So, here I am, two weeks from Sevenstock, just got all my beautification done... and I've got a strange electrical issue that just HAD to rear it's head

Situation: Stock 80 GS with full emissions and rats nest, all of which has been fully functional.

Over the course of the last week, I:
1)pulled the dizzy to clean/lube/polish it and paint the external bits
2)changed out my Type 1 RacingBeat Gauge Adapter Block on the filter pedestal for a Type 2 block (just an aesthetic change, really)
3)installed fog lights, which are relay-fed (primary straight from the battery, relay is triggered off the hazard lights fuse - works fine)
4)made a simple harness repair for the pigtail from the block temp sensor; insulation had hardened and split, so I solder-spliced a new pigtail on.

First restart tonight, and since I'd pulled the dizzy I knew I needed to re-time.

I pull out the choke, car started normally after about 3 seconds cranking, warms up, comes down off choke to idle. I set up my timing light.

Long story short, I have the following symptoms:

1) At idle, Trailing ignition is showing massive crosstalk from Leading; so much so that at first I thought I was on the wrong plug wire. Close observation shows that trailing signal is firing on correct timing about one pulse out of three (the timing mark goes yellow-yellow-red-yellow-red-yellow-yellow-etc... no pattern really.) 80's have been known to be prone to crosstalk because of the remote-ignitor design, but up til now I've never had a problem. I verified that the dizzy is wired correctly and the ignitor box ground is clean and tight.

2) At idle, my aftermarket electrical guages read normally. Anywhere above about 1500 RPM accelerating, the needles immediately peg below the zero lines. Grounding the sender shells with a jumper to the chassisprevents this symptom, which seems to maybe indicate an engine grounding fault, but one that is RPM dependent (WTF?).
3) Engine behavior decelerating is poor; multiple backfires from even light rpm down to idle. hasn't been an issue with this car before. Doesn't "feel" right, which I'm sure has to do with the trailing being out of whack. Acceleration seems smooth, though - no misfires etc. Idle is as smooth as usual.

What I've checked so far:

1)checked ground at strut tower (good) and metered ground from engine to chassis; zero measuable resistance while off.

2)Verified that the ignitor to dizzy cable is properly connected, the condenser at the dizzy is fine, and the black-with-white-strpe positive coil wire is good through to the condenser.

3)guage senders are fine, not shorted, and the oil pedestal base to strut-tower ground point shows zero ohms with engine off. Guage sender bodies to ground read zero ohms with engine off.

had to pack it in at that point for the night.

I need some suggestions as to where to look; this one's got me running in circles.

Any inspirations from our resident SA guys?

 

No - but a boatload of sympathy and kudos for heroic attempts at working it thru Glen

you mentioned the ignitor crosstalk (which I first read about on Racing Beat site) - do you have/would it be worth swapping in another ignitor (like they grow on trees...)?

Stu Aull
80 GS
Alaska

 

I just so happen to have a pair of J-105s if ya need them. Those can be hard to come by as they are only on the '80 model.

Did you try moving the pickup on the plug wire? I know that even on FBs it is common to pickup the leading signal even when clipped on trailing. Sometimes moving the pickup closer to the dizzy cap helps. Might also be worth removing the ignitors and making sure all connection are clean including the screws which go through the ignitors (helps to ground them). Put some new heatsink compound on the back of the ignitors when reinstalling.

For you aftermarket senders, are they electric based (assume so)? Where are you getting power from? Did you use teflon tape or anything like that to seal the threads at the sender? Funky electrical problems that are RPM dependant can also be caused by an alternator that is on its way south or that has poor grounding (make sure that the ears of the alt and where is mounts to are clean). If all else fails, try to run a new ground between the block and the chassis.

Also, check fuses. It might be that your ECU is not getting juice or has a poor ground. This could be the cause of the backfires. Might be worth doing some testing at the ECU if all the fuses look okay. I figure that the decel behavior is either the signal side (ECU) or receiving side (solenoid unplugged, problem with one of the othe remission devices, etc.).

 

I would really focus on what you recently worked on if it was all
working flawlessly before. Like the dizzy, the pedestal block change,
the pigtail splice, etc. It had to be some thing to do with that work.

It could be when messing with the dizzy that you pulled on the
wires that go from the dizzy over to the ignitors and now either
internally or externally they are shorting causing the cross talk.
They follow a torturous path down along the block, under the AC,
to the strut tower and into the igniter box. Try wiggling those around
some and see with the timing light if the behavior changes.

Disconnect the fog lights to make sure thats not the issue.

You have two issues and multiple unknowns to debug here so focus
on either the timing issue or the guage issue and start eliminating/nullifying
things you changed recently until you get a change.

 

I will troubleshoot it by installing a different dizzy (81+) and see how it goes. this is to eliminate one of the unknown

 

yeah if it was fine before chances are it was something you did

 

Wait a minute. So, in order to turn on your fog lights, you have to turn on the hazard flashers?

Check the engine ground near the starter, and the power cable as well.

Good luck...


.

 

OK, bunch of good suggestions here. I think I have a primary and fallback plan mapped out:

Primary:

1) isolate out all new work (foglights, mainly, but disconnect gauges too; easy enough)

2) careful visual of all work areas, look for disturbed connections or possible rub-shorts.

3) reclean all grounds (incl. engine ground @ starter) & verify ignitor box wiring & ECU connector, fuse, and ground. Signals check of ECU if necessary (PITA to get to)

4) meter pickup coil wiring, look for possible pinches/shorts

5) verify ignitor grounds and heatsinking, swap in spare ignitors & spare ignitor box for test purposes if necessary. Last ditch, swap to old pickup coils and re-test, though they were pretty old when retired.

FALLBACK;

Upgrade to 81 dizzy. Fallback because it presents smog-test complications, and because I don't have parts at hand.


Answers to specific questions:

Yep; on the list of things to try, thanks!

I have a few known-good spare J-105's, thanks. I even have a known-good (but ugly) spare ignitor box, and my old pickup coil set which has iffy bearings but is otherwise workable. So I'm pretty well-fixed in the swap department, for testing.

Yep; no difference even with the pickup right at the cap.

This was last done about 9 months ago (only about 1000 road miles) when I was chasing that phantom miss. Easy enough to check again; on the list.

Yep, electrical gauges. I installed them several years back; I don't recall exactly where I tapped power from (blush) but it's a switched accessory lead, and hasn't previously been an issue. Fairly sure it's the same feed that runs the power for the stereo. The gauges have worked reliably for ages, so clearly something has changed.

The alt was tested good less than 1000 miles ago; I will recheck it's grounding and output, though. I have a shunt meter, so I can test actual amperage out.

I did use teflon tape (NPT threads, so you gotta seal somehow) but verified that the gauge bodies are grounded properly to the oil pedestal.

I'm going to test with a block ground lead tonight.

On the list, thanks.

Yep, either that or something that got pinched/stressed in the process, while working in the general area. On the list.

I was pretty careful, but definitely bears re-examination. I'm hoping I can feed the cable without having to undo the AC bracket again; it's a PITA. I seem to remember the connector won't fit thru, though. Frankly, the whole AC design on the SAs was a real hackjob by Mazda. Most of it is even fitted with SAE instead of metric hardware!

On the list.


Yeah, though I have a sneaking suspicion they will turn out to be related. The ignition issue (and the decel misbehavior) is the higher priority; the gauge prob I can work around by re-grounding, though that doesn't really explain anything & I hate leaving problems unanalyzed.

The one thing I don't have handy, sadly, is a newer dizzy and harness for same; I've been trying to keep the engine compartment as SA-complete as possible, and so have never bought one. I'm saving this one as a last-resort due to the need for parts, but it's a worthy idea.

No; Power is pulled from the fuse that feeds the hazard lights, AC clutch, and the power mirror circuit. Upstream from the hazard switch.

On the list, thanks.


Thanks for all the insights, guys. Keep 'em coming; I'm all over this problem as soon as I get home from work tonight!

Discussing it here really helps focus the mind.

 

do the TDC via flywheel trick. I just have this feeling.

 

I didn't have the pulleys off, and I set the e-shaft to TDC (yellow mark) before I pulled the dizzy. Was still there when it went back in 5 days later...?

 

just do it. you got nothing to loose.

 

If it helps, I've switched from teflon tape to the teflon paste. Seems to give a better ground.

 

OK, here's the results of tonight's efforts; I don't have it solved, but I think I have it cornered.

First, there are two separate and distinct issues, and they don't appear to have anything to do with one another. They also appear to have nothing to do with any of my recent additions, as I isolated out all the new wiring and no change was seen.

1) The guage fault: It's not a grounding problem in the traditional sense, and although it's related to engine RPMs, that's a byproduct: I believe it's actually related to oil pressure.
2) The ignition issue (trailing firing on leading signal a lot)
More tomorrow; sleep now.

 

Still sounds like an overall grounding issue to me. Maybe (temporarily) run a wire from the engine to the - on the batter just for ***** and giggles?

Sometimes, with a bad ground (or even +) connection, it won't exhibit any symptoms until higher rpms when things start trying to draw (or produce) larger amounts of juice. A trickle can get through, but a large demand just starves out everything that's requesting it...

I dunno, but that's my gut feeling. Sleep well.

 

Looks like I left it off the details, but I did try this; I ran an alligator clip from the strut-tower ground to the condenser ground-point on the distributor. No change in the ignition issue. I also tested the condenser, and it's right on the money; 0.49uF.

Funny things about the 80 build; the physical structure of the distributor isn't electrically even part of the ground circuit, except for being the ground-point for the ignition coil condenser. The coil positive is routed all the way over to the dizzy just so it can connect to the condenser, which is physically grounded by its mounting tab.

This looks to be a physical design holdover from the 79 points ignition; it makes no electrical sense to route the positive over there just for suppression. Positive voltage is not used on the 80 dizzy at all. But the condenser was physically there for the points ignition, so they left it there for the 80 design.

The pickup coils are fully "floating;" neither side of the pickup coil connection is grounded at all, and even the cable shield does not ground at the distributor end. All electrical grounding for the dizzy is done via the ignitor box grounds - - except for that suppression condenser.

Another strangeness I noticed years ago; the six-pin connector that hooks the pickup coil set to the ignitor box does not appear on the 1980 factory wiring diagram at all. Left completely off the diagram for no apparent reason, as are the cable shields (which are two of the 6 pins). Several of the referenced connectors (like B-24 and B-26) are not shown in the right place on the locator diagram, either, and the ignitor box is missing entirely from the locator diagram.

 

interesting, if the gauges all do weird things in relation to RPM, and you're having timing/ignition problems, i'd suspect the ignitor/ignitor box. rfi?

 

OK, all is fixed as of about 2pm (working on other stuff since, plus a nice long test drive), and here's the score:

Ignition problem: trailing ignitor was on it's way out.

Swapped to my spare ignitor box (which had two known-good ignitors on it) and flaw went away.

Swapped the suspect trailing ignitor onto it - - problem reapppeared.

Put good ignitor from spare box onto original box - - problem stayed gone.

That was the easy one, once I was sure (as of last night) that it had nothing to do with the other problem, or with my recent mods, it was straightforward.

Now, for the really interesting one: the guage problem.

I was right: the problem was being caused by loss of sender grounds as oil pressure rose.

But I was also wrong: It had nothing to do with the teflon tape.

Follow along, here:

I took the adapter off (easy to do with the new design) and put it on the bench.

Measuring resistance between the two senders, we can see that both have solid, zero-resistance connection through the guage block:


On a hunch, I decided to test resistance on the "skin" of the block:


That's right, the blue anodized coating applied by Racing Beat is NON-conductive

Looking at the part of the unit that bolts to the filter pedestal, you can see that the coating is intact... except for this TINY scratch made by the center fitting when it was originally torqued.


Since the bottom of the adapter is also coated (and seals with an o-ring and smurf snot) the only contact is through the center fitting... and it's only point of contact was that scratch. Both guages were grounding through the tiny scratch's contact with the center fitting, and which only made contact under pressure.


But when oil pressure rose in the filter housing as the engine come up off idle, the fitting is drawn away from the adapter ever so slightly, and the tenuous ground contact is lost.


The ultimate solution, of course, was to polish the coating off of the metal all around the contact area where the center fitting lands:


Presto... put it all back together, and no problems at all.

Since I'd bought some pipe sealant anyway, I went ahead and re-assembled with sealant insead of teflon tape. Looking sharp, now.

Fun, huh?

 

Uh, DD:
Fun is blasting thru a tunnel on my Ducati and listening to its bark reverb off the walls ....
Fun is basking in the nods from the ricer-boys when I pull up in the Ol' Skule SA
THIS is not fun. You are insane. A genius. But, insane
Congrats on solving 2 real head scratchers...
knock 'em dead at 7stock my friend

Stu Aull
80GS
Alaska

 

If that happened to my FB i'd be craying like a little school girl while sucking my thumb. I hate electrical problems.

 

wow thats a fun one!

 

Glad you got her figured out DD. Have fun at SS.