DIYPNP in stock NA 1990 will not start
 

Hi folks,
beeing the utter noob with my first ECU procject I have no other choice to ask for help.

My Stock NA 1990 1.6 runs just fine with the stock ECU. But with planned turbo project on the horizon I decided I would try to build myself a Megasquirt DIYPNP, so I did. The assembly was not really difficult. Unfortunatly I did not order the JimStim, something I regret today. If everything else fails I will probably have to order one and wait the 5 weeks or so for it to clear customs.

So I powered it up on my desk and tried to load the 029v firmware as indicated in the MSPNP docs, but it would not accept this firmware. Am I correct in assuming that 029v is only for MS1 and since the DIYPNP is a MS2 I need a different firmware ?

Anyway I found the 2.1.0q and got it loaded. Unfortunatly I could not find a basemap for this release, so I ended up upgrading to the Alpha 3.0.3s with the base map from "91 miata 240cc injectors.msq".

At this time I installed the LC-1 lambda in the car and ran the "narrowband simulated" signal to the ecu instead of the original signal. This works out great when I run the stock ECU so my though was then to get the DYIPNP to start on this config before I change anything. Yes, I did read the note on this forum about not changing more than one thing at a time. This is the only change I have made to the stock engine.

When I replace the stock ECU with the MS2 I do get the signals I would expect from temperature probes and MAP. When I crank the engine, TunerStudio detects this and shows up the "cranking signal" in the bottom of the screen, but no RPM reading at all. I assumed this was a faulty CAS signal so I borrowed a oscilloscope and took readings from both original ECU and MSPNP while cranking. Signal seems to be fine on both, but this was an old analogue scope and it was really difficult to get a fix on either of the 4g63 signals. Beeing a bit afraid to break anything I took the CAS and MSPNP out of the car and spun it up on the desk. Still no rpm signal, and the diagnostic readings are not something I understand at all. Scope shows that signal gets all the way onto the Microsquirt daughterboard.
There is nothing at all on the Composite Logger or the Sync Error logger. The tooth logger shows up with different info depending on the frequency I use. What is the correct frequency to use when debugging ? Ended up using 250ms ! It then shows up three bars of similar height while cranking the engine?? So ?

Am I correct to assume that I should see the cranking rpm in TunerStudio, or does this output only show after the car starts up ?

I did try to reverse the cam/crank signals from the cas, to no avail.

In a final despair I downloaded the Monitor_tooth_analyzer_v2.4.abs.s19 and the tacref.exe file. This firmware shows me a clear picture of the 4g63 cas as expected. The cam line is one wide and one short, and the crank line is just four regular pulses.

I have invested too much money and energy in this project to give up. Any ideas or tips for how to troubleshoot this ?

-

First off, keep in mind the MSPNP and the DIYPNP have almost nothing in common besides the tuning software. You should be seeing RPM while cranking. I'd first check the configuration. Please post the MSQ file currently in your ECU.

Sorry absRTP for not having my profile updated. I DID read the noobs forum yesterday, introduction and profile will be fixed asap.

I´m located in Oslo, Norway.

I have attached CurrentTune and a logfile for you Matt, I hope they will be available. I´m grateful for any insight you could shed on this problem.

I have used the http://www.diyautotune.com/diypnp/docs.html site for assembly and http://diyautotune.com/diypnp/apps/n...3-16b6-mt.html for configuration with the 240 injectors.

Regards Kjetil

-

Forget to pull the ST_SIGN fuse by chance?

EDIT: NM...that affects fuel pump circuit, not rpm signal.

Thanks, gospeed. I have been very careful to pull the ST_SIGN fuse everytime I have tried it in the car. I can hear the fuel pump running for a couple of seconds each time I switch on the car.
For now I have it on the desk with just a spare CAS connected that I turn around with a powertool.

Slick!

Good luck with diagnosing you unit.

If you're not seeing anything on the composite logs when cranking (Have to crank for a good 10secs or so for it to refesh) then I suspect you didn't connect something correctly on the build. Can you take a picture of the board? did you forget the 470ohm pulls up for the VR+ and OP- (Crank and cam signals)

I do not have a pull-up for VR+ , but there is one for VR2. Maby doc is wrong ?
I added some pictures for the front, the front without MicroSquirt and the back.

Thanks a lot guys for your help and insight.
Regards Kjetil , btw. my "presentation" is not yet written. I will get around to it soon.

When I mesure VR2 (G-signal) signal on the Microsquirt pin 14, i can see the short+long pulses. Voltage is 0,35 at low and 4,95 for high. High makes sense because of the 5V pullup. On the OP- (Ne-signal) with Microsquirt pin 21 I can see the four even pulses with 0,33V at low and 11.39V at high. High voltage I guess is because of the 470ohm pullup to 12V. Is it really correct that these two signals should be "pulled" different ?

I did however find something peculiar about Tunerstudio. Until today it showed no graphs for Composite logger. But by scrolling on the mouse wheel 4 clicks up in the graph section there is something actually showing. Funny thing about these graphs is that they will appear even without running the camshaft. Noise or bug ?

The composite log you posted looks like noise.

You need to remove the opto ground jumper. ;)

Hi Ben, I grant you that it looks like the jumper on opto ground is there, when it in fact is not. I will examine for shorts to ground on that part of the circuit. Could be a faulty soldering, after all I'm not a pro.

Kjetil

I'd remove the BOOT Jumper myself...

It's hard to tell in the pictures, but it should not be installed. To test for a short, use your DMM in continuity test mode. There should be no continuity between the square opto ground pin and ground.

As Scott noticed, you will also need to remove the boot jumper, which is something you need to do part way through the firmware loading process. There should be no continuity between the square pin on the boot load jumper and ground.

Oh joy! Was tracing the cam signal into the MicrosSquirt when I discovered that a capasitor on the drawing was missing. Contacted Matt at DIYAutotune and he could reassure me that they intentionally had removed that capasitor C27 since it had cased some trouble for Miata users. Thank you Matt!
With that out of the way I discovered that signal was not actually getting past the pin on the connector. Tried to resolder, but ended up adding a wire from pin 14 to bottom of R50. Booted up and spun the cas around for a few seconds and got this picture from Composite Logger. Looks completely different compared to earlier attempts. Beeing the noob that I am, I do not know if this is expected results, but from searching around bit in this forum I know that not having the red trigger error is a good thing. Still it shows up a few times at the beginning. Thank you everyone for your help. This was not an easy one for me. I´ll verify all the settings I have been playing around with when trying to figure this error out and reset them back. Then I will pop it into the car tomorrow. Will let you know if it runs or blows up !

Kjetil

That's what your composite log is supposed to look like.

A sync error occurring right when the CAS begins to spin is not uncommon or anything to be alarmed with.

Yes, it works in the car now. My coolant temp is 30% off so I will have to fix that. It wont idle properly, but no tuning is done yet. I will close this thread now since initial problem is fixed! Thank you folks for your help!