Adamw

Do the aux 3 & 3 statuses show as inactive in the software when you are measuring them at 0V? Is there any device/load connected when you are measuring? Is it a G4, G4+ or G4X atom?

The triggers like fine. The "oddness" of the waveform is only because the crank rpm increases suddenly by 2 or 300% when the one cyl fires, then it drops back to slow for the other 7. As far as the ecu is concerned it looks quite acceptable.

Yes that one is photographed upside down compared to how I normally see them for whatever reason, but that is the coil im talking about. But I would go for a name brand like Bosch, NGK, Beru, Vemu or similar, that one is not a brand I recognise and I have seen some pretty poor quality junk from many of those generic black label brands. Also, be careful that the one you go for does actually show the Aluminium heatsink part included (this is the ignitor), I have seen some companies selling the same coil with same part number but it is missing the ignitor. I dont know any of these companies, but just for reference, here are a couple of Bosch ones for about the same money: https://www.autodoc.co.uk/bosch/757956 https://www.sparepartstore24.co.uk/757956-bosch

Whilst I cant be 100% sure as we used to buy the probes with the fitting as part of the kit, Im 99% sure this is just a standard swagelok compression fitting that is quite common in the industrial world so you should be able to find locally. I reckon the complete fitting would be part number SS-400-1-2. If you are just after the nut and ferrule then have a look for SS-400-NFSET

Yes it is a dumb coil. Im pretty sure most of the DTA's - at least the older ones I have used - had an internal coil driver(s) that could be turned on via software, so it can drive a dumb coil directly. My ign coil catalog shows a primary resistance of 0.7ohm, whereas it would state "NA" or "--" for a smart coil as primary resistance cant be measured. The common VW golf one might be a good option to swap in rather than add an ignitor and change wiring. They are a 4 post wasted spark coil with internal ignitor, good powerful spark and cheap as chips. I think they have the same 4mm terminal so leads will probably fit too. A common part number is 032905106b

The G4+ pinout used to only cover pins that were used and configurable, the G4X pinout is more complete and has all connections shown. So you can look in the G4X help file in PC link or the PDF manual that is on the ecu's webpage. But as a quick reply, you will find most of the pins in this list are "NC" which means not connected. Our ecu doesnt support the autotrans and consult port so they are not needed, 115 is the ground for the O2 heater, 16 is the main ECCS relay control.

Go to your CAN setup, change the Link CAN Lambda channel to "Receive User Stream **", and change ID to 688. So it will look something like below. Then go to the Streams tab, click on the stream number that you have chosen above, then click load stream, load the .lcs file attached to this post. It should look something like below with the stream file loaded. Apply those changes, do a store and cycle ign off/on then have a look if lambda 1 & 2 are reporting anything. If not have a look at CAN AN10 to see if the diagnostic code gives any clues. I have a vague feeling they may need an RPM signal from the ecu to start up, but dont have any info on that. Haltech WB2A Use ID 688.lcs

Its only firing on one cylinder. So could be plug leads in wrong order, stuck injectors, distributor out of phase etc. I would first try a squirt of starter fluid or petrol into the manifold to see if that changes anything - if more cylinders come alive with that then you can look more at stuck injectors or fuel plumbing. If adding fuel doesnt help then confirm the rotors are pointing at the correct post and each lead is running to the correct cylinder. I dont have much experience with the twin dizzy 1U, so im not sure if the wrong ignitor wired to the wrong ign drive could cause anything weird like that, but I would also confirm the ign 1 output is connected to the ignitor that controls the coil on the bank that has cyl #1 Im not sure if it is an optical illusion, but is the smoke in the video coming from the exhaust? It almost looks like it is coming from the head gasket area or something?

Can you do a few trigger scope captures then, do one at idle, one at say 2000RPM and another at 3000RPM, attach those and a copy of your tune.

Ok, that is a dumb coil, do you have an ignitor in between the ecu and coil?

RPM dropping to zero would usually indicate a trigger error. You will see the trigger error counter increment at the same time if that is the case. Does it happen only randomly or under some more specific condition such as high load or high rpm?

In theory yes, but I had a vague recollection they needed to be sent the calibration data every startup, but I havent touched one since way back when they were in an aluminium case so I might be mixing it up with some other device. If you already have one we can try setting it up this way to see if it works before committing to more. You would still need someone with a Haltech ecu to reprogram their ID's initially however.

Yes that is a bit of a hidden trick, you need to set the trim limit tables setting to 3D and this will also change the update rate table into a configurable 3D table. Since we are using something relative to mass flow we dont really need the other axis though.

Closed loop lambda is turned on in your map but it doesnt appear to have ever been set up or tuned so it is basically not doing anything. Most of the time it is locked out due to TPS or MAP delta, the few times that the lockout conditions are met it still doesnt do anything because the minimum trim is set to 0% (ie ecu isn't allowed to remove 0% fuel). The fuel tune appears to be so far off that CLL is never going to work well, even if set up more realistically. Your lambda is just about pegged on full rich the whole time the throttle is open.

The trigger offset would normally be around -90 as per the GTT base map so I would start there. The reference timing setting is what you use to tell the ecu what to lock the timing at while you set the base timing. If your timing light isnt flashing then that is no use to you. Have you confirmed you do actually have a spark when cranking? Are you using an ht lead between the coil and plug?

The Haltech WB's output an emulated Link CAN Lambda message and then their own proprietary message for use with Haltech ECU's. As far as I know the "device ID" only applies to the Haltech messages and you cant change the ID's that the emulated Link lambda messages use.

There is not usually anything special required for start up of something new. Usually as long as the trigger is set right and the timing is somewhere close it will start or at least attempt to run - even when fuel settings are a long way off. When you get an initial cough then die similar to your video shows that is often an indication of not enough fuel. Not always though. For your 450cc injectors, I would expect the master fuel to be more like 15-25ms, did you try it that high? A quick test I often do to narrow down what to concentrate on is to try a squirt of starter fluid or if not available dribble some known fresh petrol into the manifold somehow, if it then runs for longer or gives more promising signs of life then it would confirm the issue is fuel related. If it makes no difference then I move on to confirming spark, airflow, compression etc. A PC log of it cranking would be useful to confirm the trigger is now happy. I assume by the comment "static timing set", this means you are checking with a timing light while cranking? What does it have for an ignition coil? What size is the R5 engine?

Why would you want a Miata ecu to generate a JZX100 THWO signal? Are you putting the Miata ECU into a JZX?

The aux status reporting a fault is normal, explained in the 3rd post from the top.

Assuming the idle base position is correct for idling in neutral, then it will likely need some gear/drive offset. You can check this by noting the reported idle position when idling in neutral then shift into D and watch what the idle position moves to to achieve the target speed (give it a few seconds to stabilise). The difference between these 2 positions would be what you enter as the gear drive offset. You can also add an idle up if you find it labours too much at the normal target.

Yeah I suspect lambda target is your main issue. Someone has messed up that table. Set it to 0.95 to see if it then starts and runs better. I would only drive it gently with that single cell table however. You will really want the original table or something close to it to keep the tune close to how it was intended. The injector deadtime table looks like it has possibly been played with also, as it only has a 0 differential pressure row, but given there is no fuel press sensor that will still work as it is.

If 31 now has 12V but pin 4 is not at ground then that would suggest something is damaged I would say. You can contact [email protected] if you want it inspected and repaired. Your other option since you already have a relay there is just use that to ground pin 4. The ECU doesnt need the pin 31 input if some other device is controlling the ECCS relay. So to clarify, the relay would be wired: relay pin 30 and 85 to Ground, relay pin 87 to ecu 4, relay pin 86 to ecu 24.

Your map is reporting bus errors on CAN 2. What ecu do you have?

No outputs connect to ground when the ecu is off, they are all "open" in the unpowered state. But aux outputs do have a flywheeling diode that is connected to the 14V pin so if you have voltage on an aux when the ecu isn't powered then you will be feeding the ecu and everything else that is connected to the +14V pin circuit through that aux. This is what we refer to as a "back feed". Injector 5 doesnt have a diode so cant backfeed, the only logic I could apply to the fan continuing to run would be the ecu wasn't actually off, the ignition switch might have been off but there is some other backfeed keeping the ecu powered up.

Can you attach a copy of the tune. The lambda target is 0.4 in that log so it is possibly dumping a whole lot more fuel in that necessary if you have open-loop compensation or modelled fuel equation enabled.