with tooth number x + 1 being at the start or end or TDC or whatever you want, really.

Snice :D

As ben says, not having a REF signal means some ECUs (and the motec is sometimes one of them...can depend how many teeth / missing teeth you use) means the ECU will need to have a bit of a think when cranking. If you use a (say) 13 tooth wheel, with one missing at TDC, it can need to the turn the engine a couple of times before it decides it's sure where it is in the cycle, and what speed it wants to fire at.

What you can't do is use a 4-tooth wheel on a 4 cylinder engine, or a 6 tooth wheel on a 6 cylinder engine, and expect it to run without a REF signal. It'll probably run...but it can be a bit ropey on the timings, as there is always a delay between the crank state change and the amount of time it takes to assemble the trigger pulse (or, more correctly, the trigger discharge) for injectors and coilpacks etc. You generally end up batch firing the injectors (and possibly ignition) and, which can have costs in terms of dwell times and wasting fuel under normal driving conditions. It'd be fine on a dragster or track car, but I'd imagine it would compromise the driveability of a road car.

Generally a 13-tooth wheel seems to suffice on a 4-cylinder engine, and I have no reason to believe it wouldn't be perfectly useable on a 6-cylinder lump too.

That being said, (and I haven't done this myself) I believe it is possible to write a "timing script" for the motec, describing the signal it should expect and how that corresponds to engine cycle. This is easier for a cam sensor than a crank sensor, as cam sensors generally produce a regular waveform, whereas manufacturers seem to do all sorts of cleverness with crank wheels, to let the factory ECU fire immediately.

The best way to do it is slap an oscilloscope onto the crank and then cam sensors and check out the waveform. If it's something fairly simple then it's likely the Motec will swallow it with a minimum of fuss....even if the actual crank wheel is fairly complicated, the signal produced at 950rpms tends to be regular enough for the ECU to be told it's an 8-tooth or 10 tooth wheel or similar. This can also be a pig to crank, as at low speeds the ECU isn't seeing anything it recognises (Older Link ECUs seem especially susceptible to poor crank signal recognition at startup, taking perceptibly longer to start).

Er...does any of that help?

[scott.mohekey]

Yeah, that helps a lot. Seems I had ref and sync mixed up.

I've just had a look in ASA to confirm what my memory was telling me, and that is that the crank angle sensor on the vr4 uses a 3 tooth wheel and the cam angle sensor uses a 4 tooth wheel. Does this sound right? I can't figure out how the ECU can determine TDC from this, unless it samples the correlation over a few test cranks?

[scott.mohekey]

Thinking about it a bit more, the cam sprocket with the cam angle sensor on it will turn one whole revolution in the same time the crank angle sensor will turn two whole revolutions (I think). This means there'll be two teeth lined up for each sensor for every two crank revolutions, in other words one pair of teeth lined up per crank revolution. I assume this would be TDC?

[Turbo_Steve]

Actually...now that you mention it, I think I may have got SYNC and REF mixed up....urgh: has it been that long? I will have to go and check now.

And yes, what you describe between cam and crank sounds about right: I haven't even looked at the VR4 setup (just assumed it had been covered for FTO) so I can't say yes or no, but what you describe sounds a reasonable enough...however:

I haven't see the VR4 crank wheel, but if it's 4 teeth, I'd expect them all to be different sizes, so that the gaps are not equally spaced around the wheel. As the ECU is likely to be triggered by the "falling edge" (i.e. as a tooth leaves the sensor) this effectively tells in in one revolution what position the crank is in: chances are there will be two gaps quite close together on the wheel. ECU should see this as the 'start' or TDC or whatever, and, as you suggest, use the cam sensor (with it's many more teeth) to determine firing times.

Just a quick correction on what I was saying earlier about crank sensor wheels: I was thinking on it last night, and I don't think you CAN use a 13 tooth wheel on it's own...I was remembering a setup we did, but as I recall it DID have a cam sensor. If you're going to the hassle of custom making a wheel, I would suggest at least 30 teeth if you're not using cam, 60 generally seems very popular, though can be trickier to machine.

Question: Why are you so keen on CAN bus? Don't get me wrong, it's great to have everything connected together on one cable, but unless you're having multiple devices (e.g. Diff Control, Gearbox management, ECU and datalogger) then the benefits are somewhat limited? You can use just about any brand of funky dashboard, I suppose, and someone elses datalogger, but inevitably it's easier to implement if you stay with a single brand...just like everything else, I suppose :D Although they're all using CAN, there are varied interpretations of the actual data. Everything will communicate, but you'll sometimes wonder at the data different devices collect.

[scott.mohekey]

The other quite good thing about CAN is that you can get Atmel AVR Microcontrollers with a CAN controller built in, so I can make devices that communicate with the ECU myself.

I'm pretty sure the crank angle sensor is three equal sized teeth, while the cam angle sensor is four equal sized teeth. How many revolutions of the crank does it take for all 6 pistons to go through one complete four stroke sequence?

[Nutter_John]

720 degrees makes one complete cycle for a v6

[scott.mohekey]

Two revolutions, which is what I was expecting. I think each tooth of the crank angle sensor lines up to top dead center of one of the cylinders. Now, how does the ecu figure out which cylinder is cylinder one? is it a combination of crank and cam angle sensor? (two in every four cam teeth lines up to one in every three crank teeth)

[scott.mohekey]

This explains the GTO's setup, which is the same as ours I believe.

http://www.stealth316.com/2-ignitionsystem.htm