91 NA MS2 -- 2U and 2V weird voltage
 

---SOLVED--- see edit below
Hi Everyone,

I tried searching really hard but probably it is a very specific issue of mine so I could not find a solution (I made it bold for easier reading).. Hopefully I did not act too noob for now :)

I have '91 NA Turbo, 460 cc redtops and Diyautotune diypnp. When we switched over (I got an auto electrician to reground and swap from greddy emanage piggyback to ms2) it was working fine. I did some "safe" tuning and was playing around with launch control and proper idling. One night I felt stutter at some rpms and on the way back from another test drive the car died on the road. Long story short, it starts up with only injectors 1 and 3 plugged in. 2-4 does not make a difference. What I tested so far:
1- Spark is good - but still got a plug and play cop system ready (thought the igniter went bust with all the pops and bangs)
2- Fuel is good because the car start with 2 injector only.
3- Unplugged all injectors - tested resistance, normal ~14 ohms without much deviation
4- Unpugged all injectors - checked voltage across terminals at each plug (yeah a noob move) 2-4 had 9 V, 1-3 had 12 Volts
5- Got MS2 into test mode, fired injectors for 10 ms every half a second, I can get continuity (multimeter continuity test) on 1 and 3 - readable reaction on the multimeter but not long enough for a sound. Multimeter goes haywire on injectors 2-4 with same test. (all the characters flash) -- I also cranked with full throttle to cut fuel and clean out flooding occasionally.
6- Tested continuity on the end of the injector harness ( 6 pin one near CAS). No broken lines etc.
7- Tested continuity all the way from injector plugs to ECU pins 2U 2V and 1B, can get solid beeps
8- Tested voltages of 2U 2V and 1B directly with a needle (probe :D ) at the ecu while the injectors are plugged in, 1B is 12V (same as Battery), 2U is 10V, 2V is ~7.8V !! (all of them are supposed to show 12v, directly tested with ground, not between them). Injectors were flow tested about 1000 km ago but I am not sure if 2 injectors can die at the same time. This is the only solid out-of-order thing I noticed.

I fear that this might point out to something wrong on the ms2 board, but wanted to ask you all before I go all panicky :) I am starting get convinced that a capacitor or sth inside the ecu is fried up.

Thanks all for all the shared the knowledge on the forums, really taught and helped me a lot. Just got stuck here and all my searches end up with bad wiring (where the car would not start).
PS: Should I be happy that it can idle for almost 1 minute with 2 cylinders, albeit extremely roughly ? :)

Evo

Edit: It turned out to be the u7 mosfet on the microsquirt board, got it replaced with a "similar" mosfet, since original ones are hard to find. Injectors started firing up again. Something else failed afterwards but at least this issue is solved. Main point of the story: Make sure you are actually running on 3 cylinders before trying to diagnose why :D If I realized whole bank was gone last year, this could have been solved quicker and the car would not have had bigger problems, in my opinion anyhow, mostly due to sitting dead for 1 year. (in an underground garage but still)

Edith Piaf: Added a new final post on this i.e. bent rods :(

I mentioned my new problem on the last page but will probably open a new help topic soon :D

Just a small update, since I assume I am unique with this problem in the near past. Special thanks to Matt from Diyautotune, in at least trying to support me. There is no conclusion yet.

I opened the board and started tracing injector wiring based on diypnp web documents. Connections are correct, how ever, resistances on the Microsquirt module are different than the advertised, there should be a 1k and 100k ohm for each bank, however I have 3 1k resistance measurement and 1 47k resistance. Funny thing is, that strange resistor is on the working bank, unless I measured something wrong at 2.00 am :D

Edit: actual measurements:
R31 - R29 - R32 all measured 1k ohms (even though R30 and R32 are supposed to be 100k )
R30 on the other hand reads 42k ohms. Considering inj 2 is the problematic one, this is strange.

you shouldn't really have voltage on 2U and 2V at all. each injector has constant 12v on the white/red wire from the INJ fuse, the MS provides a ground to the other wire to complete the circuit.

use output test modes.

check for GROUND on 2V and 2U.


https://cimg4.ibsrv.net/gimg/www.mia...cd9be0ddf0.png

Hi Brian,

Thank you for the reply. I based 12v assumption based on the famous pinout diagram:
https://cimg9.ibsrv.net/gimg/www.mia...4a78a7905e.jpg2u 2v voltageI put the ECU back in the car, hoping that it was the usual strange jinx but nope. Tried to check again as you mentioned, here is what I observed:
-This time 2u had 12.7V (exactly as battery) with small fluctation of 0.1-0.2 V (just ign key on)
-2v - the problematic bank, was fluctuating from 10 to 11.3 V. Almost rythmically. (just ign key on)
- I tested white/red wire, stable voltage but above 13 V !. I also have an injector like plug near the igniter (probably PCV plug? ), hanging unconnected with white/red (same wire, same voltage) and seems like yellow/red, which has no voltage but not ground
-Testing outputs: INJ 1 very audible clicking, however this time I noticed that the multimeter reactions do not actually change with test parameters (interval and duration).
-Testing outputs: INJ 2, no reaction again, as above, no correlation between flashing continuity to ground to test parameters
-All tests used ground wire (black) on the ecu as well as hard point on the body - same results
-This time, the car could not even start on 2 cylinders. Just some fake starting sparks around 60% throttle.

Attached is my tune and my log file from the first post (cranking and 2 cylinder idle) as I forgot to log this time, along with some logs from the last time car was running - I was out trying to adjust silly stuff like crackle etc. (never tuned for spark as this problem happened before I could ask someone to drive for me while listening for knock).

I will try to order a simple led and a 1k resistor to make an indicator light, as it is not possible to notice switching. I am probably making a noob mistake while measuring - like thinking that just measuring resistor while attached to the circuit should give the right value. I edited my previous post to show the measured values.

When the injectors are connected, you should see battery voltage on 2U 2V KOEO. A voltage lower than battery voltage means that you either have current flowing into the MS driver circuit (broken) or you have a bad connection on the injector high side, or you have current flowing from the connector low side to another signal (presumably ground) between the injector and the ECU. Napkin calcs indicate roughly 300mA of leakage current, whic is pretty significant. My guess would be a semiconductor break down.

disconnect injectors 1 and 4, measure voltage KOEO, should be similar to what you see today. Swap injector plugs 2 and 3, measure voltage. If the low voltage follows the injector (swaps ECU pin) your problem is on the chassis side. If the problem stays on the same ECU pin, the problem is the MS, most likely the injector driver is blown.

IIRC, there are 2 yel-reds in the NA. One goes to the oil pressure transducer, the other to the evap solenoid. Sounds to me like you are talking the evap plug, like 2X? Once again, going off of memory here.

Thank you very much for taking time to reply, based on suggested items, here are the new findings:
1- Most obvious observation, (that was very noob of me to not do the first time because I was measuring voltage across terminals with all inj unplugged - not to the ground.)
Anyhow: Yellow/black - both injectors unplugged, KOEO-, still has the same -1v battery voltage to the ground.
Yellow wire, good bank, does not read any voltage when both injectors are unplugged . (which I think is the correct behavior)
Only one injector of each bank plugged: firing bank reads same as battery (I was boosting the battery last time) - shitty bank reads -1v battery.
--Basically, without any resistors on the line, it still reads voltage to the ground.
2- Unplugged ecu harness, tested resistance from 2u and 2v to the unplugged injector leads, 0.6 and 0.8 ohms respectively, so no suspected shorts or other mixing with other circuits
3- Measured resistance from 2u and 2v of ecu to the grounds, connected to the car with all the injectors and key off, battery connected, they read very similar value (something like 145k iirc) - no difference should be logical
4- used a test light from 2u and 2v to ground, while output test mode, get flicker on 2u (and of course injectors firing) and no flicker on 2v.
5- Unplugged all injectors and connected banks differently, voltages moved with banks (unplugged gives the same results, as mentioned in #1)

Overall, I think one of the injector drivers is acting up, since both channels behave differently even with no injectors connected. Probably U7, VND5N07 is gone. But I have no idea how to check it with my basic tools or even replace it on the micro board with basic solder iron.
I am lost as to what to do next. :idea:

Ps: I am thinking if it would be possible to add sequential injection circuit in the proto area (or replicate second injector circuit somehow) and configure it to be batch firing but use that driver for the second bank instead of this. Ideal solution is to get a new microsquirt board, I think, but I do not have the tools to update the same firmware as in my signature. Or maybe switch to MS3 :D

Don't know if this matches your HW, but maybe ot will help.

https://www.diyautotune.com/support/...jector-driver/

I just probed a ms3x here. i see no voltage on the INJ outputs when activated.

the outputs don't float to 12v like the stock ECU does...completely different circuit.


test the outputs in diode mode (beep) put the black to ground, and red to the 2U or 2V wire -- it should beep rapidly.

there should be a voltage if the injectors are connected and the drivers are turned off, no?

The ground doesn't float, see the diagram I posted of the driver.

Ok, let me rephrase

when you have injectors connected between 2U or 2V and the WHT-RED (main relay out) you will see voltage on 2U and 2V KOEO.

Thank you all very much for the support :likecat: I am now definitely confused :D is there another source to check the voltage values of the 91 na harness, besides the attached -very helpful, thanks to the creator- diagram?
This is of course based on the oem ecu.

I did not want to do that with multimeter, because in with key on, engine off, there is voltage relative to the ground at 2u and 2v. As I mentioned, if I put my homemade test led diode to 2u or 2v and to the ground, it lights up constantly with key on, on both sides. When I test output mode, diode flashes off (as opposed to flashing on) with the working bank, consistent with test interval (you can loudly hear the injectors clicking too as I mentioned). Non working bank is not responding, not giving any flashes.

Ps. I did one stupid thing, wanted to ask what could have happened, just for my self learning, the above test, I also wanted do one final test and try to crank the engine to see if the shitty side will respond the same with diode led (2v to black ground on ecu). TO my surprise, as soon as the key moved to the crank position, I heard a loud crack and the car went completely dead electrically. Turned key on and off, no response - like complete zero response, no lights gauges etc. I checked the fuses, everything seemed ok. Then I just disconnected and reconnected battery positive lead, everything went back to normal, including injector problems etc :) I assumed that since people were starting their cars with test light connected on youtube to check injector pulse, and output test mode was ok, it would have been ok. What was my mistake here ? (led diode has 2.2k ohm, as I couldn't find a 1k readily).

PPS: I found a guy in my area, who has been repairing megasquirts for some time and came recommended. I will hopefully take it to him to replace u7 mosfet. I still have the ecu with me if you guys want me test something else.

PPPS: MS can not turn the pump off in the test mode. This is not a problem normally as the car was working fine but I think I should correct this while knee deep in harness stuff. Any suggestions on typical reasons?

where is the voltage on 2U and 2V coming from?

the MS injector drivers (again, posted above) do not float to 12v on the 2U/2V side. They connect to FETs which only switch the drain to source (ground) when positive voltage is applied to the input. There's a pull-down to ground on the input to the FET.

If OP is seeing voltage on the 2U/2V side of the wiring, then there's a 12v ground fault somewhere.

yeah, this is NOT normal. there's a fault in either your MS wiring or car's wiring.

I would be very suspect of the re-wiring from the greddy emanage, especially since it initially ran.

do the same test with the MS completely removed from the harness.

The voltage comes from the injector, which is connected to vbatt through the main relay. When the injector driver is not active, it does not sink current. No current through the injector means no voltage across it. 2U and 2V are connected to the injectors.

The results from your diode test are normal, assuming one of your drivers is faulty. If you remove the harness, you will not see voltage on 2U or 2V, and your diodes will not light.

The MS controls the FP through 1C. An active high on 1C should turn the pump on. This net is also connected to the ST SIGN fuse, which is why IT MUST BE REMOVED. If the fuse is still in there, the MS will not be able to turn the FP off.

OP said he was seeing voltage on 2U/2V. Meaning the black probe of the DMM is on ground, and the red probe is on either 2U and 2V.

Yes, of course there's 12v on the white/red wire. That's the voltage supply for the injector solenoid.

Are we not literally measuring the 2U and 2V wires here??? that makes a BIG difference. And otherwise I don't see how it's possible to have 10v on the white/red on injector 1/3, but only 7v on the white/red wire at injectors 2/4, since all white/red wires come from the same source.

you think it's normal to be able to light an LED by using a ground wire as the power source?


when I read what the OP wrote, I'm picturing this:


https://cimg6.ibsrv.net/gimg/www.mia...efdfbcc6c6.png

IF the MS was wired to do so...