Those original coils are well known for doing that too......just like your experience in cars. Its a good move anyway to change them.

But both coils giving out at the same time doesn't seem likely. However I don't see the K running too well on just one coil.

Still never really worked out if its fuel supply being disrupted or the ignition side being disrupted. Check when it happens again by opening fuel tank and seeing if the fuel is being pumped back into the tank while you have the starter engaged. No fuel means it's not the coils and you should be looking on ignition control side and fuel supply side. There have been instances of poor connections to the electrical fuel pump plug under the front left of the tank causing this cut out and I have been there with it. The wires can even become detached from the metal pins.

I just went through all of the electrical connections, including the plug under the tank which actually looked pretty tidy and clean, but I used an anti-oxident electrical cleaner liquid on everything anyway and plugged it all back up. 

I had forgotten that there are in fact two coils and that it would be unlikely both going at once.  It just simply dies and then will start right back up.  Bizarre and frustrating.  I'll get to the bottom of it but I'll be surprised if it's just a faulty electrical contact.

ok, I'm really reaching out to you all.  I'm confounded by the problem I'm having with my k100. 

I ran the bike again until it died and then checked the plugs.  They were dry and what I would consider a perfect burn.  I also looked in the gas tank and confirmed that the return line was supplying fuel to the tank right up until the engine konked out.  I also checked for spark and it definitely has a nice strong spark. 

So I've changed the filter, pump, HES, cleaned all the frigging electrical connections (including the coils since I was there today), checked the plugs, flow of fuel, pressure.  What the hell is it??

I've included photos to cover demands from people that have asked for them.

Also, there seems to be a clear connection to the bike being hot or cold.  I know that tomorrow morning I can start it up and ride for a few miles, but once it has shut down, it has to completely cool off before being ridden again.  Does this cancel out the notion of the Fuel Pressure Regulator?? which would operate whether hot or cold??

How sure are you that the new HES you got is a good one?

They do go wrong and it may be that the one you got is also wrong. Similar thing happened my son's car 2 years ago, a valve failed and he got as new control panel. turned out he 'new' panel had the exact same fault as the old one. Proved very frustrating. As you say the heat and HES go together and all I can think on that is use a hair dryer to really warm up the HES then try a start. Then when you get a failure use a freeze spray to cool it and see if you get it to restart. Failed temperature sensor doesn't cause it to cut out, but will prevent a restart until it has cooled.

Did you put new Hall Sensors in or a used set which could be as bad or worse because it looks to me after reading your last post that your problem is most likely the Hall Sensors
New sensors are quite cheap but you need to fit then to the plate yourself.
Holister bought a bundle some time ago and should be able to tell you where to get some.

Here's the original thread I posted with links regarding the purchase of HES and their fitment.

I suggested earlier to check your valve clearances. Here's my post on a similar issue I had. Its quite a long thread over a long period of time so this is a summary of the problem and the solution which was adjustment of valve clearances. This is also a heat related problem. Valve clearances should not need adjusting very often and my PO had had the shims done only 12,000 km previously by a BMW dealer. Three of the exhaust valves where so far out of spec that I believe they were not measured correctly in the first place. This resulted in the valves not closing properly when the engine got hot causing the engine to stall. In the service record only one shim was replaced and that one was still well within spec.

Does the hose from the filter in this pic go to your pump?

I'm pretty sure i've eliminated the hall sensor as a suspect and now think my original hall sensor was still good.  I first tried a second hand sensor that I was able to buy locally which provided no change, so believing it to also be bad,I invested in a new hall sensor with lead from an outfit in Germany.  The problem still persists.  I have also tried cooling the hall sensor down by using a can of dust removal upside down to send out a cold refrigerant spray.  No effect, so I'm convinced it's not the hall sensor.

I guess I'll have to check out the valve clearances but I'll be amazed if that's it.

Also, Holister, yes, the fuel line does go to the pump.  I followed the arrows on the filter for flow direction, which would lead away from the pump.

Just my five cent.

1.) Could it be the ignition key: Help | Ignition intermittent of, intermittent no start ?

2.) Could it be the Ignition Control Unit?

Cheers

That unit gave me a similar problem on my K100LT. But, one major difference was that when I tried to restart there was no starter function when I pressed the green button.

It ended up with 88 [Will] offering me a swap with one to test his out but by the time I factored in 2 sets of postage swaps it was easier to get one from Motorworks as I had pretty much eliminated every other possibility. [It may well have been a pin connection because the replacement was fine and the one I took out worked fine when I tried it on the RT].

It led to us deciding to meet up and then he suggested posting up the meet here to see what happened. We got about 11 Ks out for a breakfast at Stradbally Fayre out of which grew the Irish Eccentrics, kindly named by Comberjohn.

The Ks and the Forum are life changing!!

Started it up this morning and of course, it fired right up, but only ran for 2 MINUTES before dying, so not hot, not the Hall sensor, not the temperature senor, not the valves.  What the hell is it????

I think we're all feeling your pain Jimmy but don't be so quick to dismiss valid possibilities just because you think the engine didn't get hot.

Exhaust valves are actually the hottest part in the engine and would certainly heat up significantly within just a minute causing the valve clearances to change. This happens under normal running conditions but if the clearances are below spec to begin with, a small change will make a big difference to an engines ability at lower rpm ie; the valve may not close properly and you'll lose nearly all compression.

The temp sensor can't be ruled out until you've done the test outlined in the troubleshooting page. Pull the plug on the FICU under the seat and measure the resistance between pins #10 and #13 counting from the lead end. You should measure 2.5 kilohm at 20^oC and a much lower resistance with a warmer engine. There's a graph to correlate that against.

Unless you test and measure, its all just speculation and that's neither reliable nor productive. You need to have a methodical approach to troubleshooting ruling out the most likely and sometimes easiest possibilities with evidence first. The temp sensor and valve clearances are both possible causes and easy to check.

Thru your posts you've said that it will start again right away after stalling, but you've also mentioned that it needs to cool right down before it will start...??? This brings into question how you tested the HES with a can of compressed air. I think in this case it would be better to try and replicate the fault (make it stall) by heating the HES with a hair dryer. IMO the HES has not yet been totally ruled out.

After that I would clean the injectors and install new kits.

I checked the resistance of the 10 and 13 sockets on the the plug of the FIC unit and am getting a reading 1580, and looking at the graph, that seems low. It's very hot here today, around 30 centigrade, but the bike has not been started and is sitting in the shade of my garage, but say ambient is around 25 degrees.

So how does this transfer into useful information?  Does it indicate the sensor is bad, and if so, can that stall an engine?

Holister, injectors are recent replacements without about 3k on them.  And to clarify my earlier comments, which I obviously got confused about, the engine will start again after stalling but only run for a few seconds however, it sometimes will run for longer when completely cold.

That's slightly low for a cold engine at around 25^oC.
What you need to do now is get the engine up to operating temp (when the fan kicks in at somewhere around 100^oC) or as hot as you can get it when it stalls. Measure those pins again and have a guess at the temp. You should be reading somewhere around 300 to 600 ohms at a temp of 60 to 80^oC

When she stalls, is it while the engine is idling or will she keep running with a bit of throttle?

I tested the fuel pressure today in parallel with a brand new oil filled gauge and should have been getting a reading of 36 psi, but I was getting a reading lower of around 29 psi.  As fate would have it, the engine never stalled, even after letting it run for more than 10 minutes and revving the engine to 3000 rpms for a minute, where as since my roadside breakdown, it's only been running for around 2 minutes max.  humph. 

It could be your tank breathers are blocked, with a full tank problem will appear fast as it wont take long to build up a vacuum in the tank which will starve fuel.

But if you ran with tank open i guess not!!

Blocked breather will not cause a drop in fuel pressure. The pump makes the pressure it has no reliance on atmospheric pressure. That theory is good for carbs but not EFI.

I would be replacing the FPR as that is very suspect with a drop in pressure like that.

My guess would be that the gauge pictured above is not accurate enough. If you're looking for a reading of around 36psi, it's only just off the mark on that scale of 0-500psi and it could be 10psi higher or lower depending on which eye you had open or whether you had your head tilted left or right.

This is the type of gauge that should be used with a range of 0-100psi.

I've been consulting with Crazy Frog, and the drop in pressure is due to the fact that the injectors are spraying into the cylinders while measuring, thus lowering the pressure, so the shown pressure is within norms while running and the FPR has been ruled out.  Looking to test the injectors electronics to see if there's a cut in voltage prior to shutting down by putting a 12v LED in line before the injector.

An update.  Still experiencing the same issues, but going through a series of tests to try and exclude potential issues.  I  put a 12v LED in line on an injector and also connected an induction timing light to a spark plug cable (one from each coil in two different tests, to try and eliminate ignition as the problem). 

I ran the bike until it died (which took almost 30 minutes of running time) and the timing light continued to flash right to the end on both tests connected to a cable running from each coil.  Trying to restart after the shutdown also showed the timing light flashing, so it's getting fire.

However, the 12V LED connected to the injector flashed normally up until shutdown where it fluttered a bit, and on trying to restart, it would flutter and then not light up at all, so it would seem that something is possibly cutting the power to the injectors.

Any ideas??  Any thoughts on what to test next?  Does this indicate the EFI computer is bad??  I thought these never went bad!!

If I need to change the EFI computer, does a 1985 RS model require a specific box?  Is there anyway of knowing whether a used box is good or not?

I have not replied to your latest PM. I was thinking about the problem.

Replacing parts one after the other could be very expensive and it's not the proper way to find a problem.
This troubleshooting has been a challenge, but everything had been done methodically.
To come to this conclusion, fuel pressure, ignition and fuel injection had been monitored in real time until the engine died.

By using the induction strobe light (It's the first time that I was thinking of this solution  ) in conjunction with the LED on the injector, we ruled out problems with the electronic ignition. The monitoring of both devices clearly shows that FI being the problem and we can safely assume that the computer could be the culprit.

Bert

Thanks for your input Bert.  Would any EFI computer work or would I need one that came off an RS model?  My bike is a k100 RS 1985.

Any way of testing a second hand computer to verify if it's good??

Hey!

Any EFI will do, that is for a K100 2V LT RS RT
Buy one that has warranty and you are able to return.

BMW K 100 RS STEUERGERÄT STEUERBOX MOTORSTEUERUNG CDI RECHNER K100RS #R560
https://www.ebay.de/itm/221650963907

Cheers

jumpinjimmy wrote:Any ideas??  Any thoughts on what to test next?  Does this indicate the EFI computer is bad??  I thought these never went bad!!

If it's any consolation, your computer wouldn't be the first to have malfunctioned. There are several examples at MOTOBRICK.COM where replacing the computer, or repairing broken connections within it, restored full function.

I would first consider the connectors/connections to and from the injectors. That type of intermittent behavior is typical of something loose or corroded and while is very possible that the cause could be something like a cracked solder joint in the FICU itself, its much easier to rule out external possibilities first before going to the trouble of replacing the unit.

Power to the injectors is thru a connector under the tank on the left. Make sure its clean and fitting firmly.
The injectors are earth switched by the FICU. Make sure the FICU plug and all earth connections are clean and secure. Check resistance between FICU #13 and battery neg terminal. It should be very low at around <.05Ω

I have cleaned the contacts already, but will check the resistance.  Thanks for the input.

It would seem that my trusty bike is now FIXED!

YES!


and Chocolate!  it was the Ignition Control Unit.

kennybob wrote:Posted on August 2:

If it is starved then the plugs would be dry, but if flooded then wet plugs.

If the FPR fails closed due to leaking vacuum tube then the rail pressure will be too high, which will cause a rich mixture condition (flooding/wet plugs).

Even bypassing the FPR completely (fails open) a k bike will run.

The main culprits to produce a lean mixture (dry plugs) are air leaks, or defective temperature sensors, rusty fuel rail and clogged injectors/screens, or a defective jetronic. (aka Fuel Injection Control Unit)

You had the answer all along, and could have had it fixed in 3 days instead of 3 months, if you would have just inspected the spark plugs--you can lead a horse to water...

jumpinjimmy wrote:It would seem that my trusty bike is now FIXED!

YES!

and Chocolate!  it was the Ignition Control Unit.

Great to hear it is running again  
Thank for the feedback!

Cheers

sorry Kennybob, but checking the plugs wouldn't have helped diagnose the failure of the electronic ignition module.  and btw, I did check the plugs if you would have read my earlier posting which in fact included photos of the sparkplugs.

I was quietly working with Jumpinjimmy on this one, and instead of replacing all the parts, we tried to understand what was the problem.

Ignition never fail, but what seems to fail was the signal from pin #8 from the ignition computer to pin #1 of the FI computer.
As a result, the signal to the injector was fading to the point that they stopped working.

In order to troubleshoot the ignition/FI, both were monitored at the same time.
An ignition timing light was installed on one of the lead for the spark plugs, and an LED was installed on Fuel injector. when the bike was ready to die, the LED was fading, but the ignition was still strong.
This is how the origin of the problem was diagnosed.

CF

And I'm extremely grateful to Crazy Frog for his attention and patience on helping me work through this problem, and also everyone else for their valued input and thoughts. 

Thanks again all.

Thanks for the synopsis about the means to successfully diagnose this condition, Crazy Frog. 

Unscrew the box and see if the duff component reveals itself? It's only four screws in the front.

yes, I had that idea and here's the photo.  Anybody see anything wrong??

Nothing obvious there - one of the big power switching transistors is showing mild signs of overheating but I'd expect that. How about pulling the other four screws and letting us have a look at the circuit board?

Next time I take the tank off I'll do just that and post it here.

Circle shows the general area and arrow points to where a track has possibly been burned across

Sorry, I misread the earlier posting--this is the old defective box and I can take the four screws to expose the other side of the circuit board.  I'll do that and post a picture here.