K75 hesitation when cold

[mieczkow]

Installed Replacement AFM

I received the used AFM today and installed it on the K75. Seems to have done the trick. First start-up and ride was without hesitation. I'm letting it get cold and will try again tonight when the engine is about as cold as it's gonna get in Florida tonight.

I hope this does the tick as I am scheduled to take off for the Blue Ridge Parkway on Thursday morning!

Tom

 

[gregfeeler]

Progress report

Here's on update on my progress:

THE PROBLEM:
To recap, the bike with the cold running problem is my 92' K75S with about 59,000 miles. The problem was when riding away with a cold engine it would seriously bog and try to die and then catch and go and then bog again ("bucking") as you opened the throttle unless you were extremely gentle with it, and even then that didn't always work. Once warm it ran well.

SWAPPING AFMs:
I also have a perfect running 1990 K75 standard with almost identical mileage. So Tuesday I took the Air Flow Meter from this bike and installed it in the S. This improved the cold running so that it wouldn't bog and try to die - it was just hesitant and slow to respond to the throttle. The time it took for this to go away was also reduced.

I put the S bike on a CO meter and found that at 2,500 RPM I am still getting the same 0.5% mixture with the "new" AFM as the original one. From Don's excellent article on the IBMWR web site (http://www.ibmwr.org/ktech/dyno/index.shtml) I was expecting closer to 1.0~1.5%. So either K75s run leaner than what is considered normal (and what a K100 does) or I have some air leak I haven't found. At idle the new AFM was just above 1.0% CO whereas the original was 2.5% (and the bike idled slower as a result) so I adjusted it to 2.0%.

RESULTS:
Now the S bike has a very slight "reluctance" when cold for the first three or four minutes. It's a half-block from my driveway to the first stop sign, and a block to the next, and by then this reluctance is all but gone. The improvement I got after swapping the AFM is even better now, although the problem is not completely gone. Once warm, it runs better now than when I started: excellent transition from idle with lots of power and responsiveness.

At this point I was thinking there was something wrong with the original AFM - like dirty contacts inside the sealed box area, or sticky air door pivots (although nothing felt wrong when moving it by hand). As a test I installed it as is into the donor 1990 bike. Other than a bit faster idle (to be expected since the idle CO went from 1% to 2.5%) it runs great. Good cold running and perfect warm running. So, apparently there is nothing wrong with the original AFM.

REMAINING QUESTIONS:
However, if the old AFM is good, then what explains the improvement by swapping it for another good one? Also, it appears that idle CO from 1.0% to 2.5% is not super critical. Getting it to 2.0% sweetens the starting and running but doesn't seem to be the cause of this cold hesitation problem I started with. I suppose there could have been dirty contacts on the AFM socket plug that swapping cleaned up, but even though the cold running on the S is acceptable now, it should be better, so something is still going on.

Finally, is 0.5% running CO "ok" for a K75? Don's article on the IBMWR K-Bike tech pages (and Rob Lentini's: http://skylands.ibmwr.org/tom/tech/l-jet-cal.htmlhttp://skylands.ibmwr.org/tom/tech/l-jet-cal.html) both say that as the AFM's internal spring ages they will run richer, which is clearly not the case here with two different AFMs. Just to verify that 0.5% is not something unique to the S bike, I'm going to take the other K75 to the shop on Tuesday and check it's running CO now that it has the original AFM from the S bike. Maybe I still do have an air leak down stream of the AFM on the S? AND, from what Don and Rob have tested and written about, it sounds like there was a big drop in power and torque going from 1.5% to 1.0% running CO on Don's K100, so would I get a big improvement if I could get to 1.5% on this bike?

 

[deilenberger]

Greg,

A thought. Once the bike is started - try pressing the starter button in. This enrichens the mixture by what Rob Lentini said was about 10%. The starter won't engage once the engine is running.

If at idle the speed increases your idle mixture is a bit lean. If the idle speed decreases - too rich. The ideal response is a slight increase in speed (maybe 100-150RPM for about 3-5 seconds) then a slight decrease in idle speed. This was pretty much how Rob described adjusting the CO screw on the AFM. It worked well for me on many bikes.

You might also try depressing the starter button when you feel the "bog" you're describing transitioning from idle to run speed. If the richer mixture improves things, then I'd start thinking about adding some inline resistance to the engine temp sensor in order to overall richen the mxture.

One consideration with the L-Jetronic bikes - this is not a closed loop system. The fueling map was done at a time when gasoline wasn't polluted with ethanol, which could well have the effect of leaning the mixture out. Newer systems with an O2 sensor will self-adjust the mixture, the K bike L-Jetronic has no way to do that.

As usual - YMMV..

 

[gregfeeler]

Greg,

A thought. Once the bike is started - try pressing the starter button in. This enrichens the mixture by what Rob Lentini said was about 10%. The starter won't engage once the engine is running.

If at idle the speed increases your idle mixture is a bit lean. If the idle speed decreases - too rich. The ideal response is a slight increase in speed (maybe 100-150RPM for about 3-5 seconds) then a slight decrease in idle speed. This was pretty much how Rob described adjusting the CO screw on the AFM. It worked well for me on many bikes.

You might also try depressing the starter button when you feel the "bog" you're describing transitioning from idle to run speed. If the richer mixture improves things, then I'd start thinking about adding some inline resistance to the engine temp sensor in order to overall richen the mxture.

One consideration with the L-Jetronic bikes - this is not a closed loop system. The fueling map was done at a time when gasoline wasn't polluted with ethanol, which could well have the effect of leaning the mixture out. Newer systems with an O2 sensor will self-adjust the mixture, the K bike L-Jetronic has no way to do that.

As usual - YMMV..

Don,
It's funny you mention this approach as I was thinking along the same lines. Once I check the idle and running CO measurements on the "good" bike with the AFM from the S bike, I should have a much better idea about what is "correct", but the starter button test could be very handy. I also have both bikes set up with a switch for the altitude plug so I can lean them out instantly, if needed. The problem with open-loop semi-mechanical systems like this they are not self-correcting. The good part is you can sometimes brute-force them. It's even occurred to me that with sufficient effort and a know correct AFM one could build an air flow test stand to test and calibrate the AFM units. Other than being a bunch of work, it would be a solution for bikes that are pretty old now, and maybe no one would care that much.

I knew Rob - he was just leaving the MOA Board when I started in 2002, and I'm still impressed with the quality of work and body of information he left behind. A great guy.

 

[Bobmws]

I also have both bikes set up with a switch for the altitude plug so I can lean them out instantly, if needed. QUOTE]


I'm sure you checked that you haven't accidentally hit this switch, or it has shorted closed, but it would explain the sudden onset of your problem.

 

[gregfeeler]

I also have both bikes set up with a switch for the altitude plug so I can lean them out instantly, if needed. QUOTE]


I'm sure you checked that you haven't accidentally hit this switch, or it has shorted closed, but it would explain the sudden onset of your problem.

Good point, Bob, but yes, I did check that out. It would be a lot like the "my bike won't start because I left the kill switch on" routine.

 

[mieczkow]

Long Distance Test Starts Tomorrow

Well, so far so good with the replaced AFS - it still is very slightly balky when stone cold, but in less than 5 minutes that disappears entirely. I'm heading out for Asheville and the Blue Ridge Parkway tomorrow - expect to be there by Friday afternoon. I'll send in a report on how the old girl is performing with her sort-of new (new-to-her) sensor.

Weather looks to be great over the next 5 days!



Tom

 

[gregfeeler]

Well, so far so good with the replaced AFS - it still is very slightly balky when stone cold, but in less than 5 minutes that disappears entirely. I'm heading out for Asheville and the Blue Ridge Parkway tomorrow - expect to be there by Friday afternoon. I'll send in a report on how the old girl is performing with her sort-of new (new-to-her) sensor.

Weather looks to be great over the next 5 days!



Tom

Tom,
Have a great ride, take pictures, and let us know how the bike runs out!

 

[gregfeeler]

Swapped AFM results

Yesterday I got a chance to test the CO readings of my 1990 K75 standard which now has the unmolested AFM from my "S" (problem) bike. The CO meter was on the low scale (0~2.5%) and initially at idle it pegged! I switched to high range for just a moment and it looked like it was between 3.5 and 4.0%. After a lot of CCW turns and revving and returning to idle I got the idle set to 2.0%. At 2,500 to 3,000 RPM the CO is in the 0.9% to 1.1% range. That same AFM in the "S" bike was at 2.5% CO at idle and about 0.5% at 2,500 RPM. So it is clearly running richer than when it was in the "S" bike.

The AFM from the 1990 standard bike once install in the "S" tested at 1.0% idle CO, and also 0.5% at 2,500 RPM. Looks to me that I have an air leak on the S (or something else I can't imagine) that is causing it to run lean. Also, I think this might have been the case for a long time (before I got the bike) as the original AFM appears to have been adjusted to a correct CO at idle, which turns out to be way rich in a "correct" bike.

I need to check that the crank case port for the fuel tank evaporation is correctly blocked off, perhaps temporally crimp the crank case vent hose that goes to the air box, and then carefully examine the hose from the air box to the intake manifold. All the plugs have the same identical color, so a leak around one throttle body or injector seems unlikely. It has to be something that would affect the fuel mixture overall.

 

[98lee]

Greg,

I don't see anywhere where you verified that the injectors are flowing correctly.

MR. INJECTOR http://www.mrinjector.us/ in Coeur d'Alene will clean, replace filters and o-rings, and flow test them for $16 ea. with a one day turn around or he has pre-reconditioned and tested ones that he can send out immediately.

Might be worth a try.

Have you pulled out and tested your 12,000 mile old fuel filter to totally eliminate it as a possible problem?

 

[gregfeeler]

Greg,

I don't see anywhere where you verified that the injectors are flowing correctly.

MR. INJECTOR http://www.mrinjector.us/ in Coeur d'Alene will clean, replace filters and o-rings, and flow test them for $16 ea. with a one day turn around or he has pre-reconditioned and tested ones that he can send out immediately.

Might be worth a try.

Have you pulled out and tested your 12,000 mile old fuel filter to totally eliminate it as a possible problem?

Lee,
I've checked the plugs and they were all identical for tip color and over all condition, so I've kind of ruled out a bad injector for the moment although having them checked wouldn't be a bad idea. I've pretty much come to the conclusion that I have a low fuel pressure problem - either the pressure regulator or fuel filter. I'll have access in about a week to a fuel pressure gauge set up for K-bike testing, but tonight I did change the fuel filter and checked everything inside the tank.

On the test ride, for the first time since this started, the bike road away from cold perfectly. I'm going on an early morning breakfast ride with the local club tomorrow and it will be about 45 when we leave, so this should be a good test. I need several tests to see of tonight's results are valid. The old filter only had about 4,000 miles on it, but it wouldn't take much water from a bad tank of gas to semi-plug the filter. Just the right amount of restriction could leave it borderline lean where it would run "ok" when warm but starve for fuel when cold. Maybe a long shot, but we'll know more tomorrow. Thanks!

 

[98lee]

Greg,

It IS possible for the injectors to become evenly obstructed. Quite common on bikes that have the fuel pump vibration damper starting to dissolve.

So did the old fuel filter fail the blow test (empty, let dry a bit, LIGHTLY blow through indirection of flow, should be virtually NO resistance)?

Again, if its plugged, check the pump vibration damper.

 

[gregfeeler]

It was the fuel filter!

Greg,

It IS possible for the injectors to become evenly obstructed. Quite common on bikes that have the fuel pump vibration damper starting to dissolve.

So did the old fuel filter fail the blow test (empty, let dry a bit, LIGHTLY blow through indirection of flow, should be virtually NO resistance)?

Again, if its plugged, check the pump vibration damper.

It was the fuel filter. I left this morning in 40 degree temps and put on over 200 miles with several stops long enough for a good cool down, but every start from cold the bike ran perfectly. I'd let the old filter air out since last night and it had a lot of restriction blowing through it. I had another "good" used filter to use for comparison and it had almost no restriction like you said.

When I changed the filter I checked out the tank and saw no indication of any kind of contamination and I gave the fuel pump vibration dampener a good look. I suspect water contamination of the filter. I cut it open tonight and couldn't see any accumulation of visible particles, however, the filter material was pretty dark - this after just 4,000 miles.

This is a new one on me. I would have thought it really long odds to have a filter blocked *just* enough to cause lean running but not enough to also cause fuel starvation under WFO riding, but apparently that is just what I had. All the other things I did in tracking this down were not wasted, though. The bike runs now better than it ever has, and I *know* that many things are correct and I learned a lot. Thanks to everyone for their help!

 

[pglaves]

 

[gregfeeler]

BTW, I have to thank Paul for suggesting changing the filter as a prelude to doing a fuel pressure regulator test.

 

[deilenberger]

I'll just add that there have been LOTS of cases where people say "I just changed the filter.." and the replacement filter was the cause of continuing problems. That used to be SOP on the IBMWR list when someone complained of a poor running K.. "Have you changed the filter?"

A brand new filter can be instantly trashed IF you haven't cleaned out of the tank whatever it was that trashed the old filter. The filters on a K bike will probably last about forever if you never get a dirty or wet tank of fuel (as proof - when I bought my K75S it was about 15 years old.. it still had a filter dated 06/86 in the tank - despite the former owner being charged for a new filter about every 6 months.. Once it dried out, I could blow through it with no restriction at all. It had about 60,000 miles on it.)

New BMW cars do not have a filter change interval. They are now "replace if needed." I generally do mine at 100k just sorta for good luck. I have never needed to change one. I've probably been lucky (and a bit careful) in never getting a bad tank of fuel. There is also no recommended change interval on my R1200R fuel filter (and it's one of the few hexheads that has an actual filter..)

 

[gregfeeler]

I'll just add that there have been LOTS of cases where people say "I just changed the filter.." and the replacement filter was the cause of continuing problems. That used to be SOP on the IBMWR list when someone complained of a poor running K.. "Have you changed the filter?"

A brand new filter can be instantly trashed IF you haven't cleaned out of the tank whatever it was that trashed the old filter. The filters on a K bike will probably last about forever if you never get a dirty or wet tank of fuel (as proof - when I bought my K75S it was about 15 years old.. it still had a filter dated 06/86 in the tank - despite the former owner being charged for a new filter about every 6 months.. Once it dried out, I could blow through it with no restriction at all. It had about 60,000 miles on it.)

New BMW cars do not have a filter change interval. They are now "replace if needed." I generally do mine at 100k just sorta for good luck. I have never needed to change one. I've probably been lucky (and a bit careful) in never getting a bad tank of fuel. There is also no recommended change interval on my R1200R fuel filter (and it's one of the few hexheads that has an actual filter..)

It's been my impression from changing a lot of K-bike fuel filters and doing the "blow" test that I was probably throwing money away since notwithstanding as you said unusual contamination or water, they would almost never plug from normal use. So I keep the "good" used ones in the tail compartment of all the bikes - just in case.

As it turns out on the bike in question, although it was changed only 4,000 miles ago, that was just before I rode it to the Gillette rally. On the trip I bought gas at some out of the way places. I had the mindset that a water logged filter would expose itself with fuel starvation symptoms at higher speeds and WFO throttle. The idea of one plugged just below that threshold never occurred to me. But, as I think back now, I remember the fuel pump sounding like it was making more noise than normal. Big lesson learned for this boy.

I'll say this. I changed the fuel filter on my K1200RS this summer as part of the Level II service "just because", and God help you if that sucker plugs up on the middle of nowhere. While you can change a Klassic K filter in five minutes, it's a major process requiring a special spanner to change the one on the K1200 brick bikes.

 

[Bobmws]

I'll say this. I changed the fuel filter on my K1200RS this summer as part of the Level II service "just because", and God help you if that sucker plugs up on the middle of nowhere. While you can change a Klassic K filter in five minutes, it's a major process requiring a special spanner to change the one on the K1200 brick bikes.

Are you talking about the block of wood and plastic mallet to spin the plastic 'nut' from the bottom of the tank?!
I have some chunks of oak I can start mail ordering if I stamp "special" into them.......

 

[gregfeeler]

Are you talking about the block of wood and plastic mallet to spin the plastic 'nut' from the bottom of the tank?!
I have some chunks of oak I can start mail ordering if I stamp "special" into them.......

Exactly! Actually, I cheated - I borrowed the BMW tool from my local dealer (they like me there ), which makes the nut removal and re-installation pretty painless. It would seem like it would take a precision eye and hammer arm to get that nut just "guttentight" so it won't leak, get the fuel pump aligned in the tank properly, and not break the nut in the process using a wood block and a hammer. My hat's off to you!

 

[mieczkow]

Back from the Blue Ridge

Got back from my trip up to Little Switzerland. Bike ran perfectly - did not have any issues with the bogging or slogging - given that it was cold I used a little choke for the first 3 minutes or so of my morning starts. Great weather on the way up, great weather there, two days of drizzling rain on the way back.

I loved it!