Wish me luck!
 

OK I have been decent with mechanics, but more the bonehead wrenching stuff rather than engines (brakes, exhaust, repairs). I have been spending the last year or two getting to know the Miata pretty well, and took the plunge with a turbo install project. I bought used stuff based on a Begi-S with a FM intercooler added. With advice from Reverent and Braineack I should have what I need.

Dimitris was kind enough to create a MS2v3 for me, which I have plugged in, and I have put in RX-7 460cc injectors. I have other things on deck, ready to go in. I had a little trouble getting the stock precat O2 sensor to budge so my thought is to leave that in for now, and put in the LC1 with the new downpipe.

I downloaded Megatune and have the laptop poised and ready to turn it on. The coil is disconnected and I guess I am ready to talk to the megasquirt. Wish me luck!

use TunerStudio instead

MegaTune won't work with this ECU. Use TunerStudio instead.

Argh! Foiled before I began! Thanks for the tip, I didn't see that, but I had been reading about Tuner Studio and its excellent reviews. OK, back to the drawing board. Hopefully this will not be a comedy of errors.

I'm not sure what I am missing now. I installed tunerstudio, set the parameters for the 460cc injectors, told it that it was narrowband, 1839cc displacement, etc. etc. The car does not start up when cranking though. The gauge panel in the TS interface does not show anything for the air/fuel ratio. Is that because I have not yet installed the wideband? Or is it going to be like this with the narrowband in place? Or is it something else entirely and I have missed a key step?

So far I have not yet attached DB15 plug to anything.

The wideband signal is brought in through the DB-15. Specifically, on the grey wire. Lookup the .xls I sent you a while ago. Do you have an RPM reading in TS while cranking?

Yes I have rpm, its like 240 or so. I understand the issue about the DB-15 wideband signal, but I have not put that monitor in yet, and the wire is not hooked up to anything so far. Is that the problem?

Probably fueling. Can you take a datalog?

I will figure out how!

This is some data from today. I am not sure which program opens it exactly.

That file is completely empty. When online in TunerStudio, click Start datalogging, name the file, and then start your attempts to start the car. Give it a few tries, then click stop datalogging and post the file here.

The file opens with MegaLogViewer.

OK I downloaded megalogviewer. This one does not seem to be empty.

It is not empty, but I don't see any attempts to start the car inside the log - RPM is always zero, the MAP sensor signal is roughly the same at all points. Btw calibrate your TPS.

Odd. I tried to start it 3-4 times during that log period.

TPS is calibrated.

Use this one. I understand now how this works. I hear a little stumbling and smell some gas. Probably ignition.

Have you changed anything in the basemap that came with the unit? If you did, post your new map here (File -> Save tune as)

I certainly may have changed things, I've been fumbling around trying to get the car to start. :vash:

Yoy had effectively disabled sequential inejction by turning "bank 2 specific settings" to "off" and in sequential settings from "additional drivers" to "standard drivers".

Try the attached msq.

Thanks. Actually I had just done that yesterday afternoon after mucking around unsuccessfully trying different things. I will try this set-up. If it doesn't work, I will reconnect the old ECU to see if it runs then; maybe I inadvertantly created a mechanical problem.

OK I downloaded the link and installed it. There were errors generated of incompatibility which on review of the error log, looked mostly like conversion. I used English units. I hope that it otherwise installed. Uunfortunately it still did not start. I need to go to work soon, so I can't do any more this morning (it is past 1130 AM here).

OK I swapped the old OEM ECU into the car. It started up and ran, a little rough, probably due to the 460cc injectors. So nothing is broken. I will put the MS back in, but I'm not sure what's wrong with the tune yet.

This is my most recent data log. I don't think I'm getting spark, but I can smell some unburned gas. Thoughts?

Let's test a few things. Pull out your plug wires and fit them with a set of test spark plugs. They don't have to be good, they can be your old spark plugs or anything. Once fitted on the wires, leave them on the valve cover, so that they ground through it.

Now go in Advanced -> Output test mode. Type in 12345 in that first box and hit Burn.

First test your fuel pump. Set it it on - you should hear the pump relay clicking on and the fuel pump should be audible in a quite garage.

Next, test your coils. Set the fuel pump to off, and then in the "test mode" drop down list, select "Coils". Then in "Coil output to test", select "CoilA". You should see the spark plugs on cylinders 1 and 4 sparking. Then on "CoilB", you should see the spark plugs on cylinders 2 and 3 sparking.

Try that and report back!

ok, but it did start and run with the original computer a few hours ago!

I did the tests. I did not hear the relay click on and off but I did hear a fairly loud whine that I assume is the fuel pump. The coils sparked appropriately. I only heard the whine when the fan turned off.

The fan cycles on and off every few minutes and that makes a lot of noise. That is new with the Megasquirt. I assume that is normal?

If you try to start the car while fully pressing the throttle, does it try to catch? I'm trying to establish whether this is a too much fuel issue.

That shuts off the gas doesn't it? I tried that the other day. I will try it again in a few minutes just to be sure, and unless I report otherwise, we can assume it didn't help.

I was going to re-check that as I left for work this morning (where I am now) but I forgot that I left the plugs out on the engine! :loser: I did the work last night between 4:30 and 5:30 AM when I couldn't sleep.

Dimitris, if we can get this solved, I will owe you some serious Ouzo! I like Ouzo, hope you do too. Also Metaxa. But not Greek wines so much, what is served here at least is not so great compared to French and Italian wines. (Rambling: my brain is fried from lack of sleep.) Thanks for your help so far. I will retest that in about 8 hours.

This seems out of the ordinary for starting issues. I am assuming that I am at fault somehow as it should have worked. I appreciate your patience. I am confused though as when I tried to load the msq you sent, error messages of incompatibility occurred in two different ways. One sounded like mostly unit conversion problems, and I changed mine to celsius but that did not seem to correct it. Another was a problem with the "prolog" not being compatible. I was unable to copy and paste these reports though. I included the above data log to check whether your msq did in fact load.

As weird as it may sound, I don't like ouzo at all, nor do I like any of the other greek drinks. I like beer, liqueurs and rum.

Anyway, have a try again with the plug wires in place. It may be necessary to give it a few cranks first without touching the throttle - this will let some fuel inside the cylinders to ignite. Then try with full throttle.

That is what I did before to test this very thing. I will do it again deliberately to see but I am not hopeful....

So it did not work. This is a datalog of the effort. I tried it off the datalog with just depressing the accelerator from the start. There was no sign that the engine started to catch.

You've probably already covered this, but your IGN output wires haven't been accidentally swapped have they? What happens when you swap IGN1 and IGN2?

I have been going on the assumption that the MS wiring is set up correctly. Testing the coil output via Tuner Studio suggested that the spark output was correct. I could easily try switching spark plug wires.

Do test that...even though you said that it was correct when you tested it through TunerStudio.

Switching around plug wires did nothing. I have not considered working on the MS unit itself. I am beginning to wonder if that might be the problem? Could there be something wrong with the MS wiring? This would be a challenge if so!

At this point I am completely at a loss.

Do you have access to a test light? The ones used to test if a wire in the car is live. You can use one of these to test if the injectors are getting their signals properly. Remove all of the injector connectors, and test each connector individually, using the output test mode.

InjA = Cylinder 1
InjB = Cylinder 3
InjC = CYlinder 4
InjD = Cylinder 2

Make sure to set the injection pulsewidth to a large number (ie 30ms), and number of injections to something also biggish, like 500.

I don't have a test light, but I have a multimeter. I haven't used this kind of setting before, but I think it works that way on one of the settings.

How would you suggest I do it? Have the ignitition on? Not cranking though, right?

You need a test light or something to that effect, not a multimeter. If you try to use a multimeter to measure the current, you will burn the injector drivers (transistors) inside the ECU. Connect the test light to both pins on the connector pins (on the harness obviously, not on the injector itself). Do this for each injector individually.

Then follow the same procedure as for the coils, only select Injectors instead of Coils in the test list, then select InjA for the first cylinder, InjB for cylinder 3, etc (as noted above).

OK I got a test light. I made sure the test light worked. I used the procedure outlined above. There was no evidence of input to the injectors. I did not test the coils independantly. I must admit I am getting discouraged.

Did you hit n run your bumper into Jackson Pollock or something?

Related threads strikes again.