1981 Suzuki GS650G - dies under load when throttled past ~1/3

[Nessism]

The linked thread mentions that I'll "need to run another wire down to the ignition for switched 12v power for the pickups"; the way it was phrased makes it sound like it's another thing you have to do differently on the 650.

No. It's the same old Dyna-S system, regardless of what bike.

 

[pdqford]

I checked the spark again (via the screwdriver method). It seemed really, really weak, even right off the bat at 12.6. It also seemed intermittent - not the steady every-other-go-around it should be.

Depending on how you are using the screwdriver method, if you are grounding the screwdriver to a metal surface, you are checking for spark at atmospheric pressure.
If you are grounding the screwdriver to the tip of the sparkplug you are checking for spark under compression. It takes a bunch more voltage to throw a spark under compression as opposed to atmospheric pressure.

I'd try three more quick checks:
Put your DVOM across the battery terminals and hit the starter button to see how low the voltage drops under load. I'd look for something over 11 volts.

If the battery shows that it can handle the load, put the + meter probe on the battery + post and the negative lead on the + side of the coil and hit the starter button.
This will measure the voltage dropped on the power side of the ignition system. You would like to see no more than 0.2 to 0.3 volts dropped on the power side of the ignition system.

Then put the the + meter probe on the NEGATIVE coil terminal and the negative meter probe on the NEGATIVE battery post and hit the starter button.
This will measure the voltage dropped on the ground side of the ignition. Again, you want no more than 0.2 to 0.3 volts dropped on the ground side of the ignition system.

If you are dropping too much voltage on the power and/or the ground side, you need to find and clean any poor connections.

Believe it or not, I found most of the dropped voltage on the power side on my GS750 in the glass fuzz!

 

[seinwave]

I was sticking the screwdriver into a plug cap and jumping the spark to a fin on the engine. I had to get it a hair's breadth from touching to get any spark.

Anecdotally, I grabbed the screwdriver the wrong way the first time and grounded it to *me*. It was a shockingly dull shock, a slight throb and spasm without any real pain. Not that I've properly calibrated myself to feel spark strength.

I'll do the above tests once the battery I ordered arrived; the one I bought from Autozone was two years old and read 11.5V when I got it back home. Returned that ASAP.

More evidence pointing towards an electrical fault (of some kind). On a whim I charged up the jump-pack I use on my car and hooked it up to the bike's battery. When I hit the starter it immediately fired a few times (at least a second or so), then stopped firing. The jump-pack was drained by about 15-20% off that one touch of the starter (not that tells us a whole lot - it's an old battery pack and I don't have any way to measure voltage over it).

Any chance this could be misdirection, and I'm actually facing a fueling issue and that fuel somehow seeped into the cylinder, causing it to fire briefly? I can't see any possible way that's the case but I hate this sitting with my thumb up my ass for parts to arrive so I can confirm a diagnosis.

 

[salty_monk]

There are some tests in the Clymer for the ignitor. Mine showed up bad on the one where you use a small battery to feed it from memory...

 

[seinwave]

Good news: think I've fixed it. Bad (or funny, or expected) news: it was probably the coils all along.

The new battery got here yesterday. Charged it up and threw it on after work. At first it just sputtered like when I used the jump pack, but some fiddling with idle speed and I got it to fire. It was surging/revving pretty bad; I tightened down the carb->engine boots (had left them far looser than I realized), which made things better but didn't fix it.

Today (around 3pm EST) headed out with a mind to fix the surging and get it back to baseline. Reseated the airbox->carb boots - I'm suspicious of #2 (it seems to have the wrong shape), but I did get them seated a little more evenly. That made the surge even better - now all it was doing was hanging at 4k revs when I throttled up and down. Played around with the mixture screws and got no real response. Then I realized what I thought was a hang at 4k was actually a very sputtery start-up that jumped to a stable 4k idle when I gave it a little gas. Fixed the "surge" with two turns of the idle speed screw (and a quarter-turn out on the idle mixture). Oops.

Okay, now it's back how it was before - doesn't want to get above 4 or 5k, dies smooth if you try and force it. I was planning to leave it there until the weekend but I had the coils and a little time so decided to go for it. Yanked the old ones off. While consulting the manual(s), I realized that the PO had the wires going to the ignition switch on the (+) side of one coil and the (-) side of another. As far as I can tell polarity (aka if the ignition switch wires are on the + or -) doesn't matter with a transistorized ignition - but surely it couldn't be great to have them different on either side.

Swapped the coils, swapped over the boots to the new coils. Plugged them in. Crossed myself (a thing I only ever do when I work with carburetor engines or forgot to move my car for street sweeping) and hit the starter. Sputtered strangely for a second, then it fired. I let it warm up for about 30 seconds at idle. When I last left it, it was working beautifully: nice smooth pull up to at least 7 or 8k (haven't taken it to redline at a standstill yet - let's not push our luck until I double-check the valves again).

Obvious caveat of I haven't tried to ride it yet; maybe it isn't fixed under load (doubt), or maybe there's a brand new problem that'll cause it to die. But at the very least I've fixed a serious issue and I'm feeling confident that I've got a rideable bike again.

Two Big Lessons learned for me:
- When facing a mechanical issue, choose on instinct what the issue is and immediately spend $180 on a top-of-the-line replacement. Diagnosis wastes time and lacks decisiveness.
- NEVER open up your carburetors to clean or check them. The previous owner will have always done a great job when he last touched them, and if he knew you doubted him he'd be hurt.

Feel free to add these points to the newbie advice, I think it'd help a lot of people ;^)

 

[seinwave]

And final update, for the zero people still concerned: took it on a ride and it works beautifully under load! Now I just need to replace the geriatric valve cover gasket (finally gave up the ghost and started spitting oil properly) with the silicon one I've had in a bag for a couple weeks.

 

[seinwave]

...spoke too soon I think. Replaced the valve cover gasket and tried to fire it up. Nothing. After way too much tinkering I found no spark on the righthand HT lead of the left coil. Problems persist when I swap which coil gets which line from the igniter, so pretty sure it's isolated to the coil. Same thing with spark caps etc.

It's possible I tugged on that line when replacing the gasket and somehow fucked it (but it is still continuous with the other HT on the same coil, so it'd be an intermittent connection). It's more likely (I hope) that this is what I get for paying $40 for 2 rebranded likely-chinesium coils instead of paying $40 for 4 known-chinesium coils where I would have spares. It's possible the igniter is somehow bad and damaged the coil - but again, it would be some weird damage-over-time situation, since both igniter leads produce the same effect. Can a bad igniter damage a coil like this?

I tried to test with one of the old coils. I *did* get a spark on both lines when I replaced the new coil with the bad HT with the old coil, but the old coil (as expected) produced a very weak spark. After I tried cranking it for a while and then checked the spark, the old coil had no spark, so I wasn't able to rule anything out with that.

I'm 99% sure the valve cover gasket couldn't cause anything like a no-start; it's not under compression. The one thing that could mess with it that I can think of is if a big chunk of the old gasket got wedged between a cam lobe and a valve shim and fucked the clearances up. I went through the bother of checking for this by removing the cover again and all is clear.

I'm also not 100% sure that fuel is getting through the carbs into the cylinder. The plugs looked totally dry when I cranked them (except #4, where the bad HT lead went; that looked fouled). The fuel level in the fuel line goes down very slowly, but not really in sync with me cranking it. I checked and the bowls are full.

At some point the bolt on the tightener between the intake manifold (where the boots are) and the filter box took its leave of absence. I tightened it down with a zip tie temporarily (don't have a screw of the correct size) and no difference. Can't really tell if it's properly cinched down.

The battery still seems good.

Cranking seems to produce one faint fire the first moment I hit the starter and then nothing but strong cranking for the rest of the time I have the button depressed. This feels a little like how it was once I put the carbs back on. I'm pretty sure I've exhausted every possible combination of idle speed, throttle, and choke.

I am back to losing patience with this bike. The new coils aren't going to be delivered before Friday, which is when I planned to set off on a highly-anticipated weekend trip into the mountains. I'm not going to spend the rest of the year playing this same game. If I start fixing the same thing for the third time my mind is going to turn to cutting my losses more and more.

 

[pdqford]

.. Can a bad igniter damage a coil like this? .

I doubt it, but never say never.

Now the other way around, a bad coil could mess up the ignitor
All the ignitor does is provide a ground for the coil primary circuit.
And of course the transistor in the ignitor releases the ground to induce a spark in the coil secondary.

I don't recall if you checked the resistance in the primary side of the coil.
If it is too low, too much current could pass through and fry that transistor in the ignitor that makes and breaks the coil primary circuit.

So again, I would recheck the resistance in the coil primaries and see if they are within speck.
And while checking the coil resistance, also check its primary circuit to ground.

And maybe I missed it, but what did you measure for voltage drops in the power side and ground side of each coil?
(I think I explained how to do that up above someplace.) (Post #42).

 

[seinwave]

I checked the primary and secondary resistances on the old coils, all were comfortably within spec (which is why it took me so long to settle on them as the culprit). I've only checked the secondary resistances on the new coils, which were both on spec (even the one with the bad lead). I'll check the primary resistances on the new coils before I install the new-new coils.

Thinking on it a little more: I think the new coils (on the bike now) might have had the bad HT lead from the start (aka even when I took my test ride). The plug connected to the bad lead was SOAKED, like it had seen a ton of gas and burnt none of it, way more than I'd expect to see from unsuccessful cranking (and remember, the other plugs were bone dry after the cranking). I'd be pretty surprised if the bike ran as well as it did on three cylinders. But it does point to something else being the issue behind the no-start.

The Dyna ignition I ordered due to my inherently pessimistic nature is sitting safely in my box of parts, so if the coils burnt the igniter (or I think they burnt the igniter) it's not a big lift to swap it out. I'd like to avoid doing it without some evidence of failure, though, since setting the timing then becomes another variable I can mess up.

I didn't measure the voltage drops over the coils, so giddy I was with the bike working. I'll do that with the new ones.

 

[pdqford]

Thinking on it a little more: I think the new coils (on the bike now) might have had the bad HT lead from the start (aka even when I took my test ride). The plug connected to the bad lead was SOAKED, like it had seen a ton of gas and burnt none of it, way more than I'd expect to see from unsuccessful cranking (and remember, the other plugs were bone dry after the cranking). I'd be pretty surprised if the bike ran as well as it did on three cylinders. But it does point to something else being the issue behind the no-start.
.

On these wasted spark ignition systems one pair of plugs fire at the same time, one plug is on the compression stroke while the other plug is on the exhaust stroke. The plug on the exhaust stroke has nothing to ignite hence the term "wasted spark". Typically, the secondary current flows out of one side of the coil, through the plug wire to the plug, jumps the gap and sparks from center electrode to side electrode, continues on through the engine block to the side electrode of the companion plug, jumping that plugs gap from the SIDE electrode to the center electrode, and returns back through the companion plugs HT lead to the coil to complete the circuit.

If one of the plug pairs is wet with fuel it can throw a spark that doesn't ignite anything while the companion cylinder does spark and ignite, you may be experiencing two or three problems simultaneously. Are you sure there is not a problem with that carb? Sticky float? Worn float needle? Bad float needle gasket? Heavy float?

I didn't measure the voltage drops over the coils, so giddy I was with the bike working. I'll do that with the new ones.

Measuring the power and ground voltage drops of each coil would show if you have a marginal or bad connection in either the power side or ground side, of each coil. Voltage drops will reduce amps available to the primary circuit of that coil. It​ is not going to be fully charged, and the secondary and will produce a week or no spark.

When you measure the voltage drops on the ground side of each coil, measure it back to the battery negative. That way you will be checking the ignitor grounds or its transistors

The Dyna ignition I ordered due to my inherently pessimistic nature is sitting safely in my box of parts, so if the coils burnt the igniter (or I think they burnt the igniter) it's not a big lift to swap it out. I'd like to avoid doing it without some evidence of failure, .

I'm with you there. I would like to have an aha diagnostic moment before spending money and changing parts
But most on this site have so much experience that if they could see and hear the bike, they'd know the solution from experience.​
Here's hoping you figure it out!!

 

[seinwave]

I'm pretty certain there is no problem with the carb (very recently cleaned and rebuilt, and it was pretty spotless to start with), and I am 100% certain there is something wrong with the coil. I've tried every combination of cap, igniter lead, cylinder, and plug (I'm *not* using the extremely-wet one to test) I can think of. The result is always the same: 3 leads produce strong sparks visible even in the middle of the day, and one lead (the bad one) produces absolutely zilch.

If there's an extra issue I'm leaning towards it also being electrical. These symptoms are a lot like what I had when I declared the battery to be 'bad' - except this battery isn't bad. Could definitely see hunting down a bad ground once I investigate the vdrops.

 

[seinwave]

No progress, but a clue.

Finally got a dry free hour to install the new coils - hadn't had a chance since I came home from my weekend trip. Resistances on primary and secondary measure good.

The battery is still strong. Right off the charger it read 13.3V, which dropped to 13.1 when I had the key and killswitch on.

Once I had the new coils installed, I tried to measure the voltage drops. I quickly realized that without four hands, clamping terminals on the voltmeter, or my fiancee to press into service, I wasn't going to be able to run the starter and get good numbers, so these numbers are without the starter button pressed (the key and killswitch were both switched to on). If these numbers are worthless or don't reveal anything, I'll get starter-on numbers as well (when the thunderstorms that chased me inside stop). But I think I *have* found something.

2-3 coil:
Battery plus to ignition switch wire (orange): 1.13V
Battery plus to black/yellow igniter wire: 1.09V

1-4 coil (where the 'bad' coil was when it went bad)
Battery plus to ignition switch wire (orange): 1.09V
Battery plus to white igniter wire: 10.91V

So it looks like there's something causing far too much voltage drop on the wire the 'bad' coil was hooked up to. I'm assuming that the absolute numbers don't tell much without the starter pressed, but surely one section of wire reading ten times the drop of the other sections is a big flashing signpost.

I noticed upon inspection that spade connector on the white wire was way looser than it should be. It felt tight in that you couldn't pull it off the stud on the coil without the expected amount of force, but even 'tightly' on the stud it rattled around easily if you shook it. I'll be cutting, stripping, and re-installing a fresh connector.

Apart from that connector:

- Is this likelier to be something bad inside the igniter, or something in the wiring harness?
- If it's in the wiring harness, any known points of failure I should look at?

(pinging pdqford as you seem to have gone through something similar with your 750)

Note: haven't started it yet; I don't want to risk killing another coil if that's what happened with this wiring issue.

 

[salty_monk]

Try running it / starting it in the dark & look around the head. You may find the HT wires have cracks you're not seeing and are grounding out to the head. You may find moving them around (not with a bare hand otherwise it might ground the spark through you!) makes more or less sparks etc.

Sometimes people will test for spark & it looks good but the wire is not in the same location as when running it (i.e. not close enough to the head to ground out). Just an idea...

 

[pdqford]

2-3 coil:
Battery plus to ignition switch wire (orange): 1.13V
Battery plus to black/yellow igniter wire: 1.09V

1-4 coil (where the 'bad' coil was when it went bad)
Battery plus to ignition switch wire (orange): 1.09V
Battery plus to white igniter wire: 10.91V

I understand the + meter probe is on the + battery post.
But to test the voltage drop on the coil power side, the - meter probe should be on the POSITIVE side of the coil.

To test the voltage drop on the coil ground side:
The + meter probe should be on the NEGATIVE side of the coil,
and the - meter probe should be on the - battery post.

(Effectively, the meter is providing an parallel path to the circuit being tested.
The voltage that can't get through the circuit, for what ever reason, goes through the meter, and is the voltage being dropped by that circuit.)

I'm gonna be wire color challenged as I can no longer access the GS650 wiring diagram on BikeCliffs website :-(

 

[seinwave]

The manual is very squirrelly about which side is positive which side is negative on each ignition coil circuit. I didn't find anything when I scoured, and in any case the coils themselves are unmarked. I assume the killswitch is drain to ground, but I figured I'd just measure each possible combination.

I'll go back down and measure each pole to the negative battery as well.

Sounds like the powerside drop measurement is (+) on battery plus, (-) on igniter wire side of coil. Ground side drop should be (-) on battery negative, (+) on killswitch side of coil.

Seems like I have a bad connection somewhere on the power side of one of the coils. Likely to be the igniter, or a break in the harness?

from what I can see the only notable failure point on the power side that wouldn't affect both coils is the internals of the igniter.

 

[pdqford]

let me try this way:

The power side of the coil primary circuit runs from,
battery + post,
through a connector,
to the fuse box,
through the main fuse,
out the main fuse on the red wire,
through a connector way up to the ignition switch,
through the ignition switch contacts,
out the ignition switch on the orange wire,
through a connector waaay back to the ignition fuse,
through the ignition fuse, and,
out the O/W wire,
through a connector, waaay back up to the kill switch,
through the kill switch,
through a TWO more connectors,
and connection to the + side a coil!
(Whewh! Hope the GS650 is similar to the generic GS wiring.)

So now if you put your + meter probe on the + battery post,
and the - meter probe on where the O/W wire connects to the coil,
with the ignition switch on,
your meter will show the voltage dropped by that path (circuit) described above!
Any poor (green stuff?) connections, any crudded up switches, loose wire connectors, etc,
will impact the voltage dropped in that circuit, UP TO the coil.

How many volts is your 1/4 coil power side dropping?

Note, if in your post above, you placed the - meter probe,
on the white wire (which is the ground side of the coil primary),
then you are including the voltage dropped by the coil.

The goal is to drop minimal voltage on the power side,
and minimal voltage on the ground side, and
therefor the coil will get the maximum voltage available.

Now if your power side is dropping say over 0.4 to 0.5 volts,
grab your wiring diagram and "halve" the circuit.
With the + meter probe still on the + battery post,
move the - meter probe to maybe the ignition fuse.
If the meter reading drops, you know the worst voltage drop
is in the other half of the circuit.
So "halve" the other half of the circuit and repeat.

Not saying yours will be the same,
but on my GS750E the biggest voltage drops,
were in the ignition switch,
the glass ignition fuse,
and a factory splice hidden inside the main wiring harness.

 

[seinwave]

Ah, I (think I) see -- I was wrong about the O/W ignition switch wire being the 'ground' side; it is the positive/'power' side of the ignition coil. The white and black/yellow wires are the 'ground' side.

Which means that:
(meter positive) on battery positive terminal
to
(meter negative) on O/W ignition switch wire on coil
is the power side voltage drop.

And
(meter positive) on white or black/yellow wire on coil
to
(meter negative) on battery negative terminal
is the ground side voltage drop.

(And meter positive on battery positive terminal to meter negative on white or black-yellow wire on coil (what I did above) is power side voltage drop plus the voltage drop over the coil, which doesn't mean much.)

So in conclusion the numbers I have tell us very little, except that the connections on the power side of both coils are in a similar state, and certainly don't point towards any particular wiring issue (yet).

The 1-4 coil dropped 1.09V (without the starter pressed) and the 2-3 coil dropped 1.13V (without the starter pressed) on the power sides.

I assume that the reading from battery plus to black/yellow igniter wire on the 2-3 coil is garbage data - maybe I misread the decimal point - since I really don't think there's any way for the voltage drop to lessen over the coil itself.

Thanks for your patience. I'll get proper numbers (power side & ground side, with starter pressed if I can work it) and see what comes up. This has revived hopes I just had a bad coil (the only failure mode I can think of is arcing through a badly insulated HT lead or similar, like salty_monk talks about above) and that it'll all work as desired (and keep working for more than 6 miles) when I finish my checkouts and hit the starter button in earnest.

 

[pdqford]

Ah, I (think I) see -- I was wrong about the O/W ignition switch wire being the 'ground' side; it is the positive/'power' side of the ignition coil. The white and black/yellow wires are the 'ground' side.

Which means that:
(meter positive) on battery positive terminal
to
(meter negative) on O/W ignition switch wire on coil
is the power side voltage drop.

And
(meter positive) on white or black/yellow wire on coil
to
(meter negative) on battery negative terminal
is the ground side voltage drop.

By golly, you got it!

The 1-4 coil dropped 1.09V (without the starter pressed) and the 2-3 coil dropped 1.13V (without the starter pressed) on the power sides.

You don't really​ need to press the starter button as you have shown that the battery has enough umph (technical term) to power up the starter and still have something left over for the coils.

The reason both coil's primary voltage drop is nearly identical, grab your wiring diagram, and see that each coil's power side share the EXACT same path, except for the last foot or so of the path. At that point the O/W wire splits into 2 O/W wires, with one wire powering up the 1-4 coil and one powering up the 2-3 coil.

I think that voltage drop is a little high, but at least it is what it is, and you know what it is.

Now check the ground side of the coil primary circuit for voltage drop.
The + meter probe on the white wire at the coil terminal and - meter probe on the - battery post to check the ground side drop of one coil, and then move the + meter probe to the O/B wire of the other coil to check its ground side drop.
This will check the integrity of the white wire all the way to the ignition box, the ignition box transistor and the ignition box's connection to ground and back to the battery negative post. Ditto for the O/B circuit.

Notice that on the ground side, they DON'T share the same ground path, except for the last few feet. Once they pass through their respective transistors they then share grounds with the IB ground on back to the - battery post.

So if you take the battery's voltage, minus the power side voltage drop and ground side the ground side voltage, you end up with the voltage each coil has to work with.

 

[seinwave]

Alright, took the proper tests and I think I found my smoking gun igniterwise.

Measurements on the power side look similar to last time. On the ground side, the black/yellow ground-side was dropping about 1.78V, but I found that the white (1-4) cable was dropping 11+ volts.

I pushed the white cable onto the ignition coil a little more and found that the white side cable was now dropping about 1.5V, but the black/yellow side was now dropping 11+V.

I took the white cable off the coil post and saw arcing between the spade connector and the coil post in the process. Putting back on moved the bad drop back to the white side. I tested a few more configurations and the 'bad' side moved back and forth. I didn't see any additional arcing.

There's no convenient exposed place on the ground side I could find to 'half' the circuit, but at the very least I tested resistance between where the white, black/yellow, and orange/white lines leave the coils and the molex connector where they're plugged into the igniter wires. All had negligible resistance (the black/yellow and white wire seemed to maybe show an ohm or two for a split-second a few times).

The high voltage drop and the 'switching' of where it was seems to point squarely in the direction of the igniter to me. I guess it's also possible that there's a strange sort of short inside the wrapping between the white and black/yellow wires, but that seems far-fetched and I don't have any easy way to look for a short like that without ripping most of the harness to shreds.

I'll take the next few days to prep, install, and time the Dyna-S, unless someone has another idea to where to go next?

 

[Nessism]

Ignitors are often killed by high charging voltage, a result of bad R/R grounds, or an outright failed R/R. The 2nd gen 550's and 3rd gen 750's are particularly prone to this. I think you are on the right path by installing a Dyna S. Nothing to lose.