Another Tri-spark Failure

I've been following the threads about Tri-spark failures for a few years now, and finally had one of my own to report. I installed a Tri-spark on the MK3 Commando my grandson and I built for him back in 2015. Since then he's probably put 5,000 miles or so on it with no ignition issues. But, last week he rode it to a friend's place to meet up for a group ride, and it refused to even try to start when it was time to leave. The starter spun it over just fine, but there was no spark. The ignition unit in the timing side cover was very hot to the touch. We ended up hauling it home in the back of a truck. When we got it home, I tried it, and it started up just fine. I suspect that it just got overheated enough that it quit working, but was ok after cooling off. But I can't be sure of that. I tried heating it with a hot air gun, and got it hot enough that some of the plastic coating started to peel, but it kept working, so I can't be sure that was the problem.

Anyhow, I just wanted to add another bit of data to the discussion of Tri-spark reliability. I have another Tri-spark unit that I was planning to use on my own MK3, so we could do as Jim Comstock suggested to me and just carry it as a spare, but I'm not that excited about having to swap them in the middle of a ride. To me, the risk of a failure in a really inconvenient location isn't worth taking. So, I've bought, and am in the process of installing a Pazon Altair. I haven't seen any complaints of them failing, and I like the idea of having the electronics in a cooler location.

Ken

Ken, is that the same Pazon Altair being discussed here:
https://www.accessnorton.com/Norton...ning-there-then-gone.28973/page-3#post-443250

It would be awsome for the vintage of that unit to be known. If it is just another failure of an early one with bad spec transisitor or unsatisfactory heat sinking potting then that is not new. If the failure is a later thermally improved version than that's big news. Virtually every electronics failure is caused by improper thermal design. Its usually about heat control....

ken welcome to the failspark shake and bake club. after my third one failed i sold my spare and went to a boyer micro digital. i did have a pazon analog on for a while between and was not happy with the inconsistent idle. the last one that failed was just like yours, ran fine till i stopped than would not restart the one before that ran fine than ran like crap on the restart and the first one just started to run like crap than stopped altogether.

dave of the 3 that failed 1 was the old style and the last 2 were the late version

This is a theoretical question. Would a peltier placed on the points cover keep the tri-spark or which ever unit cool enough to prevent a failure? I have no idea how much current one draws or what voltage is needed or how to control it. Just currious.
Jim

Over 16K miles on the improved version and no issues. If your tri spark stops, then later works it is highly unlikely the tri spark unit. Usually electronics fail and stay failed, they don't suddenly fix themselves.

The fact you took a hot air gun to it and it remained running says the unit was not at fault, and thanks for doing the test.

I had failures with the early TriSparks, and yes when they failed and left me stranded they restarted just fine after cooling down.
Matt at CNW was just great to work with during that time period and now I have a perfectly reliable Trispark as of 2014.

On to your problem, it could be as simple as an intermittent ground, at a connection or a fractured wire, inside the hated kill switch?
It could be a live wire intermittent at the coil or anywhere, I would not assume it is the Trispark, but how old of a unit is it?

oldbeezer said:

This is a theoretical question. Would a peltier placed on the points cover keep the tri-spark or which ever unit cool enough to prevent a failure? I have no idea how much current one draws or what voltage is needed or how to control it. Just currious.
Jim

Click to expand...

I'd doubt it could deal with the amount of heat coming off the hot engine. Better to put a finned cover plate on, perhaps even with some kind of air duct venting?

Fast Eddie said:

Ken, is that the same Pazon Altair being discussed here:
https://www.accessnorton.com/Norton...ning-there-then-gone.28973/page-3#post-443250

Click to expand...

I only saw the Pazon mentioned once in that thread. It seems to be mostly about the Power Arc ignition, with some discussion of the Boyer and Rita systems. But to answer your question, yes, the Pazon Altair is what I now have.

Ken

lcrken said:

I only saw the Pazon mentioned once in that thread. It seems to be mostly about the Power Arc ignition, with some discussion of the Boyer and Rita systems. But to answer your question, yes, the Pazon Altair is what I now have.

Ken

Click to expand...

Forgive me Ken, you’re right, it’s the power arc that’s causing Joe some issues it seems.

dynodave said:

It would be awsome for the vintage of that unit to be known. If it is just another failure of an early one with bad spec transisitor or unsatisfactory heat sinking potting then that is not new. If the failure is a later thermally improved version than that's big news. Virtually every electronics failure is caused by improper thermal design. Its usually about heat control....

Click to expand...

I bought it new direct from Tri-spark on 8 February 2015. I don't know if that is considered an early one or a later improved version.

Ken

1up3down said:

I had failures with the early TriSparks, and yes when they failed and left me stranded they restarted just fine after cooling down.
Matt at CNW was just great to work with during that time period and now I have a perfectly reliable Trispark as of 2014.

On to your problem, it could be as simple as an intermittent ground, at a connection or a fractured wire, inside the hated kill switch?
It could be a live wire intermittent at the coil or anywhere, I would not assume it is the Trispark, but how old of a unit is it?

Click to expand...

I don't think it's a bad ground or kill switch problem. When it wouldn't start, my friend checked several things (I wasn't there at the time), including power to the unit. It had power to the unit, and the green test light on the unit was lit. He also checked the coil with an ohm-meter, and it was good. But there was no spark. Because it is working again, I don't think there is much else to test, unless I was willing to do test rides until it happened again, and then try to do some quick troubleshooting while it was still hot. I'm just not interested in doing that. Better to use something that has a little better reputation for reliability.

I was initially drawn to the Tri-spark because of it's simplicity and it's idle stabilization design. From my previous life as an electronics engineer I knew that it was possible to design solid state circuits to tolerate very high temperatures with reliability, so I didn't worry much about the fact that the electronics were in the crankcase cover. Now I'm not quite so confident that the Tri-spark is up to the higher temperatures it sees. It is considerably more expensive to use mil-spec components that operate at elevated temperatures, and it may not have been possible to do so at a price Norton owners would be willing to pay. To be fair, we did get 4+ years of good use out of it.

Ken

Do we have a timeline for when design changes were made for the issues Dave brought up (bad transistors and poor potting). I bought this one in February of 2015, but I also have one I bought in June of 2016, which I have been planning to use on my own MK3. It would be interesting to know if one, or both, are the old design, or the "improved" design.

Ken

I've had one Trispark failure after 10 years and it was the older design. My own fault though - I allowed it to become loose (mounting screws) so it lost the ability to transfer heat to the timing cover. Replaced it with a new design spare that I bought from Bill. For anybody mounting any kind of IE inside the points cavity, I recommend using some heatsink compound like Dow Corning 340 between the back of the IE plate and the lip inside the timing cover to insure the best possible heat transfer.

maylar said:

For anybody mounting any kind of IE inside the points cavity, I recommend using some heatsink compound like Dow Corning 340 between the back of the IE plate and the lip inside the timing cover to insure the best possible heat transfer.

Click to expand...

Dave, you sure about that?

I thought the heat issue was more to do with the theory that the engine heat is being transferred to the Tri Spark. Not the Tri Spark generating too much heat by itself.

If that’s the case then shouldn’t we be insulating the Tri Spark from the engine rather than trying to increase heat transfer??

Fast Eddie said:

Dave, you sure about that?

I thought the heat issue was more to do with the theory that the engine heat is being transferred to the Tri Spark. Not the Tri Spark generating too much heat by itself.

If that’s the case then shouldn’t we be insulating the Tri Spark from the engine rather than trying to increase heat transfer??

Click to expand...

No. The heat has to go somewhere, and the only path to the outside air is through the aluminum timing cover. Otherwise you create an oven where the electronics' heat is trapped inside that cavity. It's not possible for the timing cover to be hotter than the EI module.

Running with the points cover off would be best, but then we expose the cam end and EI rotor to the elements.

I have an original triSpark trouble free since I bought/installed it in '08. I guess I must have the only good one.

I have to agree with a previous post - if it's now working then I doubt the Trispark was the problem. I do a lot of electronic repair/troubleshooting and never saw a bad transistor, diode, or any other electronic component fix itself.

I'd suspect the problem to be an electrical connection issue.

maylar said:

No. The heat has to go somewhere, and the only path to the outside air is through the aluminum timing cover. Otherwise you create an oven where the electronics' heat is trapped inside that cavity. It's not possible for the timing cover to be hotter than the EI module.

Running with the points cover off would be best, but then we expose the cam end and EI rotor to the elements.

Click to expand...

Hmmm, vented / removed cover I agree with as, as you say, any heat will escape rather than build up. I’m thinking a small slot in the front and another in the back to actually allow / encourage air to flow through might be ideal.

But the heat transfer bit I’m still not seeing. If the engine is hotter than the stator plate then the heat will transfer from the engine to the stator plate. Thus heating it up.

Notwithstanding any of the above, I’m personally not yet convinced heat in the cover is the issue. Lots of delicate electronics get mounted in some very hot environments on modern vehicles without failing, inc ‘black boxes’ ign coils (mounted directly onto spark plugs and buried inside the head), direct injection injectors mounted in the heads, numerous sensors, etc..

lcrken said:

.. To me, the risk of a failure in a really inconvenient location isn't worth taking.
Ken

Click to expand...

Like in the middle of the 14 km Arlbergtunnel without an emergency lane?


I had an early one fail after 5-6000 km in the same way as often described:
Engine dies, but will restart after 5 min. cooling down. Dies again after 10 -15 km, etc..
Before installing, I plotted the advance curve (can still be found here..)
Weary about the temperature issue, I added cooling fins, made air vents in the cover, placed a (kind of) heat shield between rotor and module .. still failed.

I drilled a second 1/8 “ weep hole somewhere on the bottom third of my points cover on the theory that a second hole would allow a bit more air to circulate.

I also use a gasket on the theory that it somewhat isolates the cover from heat transfer and thereby reduces heat reflected from the cover toward the TS unit

Compared to Ludwig’s precautions . . . Likely pointless (pun intended).