Need help troubleshooting...

Car in question is a b21ft running Microsquirt, DSM CAS, LS Coils. Engine itself is pretty much stock with a fresh rebuilt stock turbo, intercooler, T5 blue injectors, Spartan 2 wideband. This was a Kjet to MS swap I completed last year and, up until earlier this month, its been running great. Never lost sync while driving with the laptop in the car...aside from just needing some tuning refinement its been most excellent.

Few weeks ago I was having a little fun on some back roads when I blew the charge pipe off the turbo. Loud turbo noises, no power, but it was still running. About a minute later as I was pulling into a safe spot to address the pipe the car just died on me. So fixed the charge pipe, double checked everything was in place, and all was visually well. Car would not start. Sat for a bit figuring out my options to get home and tried one last time...car fired up and was running just fine.

Drove it another 15-20 minutes when it died on me yet again. Let it sit a while, car will start but only able to get it to drive maybe a couple minutes before it leaves me on the road. So I get a tow.

Get it home and take it out with my laptop to datalog to see whats going on. Thankfully I get it to die again. Look over the log with a friend and we see some very strange MAP spikes. So I get to probing around with my multimeter which led me to do the following:

- Found alternator ground cable melted. I was an idiot and used the old one when I did the Denso 80a swap. Looked good at the time, but yeah I'm a dumbass. Replaced the ground.
- Had alternator tested. Output and rectifier good.
- Went over all grounds in engine compartment. Found the newly purchased battery ground cable ring terminal was never crimped onto the cable. Fixed that. Inspected, recleaned, etc all other grounds.
- Pulled the intake manifold to inspect my MS wiring underneath to make sure all is good.

This is what I've found with diag both before and after all above was done:
- Key on not running my 5v reference and 12v power supply to ECU are good
- Key on with no sensors plugged in my 5v reference is good
- Key on not running 12v to CAS is good
- Tested CAS by rotating by hand with 12v supply to it and based on what I found online the output readings look good
- I'm averaging 13.6-13.8v at the battery and starter at idle
- With car at idle the 5v reference is measuring an average of 7v
- MAP output is all over the place and exceeding 12v in fluctuations at idle and when revving motor
- Backprobing 12v supply at the CAS at idle I am finding voltage exceeding 14v.

All the idle diag done above was with the car idling up to operating temp but not after the car had died. Seems maybe this is a heat related issue of some sort as the car will restart and run just fine for a short period of time after sitting a bit after it dies from driving.


At this point I'm getting beyond my knowledge and could really use some help/direction on where to go from here. Drive link below has my tune file and four logs if anyone would like to give them a gander. The new newer logs were from today and are considerably shorter than the others:
https://drive.google.com/drive/folders/1ab0P8letfjHofUqVCPQidJhoM2f8WUve?usp=sharing

Alan29 said:

- With car at idle the 5v reference is measuring an average of 7v
- MAP output is all over the place and exceeding 12v in fluctuations at idle and when revving motor

Click to expand...

Seeing more than 5v on the 5v reference is bad and so is more than 5v out of the MAP (the MAP only gets +5v for power). Where did you measure the 7v, specifically, where were the 2 meter probes? Likewise for the MAP output measurement.

I'd check the MAP ground wire carefully.

I backprobed at the map connector. Grounded meter to battery and then at the sensor ground at connector yielded same results.

Hmmm, next, I'd try wiggling the MAP wires at idle and see if you can get it to misbehave.

I don't think that the +5v Vref wire is shorting out to +12 because it's used by the TPS too, and I don't see TPS misbehaving at the same time as the MAP in the log. But something is certainly wrong to get more than +5v on Vref and on MAP output.

Also, for your 14v battery reading, check that the battery cable clamps are tightened onto the battery posts. If they're just pushed on hand tight, the battery may not smooth out the alternator voltage.

Will do with the harness wiggle. Battery terminals are on tight with zero corrosion. All connections for battery are solid.

The CAS voltage feed at idle is blowing my mind. With key on not at idle I get a few tenths under the battery voltage which is what I would expect. Don't get why it's higher than battery/starter voltage at idle though.

At idle with harness wiggle there is no change. So I did disconnect the MAP and cranked to see what happens with my 5v reference at the MAP connector. Didn't expect to run but while cranking and trying to fire I was getting over 5v. Would this possibly be an internal fault in the ECU?

Off hand, I can't think of a way that the ECU would be putting out over 5v while still working and still logging normal voltages on the battery. I'll look at the schematics tonight.

Inside the ECU, the +12v battery voltage is regulated down to +5v. The +5 is used by all the internal ECU circuits, and a filtered/polyfused version goes out as the +5Vref voltage. The +5Vref goes to the MAP and TPS in the engine harness. I'm still leaning towards something in the engine compartment shorting +12 to the +5Vref wire, but TPS in the logs doesn't misbehave when MAP misbehaves.

I'm also directing the 5v reference to the two CAS outputs via 1k pull up resistors. I truly appreciate the help.

Vcc is slightly isolated from Vref for transient voltages by L1. Vref should be further clamped by Zener D7 to somewhere around 6.3 volts. If Vref inside the MS box is getting significantly above 6.3 volts D7 would have to be damaged and unless fuse F1 opened that would have backed up to Vcc which would have damaged other things. If the fuse opened you would have other problems. Either case should result in complete non operation.

Perhaps like bobxyz, I am smelling a wiring problem in the engine compartment. My initial reaction is that the ground for the MAP sensor is damaged and you are getting a voltage on the ground which is influencing the MAP sensor output. However, this doesn't seem to jive with your statement

So I did disconnect the MAP and cranked to see what happens with my 5v reference at the MAP connector. Didn't expect to run but while cranking and trying to fire I was getting over 5v.

Click to expand...

unless you didn't observe the voltage polarities when you did that measurement. At the MAP connector, if you measured from Vref to the ground on the MAP connector it is possible the ground is floating up to 14 volts which might cause you to measure inverted voltages as high as - 9 volts (compared to Vref). The test for this would be to check the voltage on the ground terminal of the MAP sensor relative to various ground points on the car (Battery neg, Alternator ground, MS common ground). Any voltage present on the ground terminal at the MAP sensor will superimpose on the MAP sensor output which could explain why you MAP sensor output voltage

is all over the place and exceeding 12v in fluctuations at idle and when revving motor

Click to expand...

. A ground connection problem for the MAP sensor could explain why its just the MAP sensor voltage that seems to be a problem.

I'm averaging 13.6-13.8v at the battery and starter at idle
- Backprobing 12v supply at the CAS at idle I am finding voltage exceeding 14v.

Click to expand...

Depending on how you have wired the 12 volt supply on the car, it is possible to have a voltage above the voltage measured at the battery / starter. On my car the alternator + terminal is connected directly to a main PAL fuse. The load side of the PAL fuse splits to dedicated engine fuses / relays in the same box with a separate connection to a 12v buss on the other side of the engine compartment which goes to the battery + terminal and supplies all the other car loads. My CAS which is supplied from the engine dedicated engine buss will always be higher than the battery terminal voltage. Connect your voltmeter between the CAS +12v terminal and the + terminal of the alternator. Both voltages should be the same so you should measure a differential of 0 volts; however, I would expect that depending on the quality of your wiring that the alternator might measure 0.1- 0.2 volts higher than the CAS +12v. If the differential voltage measurement indicates that the CAS +12v terminal is higher than the + terminal on the alternator you have something seriously screwed up.

Thanks 142...gives me some direction for this week for sure.

I pulled my 12v supply for everything (CAS, MS, etc) from the junction under the hood to the new fuse box and relays located right behind the battery.

Looking just at the logs, the MAP sensor is fine for a long time and then pegs at max pressure, which causes the ECU to command lots more fuel, which, I'm assuming, floods the engine and it stalls.

All the sensors (MAP, MAT, TPS, CLT) use the +5Vref signal, but the logs only show MAP misbehaving. The battery voltage in the logs is also fine before stalling. Based on these, I think +5Vref must be OK and not overloaded or shorting out.

What I can't explain is the weird measured voltages on MAP and Vref in the engine compartment. Do you have easy access to a labscope? It would be interesting to see if it shows anything more useful on the MAP and Vref wires. Even easier, do you have a 2nd meter, and does it get the same weird measurements?

Is there anywhere along the MicroSquirt harness bundle that might have been damaged? MAP sensor output wire in particular.

bobxyz said:

Looking just at the logs, the MAP sensor is fine for a long time and then pegs at max pressure, which causes the ECU to command lots more fuel, which, I'm assuming, floods the engine and it stalls.

All the sensors (MAP, MAT, TPS, CLT) use the +5Vref signal, but the logs only show MAP misbehaving. The battery voltage in the logs is also fine before stalling. Based on these, I think +5Vref must be OK and not overloaded or shorting out.

What I can't explain is the weird measured voltages on MAP and Vref in the engine compartment. Do you have easy access to a labscope? It would be interesting to see if it shows anything more useful on the MAP and Vref wires. Even easier, do you have a 2nd meter, and does it get the same weird measurements?

Is there anywhere along the MicroSquirt harness bundle that might have been damaged? MAP sensor output wire in particular.

Click to expand...

I don't have access to a labscope but I'll borrow another multimeter. Yesterday I used my multi for some home electrical repair and it's reading right there too so I think it's good but can't hurt to compare.

I pulled the intake mani on it to inspect all the wiring as that is where the majority of my connections are and didn't see anything out of sorts. I think I might get a friend over here to get eyes on it as well make sure I'm not missing something.

Alan29 said:

I don't have access to a labscope but I'll borrow another multimeter. Yesterday I used my multi for some home electrical repair and it's reading right there too so I think it's good but can't hurt to compare.

Click to expand...

I use a dso138 scope for most automotive stuff now, 20$, works on a 9V battery, decent screen, cannot go wrong

As for the multimeter, I ran into measurement issues when the screws connecting the front terminals to the board inside got corroded, this is not observed with higher voltages and currents, but noticable on lower voltage and resistance measurements I chose to hard solder these connections, but cleanup should already help.

Alan29 said:

I don't have access to a labscope but I'll borrow another multimeter. Yesterday I used my multi for some home electrical repair and it's reading right there too so I think it's good but can't hurt to compare.

I pulled the intake mani on it to inspect all the wiring as that is where the majority of my connections are and didn't see anything out of sorts. I think I might get a friend over here to get eyes on it as well make sure I'm not missing something.

Click to expand...

Do you ground your MAP sensor at the intake manifold or do you run an isolated ground all the way back to the MS box? If you don't run an isolated ground back to the MS box and if you damaged the engine ground during your melting event you may have created a voltage offset for the MAP sensor.

It's tied into the sensor ground back to the MS. I am getting a significant amount of voltage at the sensor ground (tested today). DIY suggested testing my ground for the ECU and, assuming I tested correctly, that ground is fine. Still poking around a bit. Noticed a possible issue with ground for my coils that I need to revisit...showing voltage at the ground for that to my intake manifold.

So I did reach out to DIY on this too since wasn't sure if there was any sort of warranty should this end up being some kind of internal fault with the ECU. His suggestion was to replace the map and revisit my grounds. Little hesitant to drop $70 on a sensor at this point but perhaps I need to consider it.

If it's easy, can you measure a few voltages on the sensors at idle? Clip one meter lead to the ground bolt on the intake manifold, then measure:
- battery - post
- MAP ground pin
- TPS ground pin
- IAT ground pin
- CLT ground pin (if it has a separate ground pin)
- CAS ground pin

- TPS Vref
- MAP Vref
- CAS Vref (pullup resistors)

- battery + post
- CAS +12v

The grounds should all be ~0 volt offset; the Vref's about ~5.0volts; +12v about ~13.5v.

Edit: for the LS coils, do they have a separate ground for the logic level spark input and a ground for the built-in power stage? If so, where are these wired? You don't want the power-stage high-current ground sharing any of the sensor ground wiring.

I'll get back with you on findings on those hopefully today. I appreciate all the help.


The LS coils sensor grounds are tied in with the rest. The power ground I have tied to the intake mani at brace and have another separate ground going from that point to chassis where I have some non sensor grounds. I added that yesterday after poking around with the coil ground but I need to revisit my metering of that ground as I think I was doing something incorrectly.

The battery ground is also grounded to that brace at the stock location at the engine mount.

Alan29 said:

It's tied into the sensor ground back to the MS. I am getting a significant amount of voltage at the sensor ground (tested today). DIY suggested testing my ground for the ECU and, assuming I tested correctly, that ground is fine. Still poking around a bit. Noticed a possible issue with ground for my coils that I need to revisit...showing voltage at the ground for that to my intake manifold.

So I did reach out to DIY on this too since wasn't sure if there was any sort of warranty should this end up being some kind of internal fault with the ECU. His suggestion was to replace the map and revisit my grounds. Little hesitant to drop $70 on a sensor at this point but perhaps I need to consider it.

Click to expand...

Does

It's tied into the sensor ground back to the MS.

Click to expand...

mean that you have a dedicated ground wire running from the MAP sensor to a dedicated ground pin on the MS or does it mean you have a shared sensor ground back to the MS? Sharing grounds is bad practice. Dedicated / separate grounds back to the MS box with a common ground from the MS box to the engine block is the preferred wiring arrangement.

If you getting any voltage differential between the ground at the MAP connector and any other grounding location you have a grounding problem. That problem may be between the MAP connector and the MS box, between the MS box and engine ground or between the engine ground and chassis ground (on your B21 does the battery ground connect to chassis or the engine?).

The MicroSquirt ecu has one dedicated sensor ground for the sensors. I tied the sensor grounds for the map, air temp, coolant temp, wb, tps, and coils into that per their wiring diagram.

As far as power grounds...the coils go to the intake manifold, ecu and wb to negative battery terminal. My shield ground for the cas wiring is grounded to the intake mani.

B21 battery ground goes to engine and the chassis at the apron just behind the battery tray.

Hoping to get out today or tomorrow to run more tests if work and family allow.

I have seen the Microsquirt wiring instructions which say connect all the sensor grounds to pin 20. Its not a particularly good design; but, it will work if you make the single connection from pin 20 to the common point for the sensor connection short and are assiduous in the set up of the connections at the common point. The fact that your TPS and other sensors are not showing offsets indicates that the problem is not between the common point and the and the ampseal connector.