K75 hesitation when cold

[gregfeeler]

My 1992 K75S with 59,000 miles has developed a hesitation when cold. From a cold start (with or without a couple of minutes warmup) you have to be pretty easy with the throttle or it will hesitate. If you get somewhat aggressive it will bog and stumble. "Cold" in this case is about 65 to 75 degrees. Once warmed up, it runs great. It is also a bit balky to start when cold. It starts best by cranking it a few time, then laying off the starter until the fuel pump cycle ends, then crank it a seconded time with a bit of throttle (all of this is using the "choke" at either half or full position). When warm it starts in standard K-bike fashion. Last tank of gas with mixed town and highway (spirited) riding I got 45.5mpg.

Things I've done or checked:
- air filter clean
- air bleed screw adjustment checked
- valves all in spec
- crankcase vent hose as new
- air leaks around injectors checked with WD40 spray test
- plugs like new and gap correct
- two tanks of non-Ethanol gas run with 1oz./gallon of Techron
- when hot idle CO at 2.5% and at 4,000rpm at 0.5%
- engine temp sensor resistance correct when cold at pin #10 of the FI computer connection (should verify both sensor and wiring)
- TPS position correct
- fuel filter changed 12,000 miles ago, and the bike runs great at speed
- FI computer contacts cleaned with contact cleaner

So, assuming this is a fuel problem, what besides the engine temp sensor controls the cold running mixture? What else should should I check?

 

[tommcgee]

What do the spark plugs look like? They're kinda like litmus paper for engine diagnostics.

 

[LOSTBOY]

Was the CO adjusted with clean oil in the crankcase? If the oil has many miles on it, hydrocarbons rise through the crankcase vent, joining the intake stream. This makes accurate adjustment impossible. You can watch this phenomenon if you run your EGA from cool to hot engine; instead of getting lower CO as temperature rises, it gets higher.

 

[kentuvman]

Hi Greg,

Mine, an '87 K75s

Have you checked for any cracks in hoses?

You mentioned you were in your air filter box - I'd been in there and didn't correctly connect the 90 rubber hose that connects the air meter to the intake plenum - maybe not completely connected or maybe a crack in the hose?

In my case, I couldn't get bike to run past idle.

 

[gregfeeler]

Was the CO adjusted with clean oil in the crankcase? If the oil has many miles on it, hydrocarbons rise through the crankcase vent, joining the intake stream. This makes accurate adjustment impossible. You can watch this phenomenon if you run your EGA from cool to hot engine; instead of getting lower CO as temperature rises, it gets higher.

New oil and filter: maybe 200 miles on the oil.

 

[gregfeeler]

Hi Greg,

Mine, an '87 K75s

Have you checked for any cracks in hoses?

You mentioned you were in your air filter box - I'd been in there and didn't correctly connect the 90 rubber hose that connects the air meter to the intake plenum - maybe not completely connected or maybe a crack in the hose?

In my case, I couldn't get bike to run past idle.

I was not in the airbox other than what's required to change the air filter. The crankcase breather hose is essentially new and in good condition. Good suggestion to check the airbox hoses. Thanks!

 

[gregfeeler]

What do the spark plugs look like? They're kinda like litmus paper for engine diagnostics.

They where all basically white with just touch of dark gold-colored deposits at the tip of the insulator. The #1 cylinder was a tad bit darker than the other two. That made me think about a possibly leaky injector, so when I put the plugs back in I swapped #1 and #3 and I'm due to pull them to check the color.

 

[deilenberger]

So, assuming this is a fuel problem, what besides the engine temp sensor controls the cold running mixture? What else should should I check?

There is an air-intake sensor on the input of the air-flow-meter.. and I'd be looking at the AFM anyway - it's getting rather elderly, and it may be very slightly binding when it starts to open. Make sure the flapper door moves with no effort through it's entire range.

As mentioned above make certain the BIG hose between the AFM and the intake plenum is intact and both hose clamps are tight. That's tripped up more then one person with these sort of symptoms..

 

[gregfeeler]

There is an air-intake sensor on the input of the air-flow-meter.. and I'd be looking at the AFM anyway - it's getting rather elderly, and it may be very slightly binding when it starts to open. Make sure the flapper door moves with no effort through it's entire range.

As mentioned above make certain the BIG hose between the AFM and the intake plenum is intact and both hose clamps are tight. That's tripped up more then one person with these sort of symptoms..

Good suggestion. I've been thinking that since I've never checked the AFM it would be good to check it for free operation. That would give me the opportunity to look for any hose/air leak issues. A buddy of mine says he was able to remove the non-removable cover from the contacts on his AFM for cleaning, and then glue the cover back on. He said it ran better afterward, but I'm not sure if he had symptoms or was just being thorough.

 

[ncstephen]

yall are the experts here.

It has the feel of running lean which might be caused by a serious vacuum leak somewhere. Once warm things would be OK. Starting might be a little off. Gas mileage should still be OK. Plugs might be a little lighter in color. Slower speeds would be the most affected I would think.

Good luck. Give us a progress report.
NCS

 

[mieczkow]

Same for me....

I'm having identical issues with my 87 K75S. A little balky when cold, runs fine when hot. I am not having any starting problems - cranks instantly. Did much the same checking as Greg - have new plugs (old plugs looked good), new crankcase vent hose, etc.



I'll pull the air cleaner out and clean it although that was done not very long ago and I'll check the flapper on the AFM.

Tom Mieczkowski

 

[gregfeeler]

I'm having identical issues with my 87 K75S. A little balky when cold, runs fine when hot. I am not having any starting problems - cranks instantly. Did much the same checking as Greg - have new plugs (old plugs looked good), new crankcase vent hose, etc.



I'll pull the air cleaner out and clean it although that was done not very long ago and I'll check the flapper on the AFM.

Tom Mieczkowski

Post back your results if you would. I haven't have a chance to check my AFM, but I have looked at the hose from the AFM to the plenum and it looks good. Check for air leaks between the plenum and throttle bodies next.

 

[mieczkow]

Valves OK

I checked the valve clearances this morning and all are OK. Tomorrow I will take it down to my shop space and pull of the air box and see if I can find anything there. After I put everything back together today it exhibits exactly the same behavior - boggy at low throttle when relatively cold, absolutely responsive once it warms up. Darn!! Can't quite figure out what is going on!

 

[mieczkow]

Still No Diagnosis: 1987 K75 S

Well, today I took off the air filter (I have a K&N), cleaned it, checked all the hoses, checked the air flow device flapper valve - everything seems to be fine. I also checked (and cleaned) the rubber bushings between the air box and throttle bodies and they are fine.

So I am still mystified - continues to stumble at the first flick of the throttle and bog until the rpm's get up there or when the old girl gets good and hot. Once it's hot it runs just fine.

Am I finished and now must I pass it (and my checkbook) along to a dealer or pro BMW mechanic?



Tom Mieczkowski

 

[gregfeeler]

Mass air flow sensor theory...

Well, today I took off the air filter (I have a K&N), cleaned it, checked all the hoses, checked the air flow device flapper valve - everything seems to be fine. I also checked (and cleaned) the rubber bushings between the air box and throttle bodies and they are fine.

So I am still mystified - continues to stumble at the first flick of the throttle and bog until the rpm's get up there or when the old girl gets good and hot. Once it's hot it runs just fine.

Am I finished and now must I pass it (and my checkbook) along to a dealer or pro BMW mechanic?



Tom Mieczkowski

Well, Tom, I have a theory about what is going on with your bike and mine: the mass air flow sensor is out of adjustment and causing our bikes to run too lean off-idle.

I checked the CO on my bike at my dealer a couple of weeks ago, and got 2.5% at idle (spec is 2.0 +/- 0.5%), so no problem there. However, at 4,000 RPM it was at 0.5%. To make sure their machine was accurate, I stopped at one of the automotive emissions inspections stations, and much to the amusement of the operator had him test my bike and got 2.77% at idle and 0.53% at 2,500 RPM. Ideal CO off-idle should be 1.0% to 1.5%, so my bike is clearly running lean. That could/would explain the hesitation coming off idle when it's cold and the engine is most sensitive to a lean condition. That would also explain why my bike is running warmer than I think it should.

If this is the problem, then the mass air flow sensor needs to be cleaned and adjusted. It's possible to remove it from the air box housing, remove the cover plate from the electrical contacts (and then clean them), and then to also adjust the tension of the spring that positions the air flow plate. This is documented in a great article by Rob Lentini on the IBMWR K-Bike tech section: http://skylands.ibmwr.org/tom/tech/l-jet-cal.html. Key to understanding what's going on here is that there are *two* adjustments to the CO setting: the standard idle adjustment made with an Allen wrench, and then the adjustment of the air dam tension spring. The idle adjustment makes very little if any difference to the running mixture. Supposedly a correctly setup mass air flow sensor will always deliver a correct running mixture - but not if it's dirty, worn, or the spring tension is off due to age.

Lucky for me, I have another K75 in my garage that runs great cold. I have no idea how the CO is set on it, but I know it runs perfectly cold, so I'm going to swap the top half of the air box (with the mass air flow sensor) into my K75S for a quick test. If that fixes my symptoms, then I will get the CO checked on that unit to validate that it's OK across the range and do some riding tests to get a baseline of how the K75S should run. Then I'll take the mis-adjusted original air flow sensor up the the dealer and using their CO meter go through the adjustment process. It will be later this week (maybe the weekend) before I can do this, but I will first chance I get as I've now got a theory and want to test it ASAP.

 

[mieczkow]

Let me know....

Greg,

Sounds like a good theory. Let me know how the swap plays out. I did notice that BMW wants more than as thousand dollars for a new air flow meter!!!! That's a great motivator to learn how to adjust it...

Tom

 

[deilenberger]

One comment on the air-flow-meter.

The AFM used on the K75 is the same one used on the K100 L-Jetronic bikes. Identical. I believe it was also the same one used on several 1.8L BMW car engines of the era.

http://www.realoem.com/bmw/partxref.do?part=13621460447

As such - it would tend to give a lean mixture due to the smaller volume of air - unless this is corrected in the L-Jetronic analog brain. One thing that has been noted is that the flapper on it doesn't respond as quickly to a change in engine RPM (when the throttle is opened) on the K75 as on the K100 engines. That may account for the propensity of the K75 to run leaner particularly at small throttle openings and when cold.

Just a heads up on it. There is no good way to measure the output of the AFM that I know of. It forms part of a resistive bridge circuit, and if I remember correctly uses multiple tracks and wipers with steps in the tracks and discrete trimming resistors. I know Brian Curry at one time spent some time trying to come up with a definative test for the AFM and I believe he wasn't successful.

The best test I know of for a L-Jetronic AFM is replacement with a known good one (expensive if you don't have a spare kicking around.) It will be interesting to see the result of Greg's test.

One other thought - a quote from Rob's writeup: "The only reason that you should really have to go into the air flow meter is to correct for relaxation of the return-to-zero spring inside. With age, its preload weakens resulting in a richer mixture that can manifest itself by black sooty smoke from the exhaust. " - and I think he was correct. Most problems with the AFM tend to result in a richer mixture, not leaner. The symptoms that described in this thread are those I'd associate with a lean mixture.

 

[gregfeeler]

One comment on the air-flow-meter.

The AFM used on the K75 is the same one used on the K100 L-Jetronic bikes. Identical. I believe it was also the same one used on several 1.8L BMW car engines of the era.

http://www.realoem.com/bmw/partxref.do?part=13621460447

As such - it would tend to give a lean mixture due to the smaller volume of air - unless this is corrected in the L-Jetronic analog brain. One thing that has been noted is that the flapper on it doesn't respond as quickly to a change in engine RPM (when the throttle is opened) on the K75 as on the K100 engines. That may account for the propensity of the K75 to run leaner particularly at small throttle openings and when cold.

Just a heads up on it. There is no good way to measure the output of the AFM that I know of. It forms part of a resistive bridge circuit, and if I remember correctly uses multiple tracks and wipers with steps in the tracks and discrete trimming resistors. I know Brian Curry at one time spent some time trying to come up with a definative test for the AFM and I believe he wasn't successful.

The best test I know of for a L-Jetronic AFM is replacement with a known good one (expensive if you don't have a spare kicking around.) It will be interesting to see the result of Greg's test.

One other thought - a quote from Rob's writeup: "The only reason that you should really have to go into the air flow meter is to correct for relaxation of the return-to-zero spring inside. With age, its preload weakens resulting in a richer mixture that can manifest itself by black sooty smoke from the exhaust. " - and I think he was correct. Most problems with the AFM tend to result in a richer mixture, not leaner. The symptoms that described in this thread are those I'd associate with a lean mixture.

Don,
As usual, you add a lot of insight. However it's happening (sticky pivot shaft?), I think 0.5% CO is way too lean, which could cause the symptoms I'm dealing with. Going to swap AFM's today and will report back with results.

 

[mieczkow]

Got an AFM on eBay

Greg,

I ordered an AFM (used) on eBay from a running K75 which is said to be good. I expect to get it in a few days. After I install it I'll report back on how the bike behaves. Hopefully this will take care of the problem. Oh, and test your theory too.....



Tom Mieczkowski

 

[deilenberger]

Don,
As usual, you add a lot of insight. However it's happening (sticky pivot shaft?), I think 0.5% CO is way too lean, which could cause the symptoms I'm dealing with. Going to swap AFM's today and will report back with results.

Greg,

I'll be very interested in hearing what the results are. My old brain cells started reminding me of some CO readings I took on several K bikes many years ago.. and in general, it seemed they were running under 1% CO with the throttle opened (but not under load.) If this happens on a K where the intake system is known not to have any leaks - this might also indicate a change in value of some component in the L-Jetronic brain (resistors do "age".. only very good mil-spec ones do it really slowly), and after 15-20 years a change wouldn't surprise me.

Way back when - when Elsie Smith was preparing to run the IronButt on her K100 - an IBMWR list member and I had an interesting discussion on K bike mixtures. The list member told me that he had increased the performance of Elsie's bike, and gotten it to run cooler by adding a resistor in line with the engine temperature sensor, fooling the L-Jetronic into thinking the engine was colder then it really was (so the L-Jet brain richened the overall mixture.) He claimed not only better power but also considerably better fuel mileage (sounds counter-intuitive, but it isn't if the engine is running more efficiently).

I actually had purchased an engine temperature sensor that I was going to modify so I could easily plug in a resistor or potentiometer to try to tune the CO mix. Inertia of the ass kept me from ever completing the project. I had access to a dyno with CO sensor that I was going to use to try tuning it at speed and under load.

Let us know how it works for you..