Hello all, my first post, requesting some technical help and guidance please.

I’ve had my 1984 K100 for 13 years, but it has never given me any reason to learn the details of its operation – it has been virtually faultless. It has nearly 120,000 miles on the clock and has now developed a running fault.

In brief, it starts instantly as it always has done, but then dies after 2 or 3 seconds, sometimes less, occasionally a little longer, as long as you don’t touch the throttle. When running it sounds normal and all four exhaust down pipes warm up to the same degree.

The first time it did this, I had to just put it back in the garage as I was on a mission. The next day, having decided that maybe it was the injection ECU connector, I wiggled the same and tried again whereupon it burst into life. Unfortunately the next time I tried to start it the problem had returned. Since then, while troubleshooting at intervals over the last few weeks it has twice cured itself somehow: the first time I then went for a 30 mile ride without incident, but it failed again at the next start. I’m sure that again this was subsequent to very firmly seating the injection ECU connector, but I’ve tried that many times since to no avail. The second time it started and ran normally for a couple of minutes before cutting out again. Otherwise, the symptom is always the same: instant start and then die.

I have been guided in my trouble shooting mainly by the excellent document and interactive wiring diagrams created by Bertrand Vogel, as well as the existing posts for similar problems on this forum, though none seem to quite match mine, in the very short time the engine runs before it dies. I’ve also had some helpful suggestions from the UK BMW club forum.

I did find a couple of faults: an earth wire adjacent to the vacuum switch was broken from its terminal, and the clutch switch was not working. Repairing the wire and replacing the switch made no difference to the main problem.


The situation is as follows:

Fuel supply

It has a full tank of fresh Shell petrol – the last journey before the problem started was the mile back from the filling station (I should say “had”, because I got fed up of lifting a full tank and removed some). The pump runs on pressing the starter button and the fuel pressure is 2.5 bar. While the engine is turning the pressure oscillates between 2.5 and 2.3 bar, presumably as the injectors open and close. When the engine stops the pressure returns to 2.5 bar, the pump then stops and the pressure then drops away. I have a professional quality Muller Bem 10 bar pressure gauge intended for hydraulic systems, but it does the job. I have disconnected the return from the regulator and allowed it to discharge into an open receptacle: there is good flow as well as pressure. The fuel filter was replaced 13,000 miles ago (normal replacement interval: 20,000 miles ?)

All four injectors spray fuel when turning over the engine; the spray patterns look uniform.

Electrical checks

Following Bertrand’s procedure, I can confirm that:

The earths on the gearbox and under the tank are clean and tight. I have also cleaned the starter motor commutator and brushes. The starter was actually really very clean inside, thought the brushes are well worn – I’ll replace them later.

Both Hall sensors are working when checked using Bertrand’s LED test. I have another set which also test OK. I substituted this set on the engine: no change.

The voltages and resistances given at the ignition module and the injection ECU conform to Bertrand’s indications – only one is slightly different: at line 4 of the injection ECU I get only 10.4V with the starter relay disconnected and only 7.1V with the starter running, even though the battery voltage only drops to 10.5V. The ignition signal voltage of 20mV is present at line 1 of the injection ECU when the engine is turning. I used a Fluke 75 meter for the checks.

I have manipulated the wiring harness while carrying out continuity checks looking for a broken wire or bad connection – nothing to report.

I have substituted the following parts from my stock, without any change in behaviour:

Ignition controller
Hall sensors
Air flow meter (with its air temperature sensor)
Water temperature sensor

The original parts are now back on the engine.

I don’t have another injection ECU to try, but I have opened mine and did not find anything loose or traces of corrosion.

Both pairs of plugs seem to be sparking strongly.

With a strobe timing light connected to No 1 plug lead and a 12V LED across No 4 injector terminals, both systems seem to be working: I have both sparks and injection.

With the plugs out and in 5^th gear, ignition on, turning the rear wheel produces sparks at alternate pairs of plugs, pulses from the fuel pump and pulses at the injectors (tested with an LED across the terminals).

The pump stops and the injection relay opens after the engine has stopped. However, keeping the starter button pressed does allow the engine to keep running (albeit with a protesting starter motor). This would seem to be the principal finding and I understand that this richens the mixture as well as presumably providing an earth for the injection relay etc, though this does not seem to be needed.

An un-metered air leak? It doesn’t seem to fit the nature of the problem: either running perfectly (although rarely) or the start-and-die routine. Anyway, the breather hose at the back of the engine is nearly new and in good condition - I replaced it recently to cure a problem of poor running at low speed. I have carefully examined the air box above the butterflies, but have not removed it – nothing to report. I have tried spraying contact cleaner and also methylated spirits around the butterfly bodies while starting – no effect

Most things I’ve checked more than once but I’m getting nowhere. Can anyone see what I’ve missed? I would very much appreciate some help.

Anybody seen my reading glasses? Can`t find them.

George, something from my experience with same symptoms: follow the wire from right hand switch to the joint under the tank (on the right driving direction). Check on corrosion, clean, with some small screwdriver tight the female side of connector.

Thats exactly happened to me before cleaning that contact

Regards

Welcome George!

Found the reading glasses...........................Thanks to CF

Maybe not related to your problem, but the fuel pressure should stay put........after the pump has stopped.

This problem could be internal in the pump, or a malfunctioning FPR.

Have you tried to run the bike with the pressure gauge connected, untill it stops, to see if you loose the pressure just prior to the engine stops?

welcome george, you appear to be thorough with what you have done, but wishing to try the other options I would ask if you have afull set of keys hanging off the ignition key, the reason I ask is the ignition key can sometines play up just as you describe and eventually play up on each and every bump you hit. so can the stop/ kill switch if not used much, corrosion is what I'm getting at, perhaps a clean can eliminate these two from a list of possibles, did you mention the fuel pump connector at the wiring harness? cant remember and I just read it, if the connector is clean on the FI cpu then try the pump connector. It sounds like a high resistance joint going open circuit to me

Thank you for your prompt advice, which I have followed up in the garage

Nino, I disconnected , cleaned and tightened each female terminal in the connector block for the right hand switches. I forgot to mention in my original post that I had tried another (new) right hand switch assembly I have to rule out the kill switch. It made no difference, though I hadn't tightened the female terminals that time. No change I'm afraid.

Inge, on the fuel pressure, I connected my gauge again. I got a slightly lower reading this time: 2.3 bar, oscillating to 2.2 bar when the engine is turning, presumably as the injectors open and close. Yes, I have left the gauge connected while running the engine and the pressure stays up after the engine has stopped. Once the engine has stopped, the pump then stops (the injection relay can be heard to open as well) and the fuel pressure then drops away, taking about a minute to return to zero. Seems OK to me. I do have a new pressure regulator but do you think it is needed?

k75cster. I only ride with a ring and a plastic strap on the key, though we're not actually getting off the centre stand at the moment. For the kill switch I had tried another right hand switch assembly as reported above. I have a new ignition switch but it has a different connector (must be for a later model) so I have not tried it and don't want to modify it unless necessary. I cleaned the ignition switch using contact cleaner, but in case that was not adequate I then bridged the red and green wires at the connector under the tank to completely bypass the ignition switch. No change again. I also cleaned and tightened the terminals in the tank connector. It feels much more secure now, but no change to the problem.

I agree that it sounds like an open circuit, but it is remarkably consistent and you would think that it could be identified. As stated in my first post, the only reading I've found that does not conform to Bertrand's guide is the "starter running" information to line 4 of the injection ECU. I get 10 point something with the starter relay disconnected and only 7V with the starter running, well below battery voltage at the time in each case. The voltage drop occurs through the right hand switch gear - that is 12.4 goes in and only 10 comes out when the button is pressed, but confusingly is the same with the new switch gear.

I keep coming back to the fact that pressing the starter buton keeps the engine running, but without being able to translate this information into a solution. Any more ideas please? It is very reassuring not being completely alone - I was begining to give up hope.

The fuel pressure should stay a lot longer than one minute after the engine have stopped, but as earlier mentioned I don`t think it`s related to your problem.

But afterall it tells you that fuel pump or FPR needs some attention in the future.

I would take a closer look at the hall sensors and the related wiring, the hall signal is also used for closing the injection relay contacts as long the engine is running.

The starter signal overrides the hall signal in that part of the circuit, and as you say "when pushing the starter button the engine continue to run".

The problem has now been resolved with a replacement injection ECU. Thank you to everyone for the helpful comments and suggestions.

For the record and in case it is of use to anyone in the future, the logic that led me to this conclusion and the purchase of the ECU went like this:

First the situation again

The engine would start instantly but then die within a couple of seconds or so. On four occasions since this had begun, the engine had started and run normally. The first time I didn’t see how long it would hold, I presumed it was OK again and switched it off, then once for a 30 mile ride, once for a few minutes, once for about a minute.

Holding the starter button kept the engine running, but only at idle – it didn’t like the throttle.

For me this confirmed that the ignition system (Hall sensors, control module, coils) was working correctly. The Hall sensors tested OK and the plugs produced sparks, tested by removing the plugs, earthing their bodies and running the starter, or with a timing light in series.

Also, the injection system was doing something. The pulsing voltage at the injectors could be seen with an LED across an injector terminal and their operation heard through a screwdriver at each injector body. With the injectors removed, running the starter produced pulses of fuel from all four injectors, with good spray patterns.

With the plugs out and an LED across one of the injectors, ignition on and in 5^th gear, turning the rear wheel produced pulses from the fuel pump, sparks at alternative pairs of plugs and flashing of the injector LED.

The fuel pressure was correct and with good flow as well.

I had checked all of the readings at the ignition module and injection ECU connectors for the expected values and continuities from place to place – nothing to report.

I had nevertheless substituted the parts I have, to no effect: hall sensors, ignition module, right hand switch assembly, air flow meter (with its air temperature sensor), water temperature sensor, throttle switch. I had shorted the ignition switch connector under the tank to take the ignition switch out of the equation.

I had cleaned and tightened every connector I could find.

The earths (gearbox and frame under the tank) were clean and tight. I had also removed and cleaned the starter commutator and brushes.

Back to operating the starter switch with the engine running.

I understand that this does two things (apart from running the starter):

• provides an earth for the fuel injection relay to close it and power the fuel pump, injectors, air flow meter and throttle switch, via the ignition module,
  
• richens the mixture, normally for starting, by means of a 12V signal to the injection ECU.
  

I had noted before that the fuel injection relay opens after the engine stops, not before, but was unsure whether this was timed in any way. I substituted a switch for the relay so that I could control the power supply. It made no difference.

Therefore it was the richening of the mixture that was providing the positive effect, meaning that it was too weak otherwise.

So, too much air or not enough fuel. For the air, the breather hose is recent and in good condition and I couldn’t find any leaks. I had tried spraying methylated spirit all round to no effect. I didn’t see how such a problem could cure itself on four occasions either.

The amount of fuel is determined by the injection pulse width, calculated by the injection ECU and I was left concluding that the fault lay there, though I had no way of checking it and had opened the ECU but found nothing loose or burnt or any signs of corrosion.

I concluded that the ECU was not using the engine running information correctly, due to some internal fault – I will now try to track this down with another internal inspection and perhaps application of the soldering iron.

thanks for the trouble shooting guide

This was same problem as I had and the solution was ultimately the same, replacement ECU [the one under the headstock].

I now think its the place to start when this happens. In the meantime I had stripped and cleaned out the starter, replaced the starter relay and the load shed relay and cleaned off all the earth connections and electrical connections under the tank.

Hi,

I know this is an old thread but I'm having the same issue with my 1987 K100RS.

https://www.k100-forum.com/t13459-another-hard-starting-problem#159039

I note that George says he replaced the ECU but it sounds like he replaced the Jetronic fuel injection unit under the seat. I also note that "92KK 84WW Olaf" had the same issue and replaced the ECU under the head stock.

George can you confirm what you replaced please?

Thanks,
Glenn

glennpm wrote:I note that George says he replaced the ECU but it sounds like he replaced the Jetronic fuel injection unit under the seat.

George can you confirm what you replaced please?

George stated that he replaced the injection control unit—not the ignition control unit—so I think you have interpreted his post correctly.

Thanks Laitch,

glennpm wrote:Hi,

I know this is an old thread but I'm having the same issue with my 1987 K100RS.

https://www.k100-forum.com/t13459-another-hard-starting-problem#159039

We had some members with this problem during the years, where the ground
connection for the FI relay coil at the ignition ECU is lost caused by a bad
ignition ECU.
I see that you have the possibility swap this (in another thread), I would guess
this would solve your problem.

Thanks Inge,

This confirms what I expected. From Bert's diagnostics and the wiring diagrams. An improper or no signal to earth/ground from the ignition ECU to pin#85 of the fuel injector relay, could cause this.

I'll report once I get my used unit and give it a go.

Glenn

After a search through the forum it seems like I have been mixing things up a bit and
it was ground for the starter relay coil (which also is grounded through
the ignition ECU) that was the problem most of the time (also for Olaf).

But it it won't hurt to give it a try.

One way to check is to make a temporary ground wire which is hand held.

I have to admit that I get frustrated with the time it takes to check things by removing and testing them out even though it is of course correct.

For that reason I have this box with a complete set of spare electrical bits, relays, FICU, ECU, fuel pump, regulator etc etc and with the help of the forum just swap out the suspected component. It can be tested later.  It does mean that if something is faulty when I go at it that there is a spare to hand. My last issue with this same problem was found to be a crack on the FICU body where the plug on the loom is clipped into. A swap solved the bike running problem, the crack was later repaired and eventually refitted in perfect working order.

My take on having Ks is build a little parts supply by buying cheaply on eBay before you need them. They will always sell on.

92KK 84WW Olaf wrote:
My take on having Ks is build a little parts supply by buying cheaply on eBay before you need them. They will always sell on.

That is a great plan for dealing with Murphy's law on at least 2 counts:

• If you don't need it today, there will be several available on eBay that sell cheaply (fairing parts are a classic example of this)
  
• If you need it today, there will be nothing available or the price/delivery options will be crazy.
  

Also 

• If you have a spare, the Murphy probability factor works in your favour and you probably won't need it
  
• If you don't have the spare to hand, the likelihood of failure of that component, and therefore the need for a spare, is immeasurably higher.
  

Works for me!

Yes good advice and I've got to start adding to my stash!

Glenn

Got her fixed!!

I got the used ignition ECU and the Jetronic FIU yesterday. The FIU came early than expected which worked out well. I cleaned all the contacts and then applied RailZip and let them dry overnight and slept like hell BTW ;-( thinking about "stuff".

So after all the testing, refurbished FI units, contact cleaning, etc., etc., it turned out to be a bad Jetronic unit. I'm wondering what made it quit? I put the ignition ECU in first and it was more of the same wouldn't run unless I kept the starter button in. Well then I put in the used Jetronic unit in all was well! Now its behaving like it always had and starts right up. I adjusted the throttle screws with my Motion Pro mercury carb synchronizer.


Thanks for everyone's input. This is a great site. Now I'm looking into installing a K1100 throttle body assembly!

Glenn