Adamw

In traditional fuel equation, Diff Fuel Press = Fuel press - MGP. MGP is MAP - BAP. Since your MAP and BAP both match each other with the engine off it would be a pretty fair assumption that both are accurate (BAP is factory calibrated). So the fuel press sensor is the one I have least confidence in, but it appears to be giving realistic values so my feeling is the variable differential the ecu is reporting is probably real.

No I think that is a 040-2 with the little tabs poking out the back. You want the ones without those tabs. The ones I have in my terminal box are 173716-1 so they are the only ones I can say I know they fit off the top of my head without doing the research myself. They are specified for 1.8mm dia insulation so have quite big wings on them but they do crimp smaller wire ok (link wire is about 1.5mm OD from memory).

Unplug coils and use an LED test light to confirm there is a pulse at the trigger wire when in Ign test mode. If not then work backwards from there towards ecu to find where the signal disappears.

DPInst = Driver package installer, I think that is the installer for USB drivers.

Basic set-up looks ok, assuming it is a 150psi fuel press sensor. Does the MAP match BAP when the engine is off?

Would need to see the log and tune.

You really need a CAN sniffer or oscilloscope to test if there is data and the waveform is correct. One quick test you can do is unplug the dash and measure voltage on the CAN H & CAN L wires coming from the ECU. Measuring between ground and CAN H you should see a little above 2.5V, and measuring between ground and CAN L you should have a little below 2.5V if there is data.

I would be most suspicious of the reg. Is it a good name brand one? Some of the Chinese ones I have seen are absolutely terrible.

While you are at it, give us some photos of the CAN connection to the ECU and at the Lambda end as well.

Although not commonly done, those feedback (you call tach) pins should be ok commoned up like that. Most people wouldnt connect them at all as Vaughan says but the FT60 project I was involved in that used similar coils had them all spliced together and they worked fine. Can you tell us what ecu pins those wire colors relate to. But both those coils are wired to the same trigger, why do you say only coil 1 should be firing? Both 1 and 4 should be sparking when you do a test on ign 1.

You are currently on firmware 4.5.1, I dont see any major fuel or ignition control changes after that that will affect your tune so you are ok to go ahead and update to the latest. Make sure you have PC Link V4.10.2 installed on your laptop, you can check by going to >help>about in PC Link. If not you can download it from out website. Then connect to ecu, go to >ecu controls>update firmware and follow instructions on screen.

Ok, possibly 3 problems. Firstly, Engine fan 2 is set to 0°C, this will have an effect even without a Engine fan 2 aux ouput assigned, so this means the idle fan step is always active. I suggest change Engine fan 2 & 3 temperatures to 120°C, then they wont interfere. Example below. Secondly, stepper motor idle valves only move quite slowly, so they dont often compensate well for fast load changes such as fan turning on. It is best to use idle ignition control to help with this. Some suggested settings below which should be a good starting point. Thirdly, your battery voltage drops from 14V to 11.8V in some cases when the fan turns on. The lambda shows very lean at the same time, so this may indicate your injector deadtimes arent accurate. This voltage drop may just be related to the RPM dropping so may fix itself if the adjustments above stabalise RPM. But if battery voltage still drops a lot when fan turns on then it may suggest there is a bad ground somewhere. You may need to experiment with injector deadtimes to improve lambda stability.

It is not wiring related, they are just highly sensitive knock sensors. I have seen this on almost every 2J log that gets sent to me, the vast majority would be stock looms and I have seen the same effect with other ecu brands I have supported. They output very high voltage when excited near their resonant frequency. The factory ecu either has a low impedance input or a higher voltage measurement capability to work with these sensors but most aftermarket ecu’s wont. If you have knock level global values getting near 1000 like floppy’s then that means the input side of the amplifier is maxed out and adjusting gains or frequency etc won’t fix it.

You will need a CANPCB cable and a CANF plug to connect to the CAN bus. The CANF plug needs wires soldered into it.

The pressure cal is listed under its bosch part number as below. The temp calibration is "Std Bosch NTC", you will need a 1Kohm, 5V pull-up wired to the temp signal wire if connected to a AN Volt input.

attach the map.

It was correct in my original example since I was using little endian. In your example above you are using big endian so yes cosworth are counting from the other end. But I think you are also possibly confusing the terms "Start position" with "Start bit". Start position is just refering to the position in the drawing at the bottom of the link set up screen, not the actual bit numbers. Thats why it is always best to use the visual representation like you are now (when they have one, many manufacturers still dont!). So if you look at Engine speed in the link drawing it is sitting nicely accross byte 0 & 1. If you look in the Cosworth table engine speed is again nicely spread over Byte 0 & 1. Some manufactures add further confusion by starting the byte and bit count at 1 instead of 0.

Will need to see a log of the TPS error to see if there are any clues. Do you have plenty of margin between normal TP voltage and the values you have set as error values? That sounds like it has been disabled, 6.5% is about where the spring pushes the blade to with most throttles when unpowered.

Ok thanks, I will ask them to take another look.

Email [email protected], I dont have access to the sales records personally but Im pretty sure the tech guys will be able to look up the dealer or if not they will pass you on to someone that can.

CAN is only going to end in tears in that case. It might be easier to stick with analog. Connect the AEM white wire to Volt 5,6 or 7 on your expansion loom, connect the AEM brown wire to the gnd on the expansion loom.

Note gear lever force isnt calculated when shifting from neutral. Your log shows you were in neutral for the test so that is expected behaviour. Have you experienced it not working when actually gear shifting?

Attach a log showing the problem and a copy of your tune.

I will PM you a new firmware to try. The firmware engineers believe they have found the issue with the trigger. My bench test didnt show the error as I was using constant 200RPM. When tested on your engine and the RPM suddenly slows down during cranking it causes one of the trigger tests to timeout and fail.

That would have to be our Miata adaptalink (ALMX) as far as I know there was nothing else similar. So Pinout on the car side should be standard NA miata. To confirm you can slide it out of the case and it should be labelled LEM G3 - Miata/GTX