ruh roh... another intermittent ignition puzzle, TCI type

Well... I THOUGHT I had it solved. :banghead: Same ole same ole... when it runs, it starts and runs flawlessly, and has never cut out while running. But, when it sits, and I go try it... NADA. I put the spark tester on it, and no spark.

I had removed, cleaned, ohm-tested and continuity-tested every connection in the start circuit, and it all tested fine. Then it ran fine. Now it doesn't spark.

I am pretty well convinced that there is an intermittent something going on in the TCI, and just need another, or a "known good", to confirm.

Further news at eleven.....

Yes, when it craps out, immediately test for spark. If you are confident that you are testing for spark correctly, and that you have none, then it is almost always your coil or your TCI. Those are the components that can work fine cold, but crap out when hot. If you went through the whole test procedure in that TCI trouble-shooting manual, the next step is to replace with a known good TCI.
If you put a test light on the red wire to the coil it might dim if the TCI is working. But not sure of the duty cycle so maybe not. If you had something that would beep if it went to ground you could test it with that. If you had an oscilloscope you'd be all set. I don't know why more people don't own scopes.

If I had to guess, I would guess coil. My experience with coils is I've had two bad ones. One got progressively worse over a week or two until it would foul plugs in just a few miles. The other was discovered dead one morning and never worked again. Both coils measured fine.

A recent experience; I had a symptom where the bike would occasionally stall for just an instant and start up again on its own. That progressed to the point where it stop and I had to pull over and start it again. Which further progressed to the point where I almost thought I was going to walk it back home 3/4 of a mile! But it started and when I got back I pulled it apart and discovered a cracked cap on a wire at the coil that had allowed the wire vibrate out of the coil to the point where there wasn't good contact with the coil. That might not be your problem, but it's something to keep in mind. I had another cap split when I was putting it on. They get a bit fragile.
well you could be right about an intermittant Tci fault but I still feel Hmusket might be right and it could be an earthing issue.

Electrical schematics often don't show all the key earths so its easy to overlook them. For example the spark plugs require a good earth and new paint on the frame and engine/transmission mounting brackets could prevent the plugs earthing properly .

The same might go for the Tci box which I believe requires an earth to chassis through the fixing points ? someone will no doubt confirm or not.

Try connecting a jump lead from the engine to a good earth point on the frame (or battery )and see if it improves the spark
I am positive that my grounding and so forth is good... I have done marine and home wiring prefessionally at various times in my life, and rewired several other bikes (BMW airheads with both points and electronic ignition)... I had this down to bare frame, grounds for every function are fine, and I always add "extra" grounds when in doubt (dual grounds from the handlebar controls, ground from engine (head) direct to battery, plus engine to frame, etc.

The TCI is not grounded through the case, as it is a plastic case. The pins ground the TCI, black being ground. I have continuity-tested every freeking thing, I have ohm-tested the pickups and the coil from each back to the TCI pins, etc. Continuity is good, when tested....but obviously something ain't right!

Coil intermittents are, in my experience, usually a problem with heat, as the windings may short when the coil and ambients are hot, then ok when cold. This one is either A-OK, or NFG, nothing in between. And the NFG has always been when I go to start it cold, after ssitting. Mind you, my riding has been limited to the range of the gas in my funnel-to-carbs "tank", though I have run/idled/blipped to a good hot temperature while using a blower for airflow over the engine, though still not "hot" by motorcycle standards!)


Sorry to see you still have the problem, I know its probably frustrating the devil out of you. I'll assume that you double checked all your connections and they're tight. The only other thing I can think of is that there is a small break in some wire you can't see. A few years back I bought a CX500 that started and ran well, but would sometimes cut out when I turned right. It took a bit but I traced it to a broken wire in the harness right by the fork stem. gggGary suggested closely looking at the wiring coming from the pick up, that might be a good starting point. I also have a couple spare TCIs, so if you want to borrow one from an '81 for testing, let me know.
A meter is meaningless with a coil. At least the kind most commonly available for tci xs650s. You have to test it by substitution as well.
not sure what you mean by "meter is meaningless with a coil"...?

You can test for ohms through the two leads (supposed to be 2.5 ohms iirc?) which at least gives a pass/fail.

There is also a test that can be done by jumping one of the TCI leads to ground, which should cause the coil to spark.... and that test is for the NEXT time I tear into it! It tests the trigger function of the pickups on the alternator, and simulates one "sweep" so that the coil should fire.

The link posted above has some pretty trick tests to help isolate the problem, and at this point I am thinking TCI...... I have a known-good one coming from gggGary, so that will also be in the arsenal in the near future.....and the coil that is on the bike has been proven to work.
Last edited:
...another step in the link is to measure voltage at the coil... one of the leads (red and white) should have constant battery voltage (12v), but the other (orange) wire is the trigger, and it can be tested easily with an old analog VOM, set to a low-voltage scale, and it makes the needle deflect. Not so easy on my digital voltmeter, but there is something squiggling on the screen, which I have interpreted to mean that the trigger is triggering.

The above tests were all done when the bike was starting and running, not now with the no-run situtation... I just have to get back in there.... again.....

^What I meant by a meter is a meaningless test with an igntion coil is that the two TCI coils that crapped out on me measured fine when I threw them in the bin. I think there was a high voltage path in the secondary that won't show up using a meter since it tests using a low voltage. One of them had a big ridge around the middle like it thought about exploding, but it measured fine. I desribed the symptoms above.

The book looks interesting and I don't think I've seen it before, but unfortunately all I can see is the thumbs. I'll have to wait until I get to a better internet connection that will do the individual pages. Here's a fixed link to it:
Hey guys I'm new here, what up... I'm sure you've thought of this, and I'll probably sound stupid for saying it, but hell... have you made sure that you spark plug wires are making good contact with your boots? I've got an '81s that I had a hell of a time starting, but every once in a while she'd fire right up and run well until all of a sudden she'd get all lurchy and just give out. Then I could kick and kick until my leg fell off and NOTHING. This went on for about a week- I finally decided it must be a carb issue (why not?) and as I was pulling the fuel line off the petcock to dig into them, my arm brushed the left cylinder plug wire, which fell out of the boot. I reconnected it nice and snug, re-fit the fuel line, kicked her once and she's run like a champ ever since. I felt like an idiot, but it really was that simple.

I'm not saying that it's likely that's your problem, but sometimes it's the simplest things that are the hardest to figure out... I've built embarked on enough disaster-builds to know that firsthand.
I wrote the same story above but the wire was coming lose from the coil :)
I've got a no-spark situation as well on an '83 special. I swapped in a known good coil and still no spark. I've taken the TCI unit apart and it looked like one of the zener diodes was burned, but I'm about to go through the testing procedure outline in the link and see what I've got. Bah humbug.
Well, this time it IS fixed.... as I suspected, the problem was in the TCI box. Thank you gggGary for the replacement! Well worth the $$....

While waiting for a "known-good" TCI to arrive from gggGary, I fiddled and futsed with everything, re-tested electricals etc, and got the bike started and she ran like a champ. Next day, NADA.

The day the known-good TCI from gggGary arrived, I, just for giggles, tried to start the bike with the old TCI. Nothing. I grabbed the new one, plugged it in to the harness, and vrrrroooommm... started right up. So, now I KNOW that the problem was indeed in the TCI. I have not yet removed it and installed the new one, but just letting it hang there, it works.... every time..... have tried three/four times a day for the last bunch of days, and it works.... every time! :thumbsup:

I took a lot of pictures with probes, checks etc, and will write up the procedure in another thread. Guests for the weekend, so the spare/bike/computer time is at a minimum, but after that....:bike:


Thanks again for all the suggestions... all legitimate, but I was pretty sure, and getting very sure, that the problem was what it did end up being....

Last edited: