Intermittent Dead Cylinder...Switching Sides

I'll have the Fluke with me when I go this weekend, but from post 15: "The last time I checked it, I think the volts were ~14-ish at idle (when both cyls were firing) and, yes, the lights brightened when the engine's revved. It's ALWAYS done that (I've owned the bike since '75)."

Yes Sir Patience and cooperation lets see if we cant find it
lets take it from there .( Across battery measurement ) ..it is simple and quick to do .. and we then go on to other steps.If it checks out
Systematically trying to find something strange
if 14.4 fine it is checked I can be wrong not the first time ..

Then it will be to check at the fork in the cable splitting into 2 before entering coils
What voltage there
And at points how much voltage is there
Systematically step vise.
 
UPDATE:

Unfortunately, the battery in the Fluke was dead, so I was unable to gather any voltages, but I disassembled and cleaned—both mechanically and with contact cleaner—ALL of the wiring and ground connections between the ignition switch (where the key goes), kill switch and coils. Lotsa dirty/oily connections, but no broken wires or connectors....and no change in the dead cylinder or its randomly changing sides. In fact, today, I was seeing the dead cylinder switch sides while the engine was running! I'd blip the left carb and have it choke, denoting a dead cylinder. Then, I'd pull the left spark plug just to make sure it had no effect on the idle. It didn't. Then, with both spark plug leads connected, I'd blip the right carb....and, now, it chokes. So I go back to the left carb and blip it.....it revs like it should. I pull the left spark plug and the engine dies. I reconnect the spark lead and restart. Now, pulling the right spark lead stalls the engine, but the left lead doesn't..........

So, I let the bike idle while I was cleaning the rest of the lighting wire connections inside the headlight (since I was already there for the ignition wiring). All of a sudden, the idle rises ~200 RPM and it's smooth as butter—BOTH cylinder blip and fire smoothly. So I closed the headlight and took the bike out for an hour (too cold for more) and it ran perfectly.....Until I got back, turned off the engine and started it again.....Right cylinder's dead......

So, I re-checked the points for the umpteenth time, but all looked normal. I'm beginning to wonder if the public storage where I keep the bike might be located on top of an ancient Indian burial ground or something along those lines....

Ugh.
 
Sounds like a intermittent connection in headlight bucket.

Ignition connections in headlight .

A) Main switch connector.
B) Kill switch connection.

I'd also check the ground on the condenser.
 
Yes Sir this is progress.

for an hour (too cold for more) and it ran perfectly..

Mr Team Junk has it above I know it can be extremely hard to find where it is but there are notorious spots # 63

If this was me I would Bypass the front connections ( Lock / Kill / Rubber grommet into Head L ) and so on
And for testing make separate wires to Ignition parts Points and Coil. ( With a fuse )
Not for a long time solution but to make sure things are in order as long as power comes through
It can be the ignition parts also. Condenser ground ? .
 
Sounds like a intermittent connection in headlight bucket.
Ignition connections in headlight .
A) Main switch connector.
B) Kill switch connection.
I'd also check the ground on the condenser.

Thanks. From post 62: "I disassembled and cleaned—both mechanically and with contact cleaner—ALL of the wiring and ground connections between the ignition switch (where the key goes), kill switch and coils. Lotsa dirty/oily connections, but no broken wires or connectors....and no change in the dead cylinder or its randomly changing sides.....I let the bike idle while I was cleaning the rest of the lighting wire connections inside the headlight (since I was already there for the ignition wiring)."
 
Yes Sir this is progress.
for an hour (too cold for more) and it ran perfectly..

Sadly, no, it's not "progress". It's been doing this, INTERMITTENTLY and RANDOMLY (i.e., I never know if/when it'll happen), all along (for the past year and a half). This is the "clue" I thought the key to the source of the problem in one of my first posts from early 2020.


Mr Team Junk has it above I know it can be extremely hard to find where it is but there are notorious spots # 63
If this was me I would Bypass the front connections ( Lock / Kill / Rubber grommet into Head L ) and so on
And for testing make separate wires to Ignition parts Points and Coil. ( With a fuse )
Not for a long time solution but to make sure things are in order as long as power comes through
It can be the ignition parts also. Condenser ground ? .

If I had a private garage to work in, I'd try this. But, unfortunately, I don't. The bike's kept at a public storage and, because the bay I have is only large enough to fit the bike (not work around it), all work has to be performed in the (generally busy) public parking lot, where we're prohibited from working on vehicles.

I realize some who're trying to help missed last year's thread, where the prevailing diagnosis was carb problems (resulting in my rebuilding the carbs nine times, without any improvement). Because of where the bike's kept, I'm very limited in the amount of work/disassembly I'm able to do. Rewiring, even if only temporary, is just not realistic. And, frankly, I still think the problem's in or around the ignition switch. But, as I've checked/cleaned all the wiring and wiring connections between the switch and the coils and that had no effect, I remain stumped.

Again, I apologize in advance for these limitations, but it's what I'm dealing with.
 
OK still is progress keeping it simple
The Fluke needs a new battery that is something manageable at the Kitchen table ..and not at the parking lot and then some readings
Battery and later at the fork in the wiring previously mentioned . Not a big thing.

The testing harness we are talking about two -- four wires with a fuse at the most.. Between battery
And relevant places .. is that something you are able to do or find someone making one for you
If we google the stock harness and point to at which points it shall be connected.

Both these are not that big things so that it cannot be done under a street lamp even .
Here there is Car wash and petrol Stations one can rent DIY time per hour if it is to cold outside.
A few bucks there is better than haul it to a shop for electrical fault finding
 
The testing harness we are talking about two -- four wires with a fuse at the most.. Between battery
And relevant places .. is that something you are able to do or find someone making one for you
If we google the stock harness and point to at which points it shall be connected.

If the wires/fuse bypass is something I can simply hold onto two contacts (without having to disassemble anything), then I can do that. But I would feel better having a diagram indicating EXACTLY what points the wires need to contact so I don't mistakenly short anything out and create new problems.
 
If the wires/fuse bypass is something I can simply hold onto two contacts (without having to disassemble anything), then I can do that. But I would feel better having a diagram indicating EXACTLY what points the wires need to contact so I don't mistakenly short anything out and create new problems.

Ill check into this give me some time
Is that true --> Stock (OEM) 1972 XS2 650
No later installed electronics, ?
The principle is to add a parallel branch in the wiring to Coils and Points
Not going to the rest of the system Then there is certain the ignition subsystem is getting the right
physics
And preferably disconnect the original loop going to instruments .. and ignition lock
But the same can be achieved perhaps with ignition key off Ill check the wiring.

The fluke measurement of charging voltage would be a first though
 
motorcycleadventure.jpg
Ill check into this give me some time
Is that true --> Stock (OEM) 1972 XS2 650
No later installed electronics, ?

Correct. OEM 1972 XS2.
The only non-OEM item on the bike is a charging system indicator I installed on the handlebars after a failed stator caused me to get stuck in the middle of nowhere some years ago. At idle (with both cylinders running), the indicator reads normal.


The fluke measurement of charging voltage would be a first though

I'll be replacing the Fluke's battery this week and try to test the system voltage this weekend.
 
Last edited:
I don't know about charging (not really able to check it), but the battery's fully charged.

Hi freddy,
waiting until it's dark works best for this test.
Start the bike, point the headlight at a wall and rev up.
If the headlight brightens at ~2,500 rpm the alternator is charging. If it don't, it ain't.
 
Please!! Good battery voltage is only half of the story....... Sufficient voltage at coil +, coils in good nick AND connected correctly , and good conductivity from coil - , through points, to ground is all equally important. Points ignition systems are dead simple, not black magic. This thread has been running in circles, without getting anywhere. I bet the issue can be isolated within an hour, just using a voltmeter, a couple of meters of copper wire, and a handful of bullet connectors. It is really that simple.....And all necessary info has already been provided, multiple times.
 
Hi freddy,
waiting until it's dark works best for this test.
Start the bike, point the headlight at a wall and rev up.
If the headlight brightens at ~2,500 rpm the alternator is charging. If it don't, it ain't.

The headlight does brighten with higher revs, but it has done that since I've owned it (mid-1970s)..
From post 60: "Presuming similar engine rpms, I've seen no difference in the brightness of the headlight (beyond the general brightening of the headlight when the engine's revved as it's done since I've owned the bike) between the times when the engine's running rough (i.e., on one cylinder) and when it's running smooth (i.e., on both cylinders)."
 
Please!! Good battery voltage is only half of the story....... Sufficient voltage at coil +, coils in good nick AND connected correctly , and good conductivity from coil - , through points, to ground is all equally important. Points ignition systems are dead simple, not black magic. This thread has been running in circles, without getting anywhere. I bet the issue can be isolated within an hour, just using a voltmeter, a couple of meters of copper wire, and a handful of bullet connectors. It is really that simple.....And all necessary info has already been provided, multiple times.

I reckon you're probably right, and if I had a garage in which to perform the necessary tests/mods, I'd be able to confidently verify one way or the other. But, unfortunately, I don't. Also, unless something's loose that's causing the volts to randomly and intermittently drop, I don't think the random, intermittent and switching dead cylinder problem is due to general low voltage. If it were, I think the problem would be consistent, affecting one cylinder all the time. On Sunday, when I experienced the dead cylinder switching sides while the engine was running (and after I'd finished cleaning ALL of the connections/grounds between the handlebar switches and coils, to include the condenser grounds), I pretty much ruled low voltage out. Again, I'll check this weekend with (a new battery in) the Fluke and report back the result.
 
Having looked at the schematic I have to back down on the bypass wiring at this point in time
It is to risky and to little to gain now.
It would be different if was close to the bike

If Running please inspect what it looks like at the Points
There is a little sparking going on normally.
Please notice what i looks like when one cylinder is dead And look for difference when both runs.
Is there any sparking going on at the points on the cylinder not Firing ?

BUT the wiring is there
http://www.xs650.com/threads/72xs2-73tx-wiring-diagram.53244/

Downstream of the kill switch it looks like that

upload_2021-11-9_1-26-34.png


The measurements to take with the fresh battery in the fluke

1 .Voltage across battery ignition off
2. Voltage across battery ignition on
then comes the tricky part downstream from the Kill Switch goes a R/W wire ( red white I suppose )
That wire goes first to a connector

top right there

upload_2021-11-9_1-32-8.png


At that point there needs to be Around 12 V

I would probably try to measure it in the Kill Switch itself at the downside at first I have an 80 and can open it on the handlebar
That makes it possible to measure when running.

I would then measure what voltage there is going in to the points accessible after removing the points cover


upload_2021-11-9_1-43-49.png
 
If Running please inspect what it looks like at the Points
There is a little sparking going on normally.
Please notice what i looks like when one cylinder is dead And look for difference when both runs.
Is there any sparking going on at the points on the cylinder not Firing ?

I'll check the points again, but I've been checking them repeatedly for the past year and a half and they appear to be firing (sparking) normally.



Downstream of the kill switch it looks like that
The measurements to take with the fresh battery in the fluke
1 .Voltage across battery ignition off
2. Voltage across battery ignition on
then comes the tricky part downstream from the Kill Switch goes a R/W wire ( red white I suppose )
That wire goes first to a connector
top right there
At that point there needs to be Around 12 V
I would probably try to measure it in the Kill Switch itself at the downside at first I have an 80 and can open it on the handlebar
That makes it possible to measure when running.
I would then measure what voltage there is going in to the points accessible after removing the points cover

I've printed out these instructions, along with the diagrams, and I'll bring them with me this weekend....along with the Fluke.
I'll report back with my voltage readings.
Thank you.
 
The two items I would suspect are the coils and condensers. 50 year old capacitors are especially sketchy.
 
Kill switch coupled with the Electric start and the safety relay was an add on, (onto the XS1/1B wiring loom), on the XS2 wiring..........

You probably know all this, and i have no solutions to your problem, but this may be of some help to make something click some where

If you want a larger diagram i can email one to you. These have to be resized to 1200 x 880 to post. My original is 3000 x 2400 pixels.

Don't know if this will help much but here is the XS2/TX wiring diagram. The only difference is the Brake switch wire may be Y/g instead of Y.

The finished diagram is in the first post. The rest of the thread is what we went trough, (with the help of a couple of members with an XS2/TX, I don't have an XS2).
so going through it may help. lots of pics of the loom and electrical components. On the last page #5 is a great shot of the Red/White wire, from the kill switch where it feeds into a triple connector, (actually where it feeds into the connector it cant be separated from the brown wire feeding back to the stater relay), and separates into 2 other brown wires with 2 bullet connectors, one to each coil. Brown, keyed power.
72-xs2-circuit-diagram-b11325607311619 Colour aaaaa G Text 15 a 1  copy resize 2000.jpg

72-xs2-circuit-diagram-b11325607311619 Colour aaaaa G Text 15 a 1  copy resize 2000.jpg

IMG_20181016_160400.jpg

This how Yamaha married the power, on brown wire, to new safety relay and on to the solenoid switch instead of building a new loom like they did from 74 and later
 
Last edited:
Back
Top