NB SE Tacho and Odo stopped working

KevGoat · Postby **KevGoat** » Mon Aug 24, 2015 5:42 pm

93_Clubman wrote:Does that underlay below the ECU extend down far enough to wick away any moisture collecting at the base of the ECU cover? If so, you might want to pull it up a little to avoid any possibility of that happening.

Good point. I'll make sure it's placed away from the lower edge when I put it all back together.

Wonder if it's worth grabbing some foam sealer strip to put around the edge of the cover to stop moisture getting in? Or would that just be likely to cause condensation instead?

Just went out and re-plugged the ECU back in, after leaving it unplugged for an hour or so. Figured doing that might reset or possibly change something? Made no difference to anything.

Changed fuses - no difference.

Definately need to plug in another instrument panel to either solve or discount the actual cluster.

Not sure if I need an NB8C cluster or not. From memory, I think the instruments would work even with an NB8A cluster.

Anyone in SA have a spare NB instrument panel I could borrow?

93_Clubman · Postby **93_Clubman** » Mon Aug 24, 2015 6:26 pm

KevGoat wrote:Wonder if it's worth grabbing some foam sealer strip to put around the edge of the cover to stop moisture getting in? Or would that just be likely to cause condensation instead?

Think it would be better trying to stop water getting in at all, but I'm not sure how easy or difficult that is on an NB. My NAs have always remained completely dry inside, but the forum has had quite a few NBs over the last five or so years where water appears to have affected the ECU. Admittedly, some of those have been flood affected.

KevGoat · Postby **KevGoat** » Tue Aug 25, 2015 5:49 pm

OK ... so I've made some progress!

While mucking around today, I found that the circuit strips going in to the cluster's microcomputer affected things. The tacho was bouncing, but this time it didn't completely turn off - maybe because of the good clean yesterday? So from this I figured there had to be a contact issue somewhere. I kind of pushed and prodded carefully around on the circuit at the back while the car was still running and found that the tacho and odo suddenly stabilized! I had been touching around the microcomputer area. So I took it all out and carefully tried to slide the wide part of the strip that goes in to the MC, trying to improve the contacts. It didn't even feel like I moved them, but the instruments have been staying on since! Taken it for a couple of short drives around the block and still working. Left it idling for about 15/20 minutes while doing other work inside the car and it didn't flicker off at all.

BUT ... not all is perfect with the world yet! My speedo isn't reading the correct speed. It's signal seems "damped". I pulled the instrument out of the cluster to see if I'd damaged anything with all my ins/outs, but it all seems fine and intact. Checked the needle in case I'd accidentally pushed it on harder and it seems ok, moves freely back to zero if I move it around the dial.

So, I am assuming a similar issue still exists with the contacts to the MC. Hopefully, I can find the problem without stuffing everything else up!

Oh, and a DOH! moment - I realized this morning I could use the OBD2 connection to check the signals from the ECU. By bringing up the instruments on OBDLink on my phone, I could see that even when the dash tacho was bouncing (before I found the alleged fix) the signal to my phone was fine, leading me to believe there was no problem with the ECU. Correct me if that's a wrong conclusion ...

So, need to fix the speedo, but feeling somewhat happier right now! ... :wink:

Roadrunner · Postby **Roadrunner** » Tue Aug 25, 2015 8:25 pm

Forgot all about the OBD!! :oops:

I used that to check when my speedo died (OBD showed Zero so I knew it was the sensor, not the dash)
Good to see it's been narrowed down though! Worst case scenario the dash cluster is cheap(ish) to replace second hand

KevGoat · Postby **KevGoat** » Tue Aug 25, 2015 9:39 pm

Yeah was a bit of a blonde moment :oops:

I'm still going to grab another SE cluster when I can find one - just in case. Okibi put me on to one, but it looks like I just missed out. Pm'd Lou (Boyracer) a few days ago with a query if he had one. Haven't heard back from him as yet but he's always pretty busy, so I'm always patient with his replies.

I found a download with resistance values to check each of the instruments. The speedo reads correctly. With all the mucking around, I could have jolted the needle or something and accidentally moved it, but it feels very tight on it's post so seems unlikely. Not sure if completely unplugging the ECU would have changed anything? I'll see what that handy OBD thingy tells me tomorrow :lol:

KevGoat · Postby **KevGoat** » Wed Aug 26, 2015 3:42 pm

Hmmm .... not fixed :frown:

more complications ...

Took the car for a decent drive. The speedo is absolutely whacko :lol:

But it looks like it is just an innocent bystander of what else is going on.

Everything (apart from speedo) looked good for about 15 minutes. Through 50/60/80/100 zones and over bumpy sections of road and all kept working. Then, the tacho bounced. A few minutes later the odo, tacho, speedo all went dead. Turned up a small side road, pulled over and as I got down to about 2/3 kph the odo was fluctuating on/off, then as soon as came to a complete stand still everything came straight back on! Pulled out on to the road again and as soon as I started pulling away it all went off. Pulled over again and stopped, all back on! Rev the car while stationary and no problems, pull away, goes off! It's like as soon as any speed signal goes through, something bad happens.

Also noticed that the OBD dash speedo has incorrect fluctuations when the odo goes off and the tacho is bouncing. I believe (correct me if I'm wrong) that the signal from the speed sensor goes directly to the dash MC and is converted there, then travels to the speedo and the ECU, so I don't believe this is an ECU issue. I reckon th espeedo and ECU are only reacting to the same fault somewhere else. From what I've read, there's only a speed input at the ECU, not an output, so if the dash MC is wrong, so will the ECU be.

I felt around the back of the dash when I pulled over and on one side of the odo it was real hot, more than I'd have expected, especially as the other side was cool. So, maybe there's something in there shorting, or maybe there's too much voltage coming through to something in there. If the speed sensor is shorting, or part of it's cable somewhere, would that cause something in there to get hot? I removed the odo/MC and couldn't see anything wrong on the circuit board. The plug contacts all look good. But it could still be a failing component.

Engine revs fine, all engine functions are good, no OBD error codes are showing up, coolant gauge works fine, fuel gauge fine, all dash warning lights working normally and all lighting is fine.

First, once everything is cooled down, I'll try unplugging the speed sensor and see if all the other stuff works without that plugged in. If the fault still occurs and the odo still gets hot with the actual sensor unplugged, it at least rules the actual sensor out. Move along the cables to the MC from there ...

Ah well ... suppose I'd better back to it :roll:

KevGoat · Postby **KevGoat** » Wed Aug 26, 2015 6:46 pm

OK ... unplugged the speed sensor. The plug and connections looked real clean.

Took her for an extended drive - first around my normal "test track" and then followed the same route as this morning.

Not a problem!!

So I have at least isolated it down to an issue with the sensor, or somewhere along the speed signal circuit including the dash MC. (Damn! just realised I forgot to check the heat of the odo when I got back :oops:

)

I had a look along the cable, well as much as I could anyway. Most of it's within a loom and all of that looked healthy. Tried to follow it under the dash but I'd need to remove the seat to do a better job.

I've read that the speed sensors not only have problems with the two solder joints - which doesn't cause a short, just speedo error - but apparently are known to short inside the unit. I have no idea how to check out the circuit board in the MC, so first I'll google how to test the sensor for a short with the multimeter.

Getting there :roll:

JBT · Postby **JBT** » Wed Aug 26, 2015 6:56 pm

At least you're narrowing it down now. Weird problem.

bootz · Postby **bootz** » Wed Aug 26, 2015 7:06 pm

Let me know if you need a speed sensor, I have one from a 3.6 diff spare.

KevGoat · Postby **KevGoat** » Wed Aug 26, 2015 8:09 pm

Hey bootz, thanks, I might take you up on that. PM sent.

KevGoat · Postby **KevGoat** » Wed Aug 26, 2015 8:16 pm

My thinking's currently more weighted towards a fault in the MC, but if I can change out the sensor, it either fixes or rules it out. I'm pretty certain it's one or the other. If it was a straight short along the cable, I would have thought I'd be blowing fuses. I've tested the speedo instrument and that reads fine, so there's not much else in that circuit I can think of that it can be!

davekmoore · Postby **davekmoore** » Wed Aug 26, 2015 9:18 pm

Next time you're looking at the ECU take a look up in the passenger footwell at how Mazda secured the drain pipe for the a/c condenser. It's just pushed on. So take the opportunity to put a clamp on it to avoid your passenger kicking it off - especially likely if the ECU cover is ever left off.

KevGoat · Postby **KevGoat** » Wed Aug 26, 2015 9:54 pm

That's a good point davekmoore, I'll take a look at that and fix it.

bootz ... I'll definitely take that sensor. Just sent another pm, wanted to check on something before I send payment.

KevGoat · Postby **KevGoat** » Mon Sep 07, 2015 6:46 pm

OK, managed to get hold of a replacement SE Instrument Panel and a Speed Sensor.

I don't want to get ahead of myself and jinx things, but long story short, it appears to have been the instrument microcomputer.

I've actually ended up with both the sensor and the panel changed over, but by swapping the MC from my panel into the new one, it created the same problems as it did in my panel, so I think I changed the sensor for nothing. Whatever, it's working and I ain't touching it!

Had time to take the car for about a 1/2 hour drive at various speeds and revs and it all looks good, but this has been a strange issue and until I've done a good few days of driving it I'm not going to be 100% satisfied.

I'm somewhat peeved that it was the MC and not the sensor though, as now I've gone from this ...

To this ...

What is it with me and my MX-5's odo's ????!!!!