GS550e "interesting" ignition problem - dies when coils are hot

[ajh]

I have a very frustrating problem with my 1982 GS550ez (UK bike) Im slowly diagnosing. I wonder if anyone can shed some light on the problem

Its been runing fine for about a year with points in it a PO had retro fitted but started running on 2 cyllinders this week after I tweaked the timing on one side.
Complete coincidence, one of the coil connections went intermittently bad. I always planned ot fit a Dyna S as soon as there was any trouble with the ignition and assumed the points/condensor were bad because I had messed around with them setting timing and they looked quite burnt up. So fitted Dyna and of course the problem was still there! Quickly discovered the bad coil connection and fixed that.
Now the bike goes fine as long as I keep revs up. It will idle ok until warmed up and then it just suddenly dies and then wont start again on the starter button until its cooled down. It does have a spark at the plugs but to me it looks really weak almost struggling to see it. The last time it happened I tried bump starting it and it fired right up. So I suspect there is a voltage drop when the starter is used combined with a weak spark.
Got it back home and started some testing

I have 12.2 volts at the coil + feeds with coils disconnected but it drops to 9.4 when I connect the coils? Same on both coils
I have 5ohms resistance on the primary windings and 25K ohms on the secondary. (measured HT cap to HT cap) Same on both coils

By the time I took these measurements, the bike/coils had cooled down so the bike started fine on the button. Also as the measurements are the same on both coils I'm doubtful as to whether they have both gone bad but I cant see what else can be the problem

Anybody provide feedback or help on this issue please?

Cheers!

[Gorminrider]

The coils are quite a large draw on the little motorcycle battery. I wouldn't worry about that yet but it does sound like your battery might be on the way out. TIP if your headlight is ON and no switch, DISCONNECT IT. It's another large draw.
Test your battery resting voltage AT the battery AFTER it's been charged and AFTER it's sitting awhile (or turning the key on for a few minutes(not running). It should show 12.5+ on a good battery downwards to 12.2 needing replacement on a 4 cylinder bike.

Until you know what what it isn't on your bike, (crossing off all the possible Root causes, battery, charging, advance mechanism is ok, spark plugs and their caps, BAD CONNECTIONS, bad KILL SWITCH, starter button etc etc) you can't know .. There's a ton of threads on EVERYTHING here so I'd learn to use the search function in this section (=too many ideas until you learn to filter them down)

[ajh]

Thanks for the reply
I think the batteries good, normally over 12.6v but just been doing a lot of cranking and stop start before I did this testing. Keeps its charge for weeks without use. I've been charging it up and its now 13.5v so will re run tests tomorrow and also hopefully run it up to when the problem occurs and test immediately.

Bike charges ok
I've been through all connections and had a constant voltage so thats all ok as well.
Kill switch has been disabled/bypassed
Advance Mech is good and checked with strobe when I fitted the Dyna ignition
plugs are new but leads/caps are a possibility

Weird because it will start on the starter no problem once left for 10 minutes after it stalls. And it will then idle ok until its warmed up and then it just dies again

[GregT]

You've had the tank off and on doing this work and checking. Check that the fuel lines and breathers are all correct - incl the vacuum line to the tap.

Next i would pull the caps off the HT leads and ohm them.

[ajh]

You've had the tank off and on doing this work and checking. Check that the fuel lines and breathers are all correct - incl the vacuum line to the tap.

Next i would pull the caps off the HT leads and ohm them.

Thx Greg, Oh yes very familiar with the fuel line getting kinked Im 99.9% sure its ignition problem as it dies immediately like an on/off switch, no spluttering. And I checked floatbowls were full and had it on prime to take the vacuum out of the mix.
I agree, gonna recheck the resistance of coils without caps and see the difference

[ajh]

So after charging battery I did some further testing today and it seems like my battery may indeed be on the low side. After charging yesterday I left it overnight and it had dropped to about 12.7V and quickly dropped to 12.5 after a couple of tests and cranks. Turning the ignition on drops it across the battery to 12 to 12.2V and 10V at the coils. Cranking then reduces that coil voltage to below 9V
So I can see why it wont start on the starter and it does that thing where it briefly fires just after the starter stops but not enough to get it going.
When its running at about 1500-2k RPM, Ive got 13.8 volts across the battery and over 12 volts at the coils. Charging goes up to above 14 with some more revs. What I havent been able to measure is the voltage at the coils at idle because it dies before the meter settles. My battery cranks the starter over no problem contiually and I can understand why the bike wont start because not enough voltage at the coils when the starter is running but with enough power to cycle the starter even after sitting for weeks. I'm a little bit confused as to why it would cause the ignition to fail on idle when its running. Can anyone explain please?

The only other test I did was to measure the plug caps resistance and they were all 5K ohms which means theyre probably resistor caps and ok. Primary and secondary coils seem consistent at 5 and 15K ohms. And voltage drops are all minimal through all the wiring. So I think Ive checked everything possible. (I even checked the TDC marker with a dial guage to confirm timing is spot on)

I plan to let it charge overnight and see what happens to the battery after a 15hr charge. I'll also try it with a jump from another beefier battery from a CB750 and see if that lights up the spark

[pdqford]

I have a very frustrating problem with my 1982 GS550ez (UK bike) Im slowly diagnosing. I wonder if anyone can shed some light on the problem

Now the bike goes fine as long as I keep revs up. It will idle ok until warmed up and then it just suddenly dies and then wont start again on the starter button until its cooled down.

Cheers!

That sure sounds to me like a primary winding in a coil is going open when the coil gets warm. Can’t imagine both coils going bad at the same time, but I would try to unplug the coil primary leads and ohm out each coil primary while it is too hot to start and see what the resistance is.

[GregT]

Ok, I'd suspect that the Dyna is voltage sensitive - where points aren't. From experience with them on race bikes using total loss battery power, we've run as low as 9.9V measured post race. But restarting at that level wasn't even tried. A fresh battery might just solve all your problems.
Raises the question as to why the PO retrofitted points doesn't it.....

[Gorminrider]

along pdqford's idea...
Maybe put a voltmeter on and see if the bike's or the coils voltage changes just before it dies. i wouldn't worry about the coil voltage not being = to battery voltage but it's CHANGE you are interested in.

.in SIMPLER words, look for unusual voltage drops "across" components... ie: test vD across" a coil's primary.winding= "+" probe to coil's + end and "-"probe to ground... and also ,You have "Dyna"? I'm not familiar with that but if it's on the ground side of the coils you could also try the "-" probe before it gets to the dyna if that's how it's wired.

This might be the best idea:

ALSO check the sensor the dyna uses to time spark. If it's a coil and magnet setup on the crank , those little coils can be trouble too -if they are not telling the ignition to spark, the coils will stay on and not produce a spark and get warm too.

[pdqford]

.in SIMPLER words, look for unusual voltage drops "across" components... ie: test vD across" a coil's primary.winding= "+" probe to coil's + end and "-"probe to ground... and also ,You have "Dyna"? I'm not familiar with that but if it's on the ground side of the coils you could also try the "-" probe before it gets to the dyna if that's how it's wired.[/U][/I]

Just to elaborate, with the voltmeter on the 20V DC scale, and the “+” probe to the coil’s + end, BUT with the “-“ probe to the coil’s negative end, that will measure the voltage dropped across JUST the coil, doesn’t matter if running Dyna or points.

And for us electrical newbies:
To measure a voltage drop across a component, the component needs to be hot (meaning current must be flowing through the component).
The meter will read voltage dropped by the coil primary (higher readings are better).
But if the coil primary goes open when the coil gets warm/hot,
the meter will NOT read zero, it will read the voltage available at the + end of the coil.

Hopefully, this will either blame or clear the coil as the cause of the issue.

[ajh]

Thanks for the replies all.
So I think this is part 3 of this epic series
More testing today and am now satisfied the battery is actually ok. I tried a known good more powerful battery and still had same problem. Charging is also good at 13.8 to 14.5Vs. So Im back to the coils and wiring around them.
I have been testing primary side of coils and the resistance does increase as they get hot. Cold they are 5ohms but after running the bike to when it dies, they then measure 5.8 ohms. I tried 3 times and it happens every time. After 10-15minutes they drop back to 5ohms and the bike starts first time. Now I have to caveat this a bit because of the sensitivity of the Multi meter and the probe connections but so far these findings have been consistent 3 times. Will try a few more times
Voltage at the coils (switch side +) is over 12 volts when running but it does jump around a bit whilst running. Again Im not 100% sure if this is the meter or probe connections or it is showing something wrong but it doesnt seem to influence the engine running. However, Im not happy with the coil connector block connections so next job is to replace them and then run voltage / resistance tests again.

Any thoughts on the primary coil resistance increasing by .8ohm? Doesnt sound like much to me given the rise in temp but i believe 5ohms is the recommended limit for these OEM coils.

[Gorminrider]

Resistance of copper increases with heat. That doesn't mean they're cooked. Not yet. When they're cooked, they will melt,short ,and your (correct!) fuse should/could blow. or, the winding will melt and that's a permanent OFF.I doubt the bike would start after that.

The gist of it is that it's not sounding good at the coils...I think they have stopped sparking but you can test by pulling a plug just when the bike has died.

But if they haven't you are maybe in the wrong place altogether and should goto "fuel"

NOTE on Multimeters on ohms scale- cross the probes, note the reading, subtract this from your subsequent reading of a device. This is "zeroing" the meter and it's important for low ohm readings to get closer to the actual resistance..but low ohms are problematic on cheap meters anyways...The ones with a 9 volt battery in them probably do a better job?

I'm not suprised the meter jumps around a bit when the bike is RUNNING your coils are being switched off and on by the Dyna (or points) and the meter is sampling at a different frequency-also a collapsing field as the coil is going off and on is likely affecting readings-maybe even through the probes themselves!

I'm a little worried by the dyna- i don't know if IT is designed for these coils (could it be overloading them by being too efficient?) or is it functioning correctly? ( my query above per sensors that tell it to)

I'd double check the literature on the Dyna model for sure.

[pdqford]

The gist of it is that it's not sounding good at the coils...I think they have stopped sparking but you can test by pulling a plug just when the bike has died.
e.

OP could compare the resistance on the coil’s secondary side (including the wire and caps) when cold vs when it dies.
Or, connect up a timing light to the secondary and see if the flash gets flaky just before it dies.
Or, do you have a scope?

[ajh]

Gorminrider - agree with all you say, thanks! Dyna is simple points replacement and is fine with OEM coils. BTW their support is second to none, really recommend them.
pdqford - I had completely forgotten the old trick of using a strobe, manys the time 30 plus years ago I would drive around in a car with the strobe wires through the window so I could see if the spark was failing. thanks for reminding

Heres todays update, bit brief as rain stopped play

I decided to replace the coil connector blocks as although they locked together nicely and I could not confirm a voltage drop or problem with them on the multimeter or moving them around, they just looked a bit green inside. They're hard to inspect inside without damaging. I was loathed to cut up original loom but decided to chop them out. I opened one up and sure enough the bullet connections inside werent good. And I couldnt get solder to take to the loom wire end even cutting it back, cleaning and fluxing so thats probably a sign. Anyway crimped new connectors and it seems to have done the trick. Havent been able to ride it yet but I ran it twice for a few minutes and it idled sweeter than ever, no problems.

Hopefully I can test fully it tomorrow

The 2 learnings for me -

1. connector blocks may look and feel ok but its hard to confirm 100% if the actual connection inside is good. Any doubts replace them
2. Its very difficult to test the voltage at the coils whilst running with a digital meter because they constantly sample the voltage and you have the switching from the ignition (thanks to Gorminrider for also confirming this)

And I have a theory (not confirmed yet) that on these old coils maximum output might be close to the voltage required to generate a spark so if primary volts drop below 10v, (almost 20% less than 12) the impact on coil HT output could be 20% less dropping it below whats needed to generate a spark? And then if the starter is taking more volts away......

fingers crossed I can confirm its fixed tomorrow because otherwise Im out of ideas!

Cheers

[ajh]

Just to elaborate, with the voltmeter on the 20V DC scale, and the “+” probe to the coil’s + end, BUT with the “-“ probe to the coil’s negative end, that will measure the voltage dropped across JUST the coil, doesn’t matter if running Dyna or points.

My head hurts wouldnt you have to make sure the crank position is with the points/dyna 'closed' on dwell so there is a permanent ground?