Hello everyone,

This is in regards to a 1992 K100RS. There are 100k miles on the bike, but only about 31k miles on the engine. Everything is fine and dandy when the bike is cold, but once it warms up, the engine starts to misfire. It's nothing that keeps it from being rideable, but it does cause noticeable drops in power when it occurs. It misfires both when revved in neutral and while in gear / moving. The misfiring goes away at about 4k rpm and up. The issue has been worsening noticeably over the past couple weeks.

Things that are known-good:
-valve clearances were done by a BMW shop ~2k miles ago.
-all rubber parts that could cause and air leak are brand new (2-3k miles). This includes the plastic boot with the O-rings underneath the TBs
-Air filter is brand new
-Just replaced some wires with melted insulation that went to the starter relay, and some of the wires around them (replaced the starter relay as well)
-Spark plugs are brand new (2-3k miles), I also recently cleaned them which helped this issue a little bit but did not solve it.

Things I am thinking:
-Is it possible this type of thing could be caused by a loose timing chain?
-I read around the forum a little bit on plug leads as a culprit, but that was mainly when under load from what I read. I assume this is still a possibility, and I think the ones on this bike are original, but not sure.
-Possibly something to do with the coils? Connections? Cracked coil casings causing them to arc?
-Fuel pressure regulator? When I got the bike back from the shop, the mechanic told me that the FPR gave a wacky vacuum reading at idle and low RPMs, but that once he got the bike out on the road and up to speed that it evened out. The fact that he said idle and low RPM (the range my issue is in) makes me think of this as the most likely culprit. I should buy a new one either way, but money's just that tight right now.

What gets me is that it's normal when it's cold and only starts to misfire once it warms up. Seems backwards to me.

Thanks in advance!!

Remove the small metal cover on the front of the timing chain cover and view the hall effect ignition sensors. They are known to fail when hot & be ok when cold. (However, they usually quit altogether when hot or aybe just 2 cyl. of the four)

Put a hair dryer to blow on the sensors and heat them up a bit, then start the bike cold to see if you can duplicate the failure.

not sure how much heat you need but probably want to get the area up to 200 - 250 deg. F or so to test.

Lot's of info on the forum if you search Hall Sensors or Ignition, etc.

Whatever occurs, please make a post about what was found when you are sure that you have solved it.

Regards, ibj...

Ohhhh boy, I hope it's not that. I replaced that about 5k miles ago with a used one. ($450 something brand new, ergh!). I'll see if I can get a hold of a heat gun or something and test that out. Thanks for the addition to the culprit pile!

"the mechanic told me that the FPR gave a wacky vacuum reading at idle and low RPMs, but that once he got the bike out on the road and up to speed that it evened out."

Never heard shuch twaddle in all my life. The 16v FPR does not have a vacuum connection as the 8v K100 does. He might as well tell you that the fluzgrutz in the oil foam accumulator needs a set of new bubbles.

I have just been through a remarkably similar experience and after months of spending time trying to find the problem it ended up being an TPS that was not adjusted correctly.
It may not be but it sure is worth a go to make sure its set correctly.

The TPS rational position can be adjusted by loosening the two bolts the hold it to the back of the throttle body assembly.

It should be positioned so that with the throttle closed and the bike turned on (but not running) you should get a reading of 0.375V between pins 1 and 4 of the TPS connector. Snap the throttle once or twice to make sure the throttle is fully closed.

FYI: It needs to be plugged in for this test. Just slip some thin wires into the connector when you connect it to monitor the voltage.

Give it a go and lets know how it is.

Thanks Rick, will give it a go in the morning. On another note, the FPR has a little hose connected from the end of the FPR to the forward-most TB, whereas the other TBs have the little rubber caps on them. When disconnected, it supposedly increases fuel pressure by around 8-10 psi if I remember right. -(something i've read on the forums). The '92 is not supposed to be this way????

All the 16v engines had a motronic 2.1 up till late 93 and as far as I know there was no vacuum line to the FPR but I will check it out futher. Maybe crow is on the menu.

Well I hope the feathers dont taste too bad, apparently they do have a vacuum line to the FPR.
Give the TPS adjustment a go and there is an adjustment for CO2 which is at the front of the electricals box. I have never played with one and do remember hearing that an initial value of 590 ohms but thats all I know and I cant find anything about it except that it has about 20 turns from one extreme to the other.

One thing I found is that there is the same part number for the motronic 2.1 ECU regardless of which model its on so now the brain is starting to get addled. I don't think this bears too much thinking. According to Real OEM my K1100 should have one and it hasn't and I know of one K100RS 16v that doesn't have one and that is the one I was going on but Motobins say that the vacuum FPR is all K100 and NOT K1100. I dare say that BMW has a very logical explanation for an illogical caper.

Haha, no worries. Thanks for looking in to it. Well, I'm taking a look at the TPS, and it was at about 0.359v, a little low. I know how to set the TPS by turning it, but (here comes the part I just don't get) do I set the idle screw first and THEN the TPS? or vice versa. Turning the idle screw moves the butterfly, thereby changing the TPS reading / position. But then on the other hand, turning the TPS changes the idle speed as well. How do I go about setting the TPS and idle screw together?

Loosen off the 2 screws on TPS first,check that it is not fouling the idle setting, then set idle speed and lastly reset TPS.

I remove the TPS when setting up,no need to but it's just me.

Okay, so TPS last. Couple question: So it doesn't even need to be there to set the idle? And what do you mean by "fouling the idle setting"? Thanks!

Loosen the TPS move it both ways, just check that you can move it without affecting the idle speed.

When I loosen the TPS bolts and turn it while the bike is running, it does indeed cause an increase or decrease in idle rpm depending on which way I turn it. This makes sense in the motronic system though as it controls mixture, yes? If not, what would that increase/decrease mean?

That's why I remove it completely, and the reason for my comment on affecting the idle speed.

I had a similar problem. We found that one of the plug leads was loose and shorting from where it plugs into the coil pack to the body of bike. This could not be seen untill we removed the coil packs. Hope it is that simple..

Ah I see.  Yeah, I removed it and set the idle screw and have got it all back together.  Same thing was going on, and there is likely an issue somewhere else as well.  The CO2 sensor on the front of the fuse box seems off to me.  I've never futzed with it before, but turning what looks like the adjustment screw on it seems to not do anything.  It also just keeps turning forever with no stop.

I checked all the plug leads to the coils, as well as the connectors that come from the harness to the coils.  Do you mean you couldn't see the ground behind the coil and that's what was shorting?

As of right now, I experimented a little bit with the TPS and idle screw.  The TPS is set to about .530v at closed throttle and the misfire issue is a lot better than it was before.  The bike rides a lot smoother too.  This tells me there something up elsewhere.  Any more ideas are welcome.  I'll take the coils all the way off and check the grounds back there now.

Rabidchiwawa007 wrote:

I checked all the plug leads to the coils, as well as the connectors that come from the harness to the coils.  Do you mean you couldn't see the ground behind the coil and that's what was shorting?

.

Yes..My bad..Just home from work.. . We could not see it or hear it. Only once we romoved the coils could we see the scoring from arcing.:

Wow

If you have a high voltage leak such as stated above, you may be able to see some blue sparks jumping or the glow from them if you idle the bike in a completely dark place with the spark plug dress cover off.
Also, you can enhance the failure if it is present by putting a few drops of dishwashing detergent in a spray bottle of water & mixing it, then spritz lightly around the coils, leads etc. The soapy water will enhance the path to ground........kinda like testing the system for a fault that occurs on a foggy morning/rainy day.
I haven't read back through all the thread yet, what's the condition of the coils themselves? Have they been Ohmed out and do you have the "orange Tower" ones?

I took the coils completely off and inspected them, along with all the plug leads. Everything looks to be in good shape, no cracks in the coils, clean contacts, etc. the coils look to be the original coils that came with the bike, meaning they have 100k miles on them. I haven't 'ohmed them out' as I don't know how to do so. By 'ohm them out', I assume you mean test them.

Ungas is talking about a K100 with L Jetronic injection where the TPS does 2 things WOT and Closed.
Rabid is talking about a Motronic 2.1 on a K100 RS 16v where the TPS monitors the actual position of the throttle butterflys at all times so as to be able to give the correct mixture.
You are very much at crossed purposes and I was wrong about a vacuum control on the FPR (we will laugh about that later).

As with the L Jetronic the Motronic 2.1 should have the butterflys closed to the stop set at the factory when at idle and air is getting in through the air bleed idle adjusters (the Big Brass Screws).
Idle is set after the vac is equalised by turning the BBS out or in equally and keeping the vacuum equal. So by that logic the TPS can be set at idle at any time in the process.
On a Motronic 2.1 the TPS is set by adjusting for a voltage of .375 volts.
When I adjusted my TPS on the K1100 I just turned the TPS till it stopped doing it's fut fut fut routine and then right on 3200rpm it would cut back to 4 cyl. It doesn't quite make sense to me that it was cyl 2-3 that were the ones missfiring. That is what I could see by looking at the plugs.

It was only doing it when hot because that is when it is lean and before it gets to that temp it is running on predetermined settings and not the readings from the various sensors. A bit similar to the closed and open loop settings the Motronic 2.2 has with the O2 sensor.
When the 2.1 drops back to idle it goes to the predetermined settings as does the 2.2 and that is why at idle it's OK and as soon as you open the throttle and load it by letting the clutch out it does the fut fut fut routine.

When my K1100 started the caper I immediately thought electrical and chased coils, hall sensors, connections and each time I played with something it would go away for a while then come back and I was so focused on the electrical thing I overlooked the temperature thing. While I played with it the thing cooled down.
Some here will remember the time I found the "faulty coil that didn't read faulty" no wonder it didn't it was OK after all.
I went round and round chasing an electrical problem that was not there. I removed the TBs and found they were not syncronised correctly and replaced them with a spare set then set them up using 4 dial gauges and made a jig to hold them.

When I eventually found that the TPS was not adjusted correctly everything that was happening suddenly made sense.
I only moved the TPS about 1mm max counterclockwise and all went well.
When I finally found what was wrong and fixed it, it was a different and much better bike to ride.

To put the finishing touch to this MOTOHANSA always did all the previous work on this bike from 61000km (when the PO bought it at auction from NSW Police) to 155000km (when I bought it) because the guy I bought it from worked very close to them and didn't know beans about it he just relied on the expertise of the major BMW dealer in Sydney. When I got the bike it was in an appalling condition and he was told to "sell it to the wreckers its stuffed" and they had caused the problem by not doing the work as scheduled or not doing it correctly, well both actually.

Hope this helps and please lets all try to make sure which model we are talking about as things get very confusing and have the model in our signature line.
I am just as guilty of this as anybody else.

Very good explanation by RickG. It seems he probably has the answer to your issue. You guys are both talking about the closed loop systems that I have a lot of experience with in automotive, but not on the K's. Mine have always been early K100s.

If I have ricks procedure correct, He backs off the throttle idle mechanical stop screw all the way, until the brass butterflies inside the TB are completely shut and then brings the idle speed up from there by opening the air bypass bleed screws?

That would be completely foreign to me as with my early K I set the bypass bleed screws with the butterflies held slightly open via the idle stop screw, hence giving me the ability after balancing to vary the final idle speed up or down by the stop screw.

I was told later, that the true final balance of the air bypass screws should be done at partially open throttle at 1500 RPM.

My source for that was a former BMW corporate employee.

In my particular instance, I had to resynchronize the mechanical position of the 4 butterflies relative to each other (not for the in-experienced tuner) before balancing with the bypass screws.

My TPS in comparison (as shown in Rick's post) is a simple matter compared to the closed loop K1100 that you guys have.

Interesting. That would mean the idle screw is completely null and the idle is set purely by using the bypass screws? Unless I'm misunderstanding that.

I do get a 'thud thud thud' sound at idle that sounds a lot like a butterfly out of sync. When I took the vacuum hose off of the front-most TB that went away, so I'm thinking I'll take a look at the bypass screws tomorrow morning. I have a single vacuum gauge, so I suppose I'll go through one by one and make sure they're all the same. I remember awhile back when I first bought that gauge I tested it out on this bike. When I got it hooked to the vacuum port on a TB, the needle didnt stay in one spot as it idled, but moved up and down within a certain defined area (of about a 1/2 inch on that gauge) very quickly, probably as the valve opened and closed. Is that how it's done, or am I missing something? (aside from 3 other gauges, one for each TB)

On the gauge sets you get small taps that almost close the airway so that the gauge doesnt react quickly.  I bought some taps for a system I made from an aquirium (fish) shop they are for varying the air flow through the bubblers that put the oxygen into the water. Irrigation supplies also have very small taps designed to trickle irrigate.
That id how I have always set the idle but I suppose that if it is a tiny bit open then you dont open the BBS as much for the same effect

Makes sense. I can probably just finangle something together by using a hose clamp to squeeze the hose shut. If not, the local pet store here opens at 9am! I'll post back results in the morning when I've done it.

possible to insert some kind of restriction plug (commonly called a "pill" in the gauge vacuum hose (maybe a piece of foam rubber, cigarette filter or the like) to restrict the rapid flow if you don't have the valves. You might also form some small balls of solder and drill a very tiny hole in each one to insert those in each hose for a flow restrictor (or some small pieces of dowel wood?
When I made my set, I did exactly as the previous post suggested and got miniature air flow regulator valves from the pet supply store. Unfortunately, I also used their fish tank silicone air hose for hook up at the engine, it proved too soft for vacuum use and tended to "suck flat" during the tests.
Regards, ibj...

Rabidchiwawa007 wrote:When I took the vacuum hose off of the front-most TB that went away,

It could be a idea with a flow restrictor at the FPR vacum hose also.

http://www.motobrick.com/index.php?topic=4498.0#msg28134

Inge, I think it went away due to the increased air coming from the open vacuum port.  But out of curiosity, are you suggesting essentially decreasing the vacuum to the FPR?

Well here's my solution: https://i.imgur.com/JpQqPr6.jpg
It's a little quirky and I had to be careful of moving the vice grip as it would change the reading I would get, but I got it done eventually.

I took a reading from all of them before adjusting, and there was about a 2 psi range difference across the 4 cylinders.  Not knowing what it was really supposed to be, I just averaged the readings and set them all to the average. Idle immediately got a lot smoother and better.

I messed with the TPS again as well after setting the butterflies.  It's at about 0.5v.  At 0.375v, it misses a bit more.  I do have one of those Harmann chips in the computer, and I think those are designed to send less fuel thereby smoothing out the power band and increasing mileage.  I just wonder if it doesn't want to cooperate with the bike and the 0.5v TPS is in fact just increasing the amount of fuel used.

A few slow revs from idle to 4k showed improvement as well.  There's a slight hesitation at 2k rpm, and then a little *tiny* bit of "fut fut fut" (within normality I think) up to 3k/3.5k. A good twist of the throttle shows no hesitation at all

All in all, it's a ton better than it was and I've probably done all I can do with my tools / skill set.  A new FPR is on my list as soon as I get some money in.

Sincere apologies for the misleading information posted yesterday, It's not often I'm right but wrong again.

I think Inge was suggesting dampening the vacuum signal to the FPR, not reducing the fuel pressure.
By restricting the vacuum flow, It slows the reaction time of the FPR in relation to the normal pulses from the intake valve port.

You are talking fuel pressure, correct? Not the reading from the vacuum port?

Although pulsation in the vacuum signal is totally normal, I haven't witnessed that causing pulsation in the actual fuel pressure.
Regards, Ibj...

Not wrong info, it still technically works just fine

I was referring to restricting the vacuum pressure to the FPR, sorry. I'll see if that makes any difference as well when I get a chance.

I should also mention that the bike seems to burn about a half quart of oil every 800-900 miles or so. I remember reading around online that it was okay for the K-series bikes to burn through a full quart in around 600-700 miles and that they were designed like that on purpose, for cheaper production. Is this true?

My 85 K100rs with 90,000 miles on it uses none. I go 5,000 miles between oil changes and never add any, seldom even look in the sight glass.
One thing though....I almost never leave it on the side stand if there is any way to put it straight up.

If you are getting a lot of big clouds of oils smoke after using the side stand, that could explain some use. You might try a different oil???? I'm using Castrol T4.

Can we assume you have NO leakage?

Regards, Ibj...

I have no experience with 1100's or 1200's

This is a '92 k100rs.  Yeah, there is no leakage, never any drips under the engine, no oily residue on or around the engine.  No big clouds of smoke, but if I give it a good hard rev you can see a little smoke shoot out the pipe.  Sort of looks like two-stroke exhaust with maybe a quarter to half as much smoke.

And I never leave it on the side stand, only center stand. Have also tried a few different oil brands and weights. Right now its got castrol 20w-50 in it.

That would be a bit excessive for my liking maybe a leakdown test and or compression test.
How many miles has it got on it.

valve stem seals ?

Bike has 100k miles. I swapped a different engine in when I bought it due to dead 1st cylinder. This engine has about 40k miles

If it had a cast iron cylinder liner I would say take it out to a lonely road and give it one hell of a thrashing you know ride it like you stole it and get the glaze of the bore.
Give it a leakdown and compression test it may show something up as I would not expect it to use that much. BMW does say in the manuals for the airheads that any oil consumption up to 500ml for 1000km is acceptable and OK but not in my books is it acceptable.

If you have a substantial amount of miles on the current set of spark plugs, an inspection can help reveal some things.
As oil burns off, it leaves a fine tan sandy (ash) deposit build up on the firing ends of the spark plugs.
If a specific cylinder is burning oil, that plug will have more deposit's building up around the electrode and inside edges than the others.
I would think that an engine burning that much oil would accumulate a definite viewable excess of sandy ash on the plugs.

On an engine that burns a lot of oil, (not necessarily yours) that sandy ash builds up until it finally "bridges" the area between the center electrode and the earth tang above it. At that point the plug stops firing and immediately, the dry brown sandy deposit gets fuel/oil saturated and comes out looking black & oily.
Years ago when technology was not as good as today, I had lots of events where a customer would come in with his V-8 running on 7. I would pull the 1 contaminated plug, chip away the build up of ash & crud built up on the electrodes, spritz it off with carbi cleaner, hit it with a blast of air and put it back in.......Good as new till the next time that cyl (or another one) bridges a gap.

If you find only 1 of the four that has significantly more ash built up than the other 3, that would indicate which cylinder at least.... has a problem.
If it were defective valve seals I would imagine that given the proper procedures & tools, the valve stem seals could be changed without removing the head. I don't know if it's possible on this engine or not.
As valve seals go, oil seepage is usually more severe on intake valves than on exhausts, because they operate in a vacuum.
So much so that some engines are actually designed with valve stem seals on the Intakes only.

At any rate, your oil consumption although more than desirable is certainly manageable and would not warrant a tear down.
Regards, Ibj...

OBTW.....My Castrol T4 is 10W40.

Great info ibj, thanks. I cleaned the plugs maybe a few weeks ago and they were slightly ashy-looking and all of them were the same. The plugs were brand new about 2k miles ago. I was just thinkin of excessive oil consumption as a reason for misfiring.

With that relatively small quantity of oil useage, I would not expect it to be a mis-fire issue.
You might think about changing to the T4 oil as it is thought to contain more zinc as called for with these engines.
If you do so, you could also try the 10/40 as opposed to the 20/50.
At least just for a trial next time you do a change.
Also, keep the oil level slightly below the center dot in the window as opposed to slightly above. It would be my guess that the lower the level is in the crankcase, the less it will use.

Hi guys

re. "The CO2 sensor on the front of the fuse box seems off to me.  I've never futzed with it before, but turning what looks like the adjustment screw on it seems to not do anything.  It also just keeps turning forever with no stop."

Inside the CO controller plastic casing is a tiny, 1000 ohm, multi-turn potentiometer. Its dead easy to turn the slider of this off either end of its screw thread if you use a normal workshop screwdriver. Once off it is difficult to get it back on and the Motronics 2.1 will drop into 'get you home' mode, (maximum rich). This will make the bike run unevenly.

To fix it, take the CO controller off and set a DVM to 'resistance' and connect to the centre and one end connection. Use a small screwdriver to rotate the external screw back and forth until you see the resistance change (which means the slider is back on its thread) Then set the resistance to about 500 ohms (the midpoint) and put it back on the bike.

From memory these things are ten turns from end to end, so from the mid-point its best to count turns to avoid running off the end of the resistor.

You should see a significant changes while adjusting. Rotate CLOCKWISE TO WEAKEN.

I didn't have access to a gas analyser so I turned the screw to find the highest tickover speed then enriched it by half a turn. I don't think its a million miles out, the tickover is smooth and the engine has a smooth pick-up and plenty of power.

Hope this helps...

Thanks Gerry for that very handy bit of info about the gizzmo going all the way to nowhere and falling off. I have never touched mine as I have never felt the need but finding that out makes me want to check that on mine as I have been having problems that aren't still quite right. And the history of a real mess with anything that could be adjusted was done wrong I would suspect that it is also wrong.

What Rick said, awesome to know. I've been sick the past couple days and haven't had a chance to do anything. Feeling better and better, so I'll check it out soon. What do you mean by "connect to the centre and one end connection"?

Three connections on a potentiometer, 1, 2, 3:

     1~~~~~~~~2~~~~~~~~~~~~~3

Where ~~~~~ is a long wire-wound resistor.

Between the end connections 1 and 3 you have 1000 ohm resistance. The centre connection 2 is a slider which normally moves up and down the resistor on a threaded rod. When its in the middle you have a resistance of about 500 ohms between 2 and each end. Turning the screw moves 2 and changes the resistance between 2 and the end contacts. As resistance [1,2] goes up [2,3] goes down etc.

You need to find the pair of connections which gives a reading of 1000 ohms. These are the 'end' connections. The remaining connection is the slider 2.

When the unit is broken (turned off one end or the other) connections between 2 and either end will be 0 ohms or 1000 ohms and turning the screw will not change anything.

Connect the DVM to 1 and 2. Then by judicious fiddling with the screw its possible the re-engage the slider on the screw. You know this has happened when the resistances start to change as the screw is turned.

The alternative is a new unit.

Good luck...

ps. this might be useful.

http:\\nearlyfreesoftware.com\pics\k1100_motronic_2.1_by_Frank_Warner-1.pdf

Got it, thanks so much. Sounds easy enough

Update: finally got the CO potentiometer off, and I believe I've found the culprit! One side read 4 ohms, the other side 1014 or so. Going to go work on it now, and if I can't fix it, I'll buy a used one on ebay for $65-70. I'll update again when I've got it back together. Excited!

Sounds hopeful

another update: I'm learning (a lot). Vehicle electrics like this are all magic to me.

Using this link: k1100_motronic_2.1_by_Frank_Warner-1.pdf that was provided above in this thread by someone, I went to the motronic connection and did the following:

"1. Measure the resistance from pins 31 to 25, this should be 1k ohms +/- 10% and is the value of the pots resistance.
2. Measure the resistance from 28 to 25, this is the slider to the bottom resistance.
3. Measure the resistance from 28 to 31, this is the slider to the top resistance.
4. The slider resistances should approximately add up to the total measured value of the CO pots resistance. If either or both of the slider resistances are well above or below the CO pots', it is faulty - get a new one."

SO, I did these measurements and got the following:
-Pins 31 to 25 = 724 ohms (total for CO pot)
-Pins 28 to 25 = 1246 ohms (slider to bottom side)
-Pins 28 to 31 = 456 ohms (slider to top side)
1246 + 456 ≠ 724 at all.
-I need a new one.

From what I've learned from this thread, the CO pot can cause unevening running similar to what I'm experiencing with the misfiring. As such, I'll order a new one (probably used on ebay? because I'm a poor college kid) and pop it in, adjust a few things, and see what happens. I'll report back here when I get that done, probably in a week or so. I'd love a confirmation that I have done these procedures correctly and ordering the new CO pot is the correct course of action.

HUGE thanks in advance for all the help on this thread!