Thought my problem was solved, but it's not. Quick recap: I have a new HES unit. Used (but known good) coils from a forum member. Also put in new spark plug leads.

After putting in the new HES, the bike fired right up on all cylinders. All seemed good. Didn't ride it then because it was late night. Took it out today. It started right up, running on all 4 cylinders. Rode down the street and had major backfiring and lurching/hesitation. Thought my timing was off so I moved the HES plate a bit counterclockwise (looking at the engine from the front) and took a ride. Same symptoms, lurching, backfiring. Though at full throttle as it approached redline it ran fine. Adjusted the HES the other direction (clockwise) and now I'm only firing on 2-3. Put the HES back to the original position and have the same no firing result on 1-4. Pulled plug 1 and have no spark and wet, unburned gas.

At the + input on both coils I get 11.2v with the key on and engine not running (my battery shows 12.5 or 12.6 in this state; headlight is on. With the engine running (on 2-3 only) I've gotten inconsistent readings. One time both + terminals were showing 7.2v. Another time both were showing 13v.

On coil 1-4 the - side always shows a low number, about 3.5. On the good, firing 2-3 coil the - side always reads what the + side reads. I also wired up an LED to attach to the +/- at the coils. Both coils show a pulsing light. Further, I have my tachometer sensor wound around cylinder 1 plug lead and I get tachometer readings.

Also have swapped the plugs around and the 1-4 plugs worked in 2-3. So the plugs are good.

At a loss here. What do I do next?

A bit of a mystery Justin. Must be very frustrating for you mate.

Have you checked the coil earthing. Give the spade connectors a good clean/deox and check for continuity back to the GB earth with a DMM.

You've replaced the HES. Did that unit from EME include the harness back up to the connector under the tank? While the sensors themselves can fail, the harnes wiring can also. The insulation can suffer over time

Hang in there.

Yes, extremely frustrating.  Thanks for the encouragement.

The new HES unit has a new wire and plug.  The plug is different than OEM, so I had to cut the OEM plug from the loom an install the new plug.  The new plug is molex-type.  I didn't have the necessary crimping tool to attach the pins, so I soldered them to the wires.  It is quite solid.  It could be the problem, and I'm going to double check, but I doubt it.  

I have earth on both coils going back to the gearbox. I tested it both with the earths attached to the coils and unattached to the coils. In both cases there was continuity. For good measure, I also cleaned the earth connectors with deoxit.  Tried starting the bike again and it started and ran on 2-3 as before.  

I turned it off, waited a moment, and tried a second time and my starter relay got stuck on. So, I disconnected the - from the gearbox to make it stop.  This happened before, and I got it unstuck by giving the relay a firm whack.  I'm still using my old relay; a new one is on its way from China.

When this happened before I assumed that it was because I had a low battery.  But I have a brand new lead-acid battery installed and the voltage, even after cranking for a while before I disconnected it, it read 12.5 volts.  So, my battery is not low/weak.  

Given that I've experienced this twice now, I'm wondering if the relay is somehow related to my no-spark situation?  That seems hard to believe.  I'm thinking the relay is a separate issue.

As before, all suggestions welcome.

This sounds like an intermitant fault reminiscent of a loose plug. Can you confirm that the ICU (up front under tank) connections are clean and the plug is firmly secured?

Kaptian to the rescue again!  I cleaned the ICU plug with deoxit and I'm now getting all four cylinders again.  

I'm not going to declare victory yet.  It's too late tonight to go out for a ride.  I'll take it out tomorrow and if all is well then, I'll report back.  

Interesting point of history: a while back I had a no start problem.  After much troubleshooting, cleaning all under-tank connections, I found that I'd knocked the ICU plug loose while running some other wiring.  I plugged it back in, and things worked, and away I went.  But I don't think I ever bothered to clean that connector before.  Hoping that this was the problem and finally this issue to rest.

Awesome

Re: your starter relay. Sometimes you get a little blob of molten contact metal solidify between the contacts. Clean with some fine wet & dry and they'll last a little longer. Keep as a spare once the new one is installed.

Hats off to you for your tenacity, Justin!
Good luck!

Took the bike out for a quick ride today. It appears that my problem is SOLVED. Looks like it was a dirty ICU connection. This makes me wonder if everything else I did was necessary. May not have been, but it probably was a good idea to do the replacements I did since all items were known failure points, were 30 years old and had a good number of miles on them.

Only rode about 10 miles, but that allowed me to go through all gear, speed, and RPM combinations. Even the timing seems to be correct, though I might tweak it in the future as a performance enhancement.

Brickrider, thanks. I must confess, my tenacity was waning. I was giving serious consideration to giving up and parting the bike out. I'm glad I got this fixed.

Thanks again to all that helped on this thread in the others.

"I was giving serious consideration to giving up and parting the bike out"  

Sounds like we saved another K from certain execution. Enjoy the ride.

At the + input on both coils I get 11.2v with the key on and engine not running (my battery shows 12.5 or 12.6 in this state; headlight is on. With the engine running (on 2-3 only) I've gotten inconsistent readings. One time both + terminals were showing 7.2v. Another time both were showing 13v.

Glad you've got if fixed Justin.

On the inconsistent voltages: when the engine is running and the coil circuit gets interrupted by ECU to make a spark this will generate quite high negative voltage spikes in the primary winding too, ie where you are measuring. There is probably some clamping mechanism to protect the output of ignition circuit but it's going to be very noisy.

If all is working, the coil is going to be connected for something like 2/3 of an engine cycle then cut off by the ECU. You can work this out by measuring the length of the slots in the cylindrical vane that goes through the HES pick-ups.

I suspect 7.2V is probably about the right time-averaged DC equivalent.

If the ECU is not triggering you will probably have battery voltage connected continuously to the coils which are not firing, so they will be close to 13V

There will also be some noise picked up in the voltmeter wires which are in proximity to the two which are firing. A lot of EM pickup and stray magnetic fields. A cheap analogue voltmeter may average this out better than a DVM.


You are measuring a very noisy alternating signal not a DC voltage.

Beamer, that's very interesting, thanks. And it's consistent with what I saw and was experiencing. I only have a digital meter, and it was jumping around a lot, very hard to read.

This is a picture of the signal you're measuring.  It's the waveform produced on the primary of coil #1 on a K75 at idle.

That's very interesting Rob. What is the voltage of the baseline on that trace?

Do you have a link to a circuit diagram for the amplifier unit which drives the coils?

The baseline is 0V.

Dieters K Jetronik Kompendium is probably the closest you'll come to a schematic of any of the Bosch L-Jetronic components.  Bosch Anything is very much a black art.  Read the document Messungen an der L-Jetronic von DieterMN.  He shows it's basically a Darlington pair that drives the coils.

Many thanks, that looks useful.

As I suspected, there's a zener diode to control the negative spikes.  While there's not any proper doc. from Bosch, apparently, it's all pretty standard stuff and a lot more transparent the microprocessor controlled stuff where the firmware really is invisible.

This is a picture of the signal you're measuring.  It's the waveform produced on the primary of coil #1 on a K75 at idle.

Am I right in thinking that that is the waveform of the collector of the output transistor connected to the negative of the coil and what we see in your snapshot is both the turn off and turn on events?

It's the voltage seen on the Black/Green wire on coil #1, relative to ground. I can't imagine that these are the turn on/turn off triggers. A coil is not a monostable device. Sorry, I can't comment on what I'm seeing here since I haven't analyzed the circuitry behind this. I was scoping the wire at the time I was preparing to connect up my aftermarket gauge, checking that the signal wouldn't exceed the specs of the gauge.