I have been reading several threads, both new and old,  on the forum about K bikes running too rich looking for clues about a current problem with my 1983 K100. Mathew suggested I open a new thread to describe this .

Since I dropped the bike on October 2019 the bike was sitting waiting for a few parts and more recently waiting for the tank to be repainted. When I got the tank back the painters had lost the fuel pump from within. After a month of searching the paint shop they have provided me with a new pump, mounting and filter which are now fitted. I also have a new battery.  Other than this the bike is essentially as it was before the drop. The temperature sensor was new about 18 months ago.



Finally this last week I had a chance to ride again. The bike started fine but was not happy at idle. I fitted new NGK spark plugs and went for a 200 km test. This included about 60 km of high speed expressway. All went well except:
#1 Popping on overrun
#2 Poor idle (sounds rich) and if stopped at a red traffic light very poor power and spluttering pulling away as if flooded. However at speed pulled well.
#3 Noise from fuel tank

When I arrived home after a hot fast run I pulled the"new" plugs. All very black, maybe more so cylinders 2 and 3

I guess I first need to check fuel pressure and resistance through the temperature sensor and it's wiring. Alternatively could the tank noise be from the fuel return line non return ball valve causing higher pressure at the injectors? I also had the injection control unit out of the bike so the large connector might be suspect.

I will work through these possibilities. Any further suggestions? Any ideas gratefully received !

I'm thinking maybe too early to say it's running rich.  Is it the right pump,  or any old pump. That's a long bow once again. I'm calling long bow cause one of my bricks had an AU Falcon pump flapping around but worked fine. I'm sposing the regulator has more to do with the pressure than the pump. Maybe add vacuum to your checklist.

Hi,just a thought,having had the tank off is it possible the (choke) cable has moved,i would disconnect it and ensure the rail is closing completely

Thank you gentlemen.

Firstly the pump is the Quantum K100 kit, not a BMW part but I assume correct for the bike.  Agreed, fuel maximum pressure should be controlled by the regulator unless there is a restriction in the return line from regulator to fuel tank.

The "choke" lever works normally and the throttle body butterflies all return to their stops as per normal.

In a long shot it sound like a failing temperature sensor.
But first, take off the big connector of the EFI, and connect it well again. Then run the engine.

Check the return line from the FPR they can collapse on the bend into the tank, non OEM are more prone to this. The Ozzie summers with their excessive high temperatures can contribute to the collapsing. A restricted return line will cause it to run rich. Also check the FPR pressure which should be 36 PSI.
Regards Martin/Matthew.

I forgot to reply to your one way valve question. You could bypass it with a temporary hose from the regulator through the tanks filler opening, to see if it then works well. Or just clean it regardless. Mine was gummed up with rust when I got it. It was easy to clean, but is probably a different type than yours.

1983 K has a vacuum sensor beside the fuel pressure regulator, its about the size of a spark plug. There is a vacuum line off no 1 throttle body. Make sure that vacuum line is good or put a vacuum cap on it. I think recheck all the vacuum lines and caps are good as loss of vacuum means rich running too.

The more knowledgeable here tell me that vacuum sensor advances the ignition timing.....my 83 K100RS also has the engineering screws on the cam cover and is a lot sharper than my 84 [doesn't have them] or 92 K100. They all totally dislike 15% ethanol gas and very much dislike 10% ethanol gas and it all seems to have 5%.

More thanks!
The vacuum lines and caps are all looking good but I  am yet to remove tank and aerobic to see behind the fuel rail where presumably the heat is most damaging.

That one way valve on the tank return.... My bike is very early production, #701, so presumably has the early spring loaded ball valve. I cannot blow air into the tank by mouth thru the valve but I  guess if the spring is still in place that would be enough to send me blue in the face. I will need to siphon fuel from the tank before  I  can investigate further. Being retired I  am far too busy right now to make more progress

My fuel pressure gauge predates injection  systems. 15 psi will not do the job!  I will get a 100 psi gauge early next week.

Michael Sydney wrote:That one way valve on the tank return.... My bike is very early production, #701, so presumably has the early spring loaded ball valve. I cannot blow air into the tank by mouth thru the valve but I  guess if the spring is still in place that would be enough to send me blue in the face.

Overread you have the old K.
I recently bought a new valve, should blow through with a little pressure.
They won't cost a leg, replace it and be happy

Success! Rich running solved.
I checked resistance of the temperature sensor circuit my measuring the pin at the multiplying under the seat at injection control unit. Well with specs.

I ordered online a 100psi liquid damped fuel pressure gauge. Together with a tee piece and some odd bits of those I discovered a steady 100 psi at the fuel rail inlet! I wonder if that injects three times correct fuel volume?

Clearly a problem with pressure regulator or that one way valve on the tank return.

I found that by rerouting the fuel return to the tank via the filler cap dropped the pressure to around 30 psi. Blocked one way valve!
It was jammed solid with rust and debris.the ball and spring were beyond saving so reinstalled as unrestricted.
Check again and all is good. Running sweetly.

Looking forward to a run through the hills tomorrow 

good solve ol mate 

have a good ride

Thanks Charlie
I am just astounded the bike would run with about three times normal fuel pressure. Does that mean it was injecting tripple fuel? It is an open loop system with no feedback to the processor  if over rich. 

Consumption was  high but no problem with power above 2,000rpm

Anyway, the parts for my "test kit" cost less than $30. Will be useful again I  am sure.

Does getting over 100 psi from the pump [the stock pump only gets to about 60 psi] make any difference?

92KK 84WW Olaf wrote:Does getting over 100 psi from the pump [the stock pump only gets to about 60 psi] make any difference?

I believe the standard pump is capable of achieving over 100 psi if it encounters a blockage such as this. There is an over-pressure valve built into the pump to limit it to this.  The maximum pressure capability of the pump is not relevant however.

In normal use the fuel pressure regulator constantly bleeds off excess pressure ( and returns fuel to the tank) to maintain approx 32 psi in the fuel rail which is critical to ensure the injectors deliver the correct fuel load. 

In the case of my bike the regulator simply could not return fuel to the tank so pressure rose to excessive levels. Since the engine management opens the injectors for a specific period of time to suit demand, as communicated by the air mass sensor, excess pressure allows too much fuel to enter the inlet ports in that time compared to the quantity at 32 psi, hence the rich running, excessive fuel consumption and sooty plugs.

These older K100 have a very primitive fuel injection system relying on pre-determined  mapping to deliver fuel. Later system (on later K bikes) have a closed loop system which constantly monitors the exhaust gas and adjusts the "mixture" accordingly.

-Just a qualifier on closed loop. Not all of the fuel mapping in the ECU relies on closed loop control. I discovered this when using a Dynojet/Power Commander on a boxer twin fitted with Motronic. Generally, the 'extremes' of the various parameters e.g. full throttle, high revs, sudden throttle changes etc cannot rely on closed loop because the systems just can't (or couldn't) hack it. The closed loop areas of the maps tend to be in the steady throttle, cruising areas. I guess this state of affairs has evolved with improvements in FI, but certainly the early Motronics were like this.

Remember, catalysts and closed loop control were brought in for emissions control, not performance, and those emissions & economy tests were artificial, steady running affairs.

Excellent investigation! Do you know what kind of fuel consumption you were getting while this was going on?

Great you managed to solve  it

Does that mean it was injecting tripple fuel? It is an open loop system with no feedback to the processor  if over rich.

No.  Simplistic equations for unperturbed linear flow in pipes give flow rate as proportional to square root of pressure. So more like 70% more.  But I would expect turbulent flow to be even more non linear. That rapidly gets into complex fluid dynamics , Navier-Stokes equation and no simple equations with nice simple rules.

I'm a little surprised the plugs were so different since the pressure must be nearly the same for all. You may want to get the injectors cleaned/checked.

I'm wondering if the painters gave you a wrong pump. If the system is blocked, the relief valve in the pump should open at 60 psi. Maybe that's being fussy as it doesn't matter if the regulator and the rest is OK, I spose. But it's good to know that nothing went pop at 100psi.

If you did read 100psi, maybe it was even higher, since your gauge only reads up to 100.

Hi, a few things that BMW did to my 1985K100.
Disconnected the TPS. Balance the vacuum across all intakes, you will notice the vacuum is only on the rear cylinder at moment. Take a reading before and after, I saw them do it and i could see the difference. PS some mentioned choke, there is no choke, the lever just lifts the the revs. Good luck

daveyson wrote:I'm wondering if the painters gave you a wrong pump. If the system is blocked, the relief valve in the pump should open at 60 psi. Maybe that's being fussy as it doesn't matter if the regulator and the rest is OK, I spose. But it's good to know that nothing went pop at 100psi.

If you did read 100psi, maybe it was even higher, since your gauge only reads up to 100.

Thanks for all the helpful suggestions. It was good to solve the problem but certainly interesting to contemplate the details.

Regarding the  pump, The painters lost the original pump and supplied a brand new replacement of unknown origin. I fitted it and rang the bike but a) noticed a strange noise from the tank and b) realized it was running rich. I then fitted a brand new Quantum pump that I had obtained for my other K100 (A work in progress), but had same symptoms.

When I got my test gauge I believe I got a reading of a little over 100psi, but that was certainly towards to upper limit of the gauge capability. I believe that the pressure was limited to this by an internal pressure relief valve within the pump. I have subsequently discovered that the pressure regulator works fine by opening at approx 32 psi but could not function due to a blocked return line spigot.

Now, I understand that the new replacement pumps are capable of generating much higher pressures that the older pumps. Maximum pressure from older pumps may well have topped out at approx 60 psi but I have read that current pumps can in fact generate more:

The Fuel Pump This type of high-pressure fuel pump (Fig 1.0) is called a roller cell pump, with the fuel entering the pump and being compressed by rotating cells which force it through the pump at a high pressure. The pump can produce a pressure of 8 bar (120 psi) with a delivery rate of approximately 4 to 5 litres per minute. Within the pump is a pressure relief valve that lifts off its seat at 8 bar to arrest the pressure if a blockage in the filter or fuel lines or elsewhere causes it to become obstructed. The other end of the pump (output) is home to a non-return valve which, when the voltage to the pump is removed, closes the return to the tank and maintains pressure within the system. The normal operating pressure within this system is approximately 2 bar (30 psi), at which the current draw on the pump is 3 to 5 amps. Fuel passing across the fuel pump's armature is subjected to sparks and arcing; this sounds quite dangerous, but the absence of oxygen means that there will not be an explosion!



https://www.autoequipment.com.au/image/data/Support/EFI%20part%201.pdf


Obviously at higher pressures the pump can deliver lower volume to the fuel rail, and at higher pressure the injectors, for a given opening pulse time, probably have difficulty handling vastly greater flow. I do feel, however, that with the high pressure at the fuel rail my fuel consumption was approximately doubled!

just to add a note , pumps are rated for flow volume and pressure expected , so you might find a replacement pump may be rated for a higher flow per minute  at rated pressure higher than expected on a k bike ....lets guess a v8 pump rated FOR almost double the fuel consumption  through the system , this pump may also be capable of much higher running current (amps ) to make the effort to produce  the volume as designed ... its all a matter of relevance .

but in your kbike application it may be just cruising running the 2.5 bar and lower volume for the bike  , but maybe slightly higher current than your connectors may be rated for  , just keep this in mind and check, clean, whatever, the tank connector periodically looking for signs of over-heating , distortion of any of the connectors .

Also, I'm thinking you still have a problem, since the relief valve failed to open at 65psi. It might be seized. You might be better off putting the pump from the painters back in.

The original pumps are also the roller cell type. It's a safe bet they could also provide enough flow to allow 100psi, that's why they have a relief valve, to lower it to a maximum of about 60.

Let's say the pump is capable of allowing 120psi, the fuel pressure regulator reduces it to about 30, the relief valve reduces it to a maximum of about 60, in case of a blockage.

I think you should be entitled to a replacement pump, in which case you could consider that the pressure gauge has pretty much paid for itself.