What you see when the ignition is on but the engine is not running is correct (36 psi).

Needle vibration is caused by pressure drop when the injectors open.

This may be a result of fuel supply constriction not allowing fuel flow to keep up with demand. Could be caused by any of the problems listed in my last post plus blocked fuel filter if you haven't changed it. Very important to install it the right way round.

Injectors are turned off during deceleration hence no needle vibration. 28 psi is too low, should stay at 36 psi (I don't think the pump is off).

Best,

ps. Be careful of the pipe to the fuel pressure gauge, you've fuel at 36 psi in there, if it comes off at speed you've got a serious fire hazard.

Testing the pressure between the fuel regulator "outlet" and the return port to the tank, yields NO information. That line is just a dump line flowing freely back to the tank for unused fuel. Should be little or no pressure there. What pressure might show there would be from whatever restriction might be created by the return backflow one way ball check valve. Discontinue taking any pressure readings on the outflow side of the regulator.

Pressures on the rail side (inlet side of the regulator) seem correct. I believe the pulsation is normal due to vacuum pulses in the signal hose from the throttle body, as well as the pressure drops from the injection cycle.

Again, consider it normal & move on. The test is complete. You passed.

(Some autos actually have a fuel pressure dampener installed on the end of the pressure rail to smooth those pulses).... Not required on our bikes.

Time to look elsewhere for the miss or performance problem.............
Regards, Ibj...

Understood, thanks.

I'm back in town and my new FPR and spark plug wires came in.  Haven't put them on yet, but will later today.  I've been inside the tank and found a fairly dirty fuel filter.  I wiped off the end and blew through it and it had quite a bit of resistance compared to the new one I just bought.

Fuel pump: I covered the end of the fuel pump hose with my thumb and turned the key.  The pressure was indeed enough to push my thumb away.  As far as flow goes, it looks okay to me, but I'm not totally sure what it should look like.  As such, here's a few seconds of video of the fuel flow.  https://youtu.be/aTzaQm2Vavc
Let me know how it looks.  Thanks.

Time to pop on some other parts.

That looks a bit down on flow to me. Would pay to remove the pump and check the pre pump filter as they can get full of crud and limit flow a bit.

New FPR, new fuel filter, new non-metal cap plug wires all installed. Issue is much much better, but still slightly remains. It's a little more apparent when the TPS is adjusted back to 0.375v, so it's staying at the smooth 0.5v again for now.

I'll look in to taking the pump out and checking the pre-filter. When I was in the tank, I did notice a couple chunks / collections of dirt about the size of a U.S. dime. What a pain to work on things in a gas tank!

Getting there!

It's relatively pain free to work on the pump with the fuel cal completely removed. Once you disconnect the fuel hose clamps they should twist off with a bit of effort. with the filter & hoses out of the way, the snap ears on the pump bracket can be released and the whole pump, bracket &filter screen can be lifted out. the wires are long enough to unhook with the pump up at the opening. Likely you'll need a new screen basket and some substantial cleaning out in there.....probably need a new isolator rubber also.

Don't forget the cap assy should be installed to open toward the rear with the hinge closest to the rider.

Oh, so I don't need to remove those 6 something bolts holding the fuel pump down, theres just some ears / snaps? This makes life a lot easier. I'll take a look, I've got about 4 hours before I need to be anywhere.

Out of curiosity, why does the cap assy have to be installed that way?

Rabidchiwawa007 wrote:Oh, so I don't need to remove those 6 something bolts holding the fuel pump down, theres just some ears / snaps?  This makes life a lot easier.  I'll take a look, I've got about 4 hours before I need to be anywhere.

Out of curiosity, why does the cap assy have to be installed that way?

There is a drain on the left side and it needs to line up with a cutaway in the base of the cap.

Push the tabs inward to remove the whole assy. Note when re-installing, one tab is wider than the other corresponds with openings in the tank. Goes in one direction only.

Another update:

Took the fuel pump pre-filter out. Here is before:



And here is after I took it off the pump and cleaned it.  It cleaned up well:





So today, I accomplished the following:
-New FPR installed, fuel pressure seems to have stabilized a little, but I still get around 29 psi showing on the gauge upon deceleration.  If it's idling and I shut the bike off, the gauge stops at about 35 psi and doesn't lose pressure.
-Installed new fuel hoses from fuel pump to fuel filter and fuel filter to metal outlet tube inside the tank.  No kinks or leaks in any hoses in the fuel system.
-Sucked quite a bit of dirt out of the tank using a syringe with a tube.
-Installed new fuel filter.  
-New non-metal cap spark plug wires installed
-Re-applied heatsink paste to ignition amplifier

Yet, the issue remains.  It has become less apparent yet again, but it is still present.  The engine still doesn't sound so happy when I let the clutch in or at low rpm.  I'm beginning to suspect the hall sensor, or possibly the computer(s).  Any other ideas?

After reinstalling the fuel pump still no change?

That filter basket has a correct hose connected to that nipple when in the tank?

I've lost track of what's been covered so far on this one.

Where are we at with coil diagnostics on this one?

Understandably, I've almost done the whole bike.  I thought about doing coil testing, but someone posted back that it wouldn't be very informative.  I've heard things about the coils with the little orange towers (I have these coils) and I am suspicious of them too.  I looked for jumping sparks around the coils and saw none, then let the bike idle while I sprayed the coils with a garden hose.  Idle only slightly changed / remained uneffected.  What's bad about the coils with the orange towers?

Sorry if I have missed it but have you changed the plugs. I dont think the plugs in the photo are correct. According to my book they should be Bosch X5DC or possibly for later models XR5DC(4V)which are both single electrode plugs. NGK equivalent D7EA or DR7EA (4V) also single electrode. The dual electrode plugs are definitely used in later BMW bikes such as my R1200GSA but not the K100s. Just an observation.

Someone posted about this earlier but then corrected themselves saying they were the correct plugs. When I first bought the bike it had dual electrode plugs like that, and then the local BMW shop replaced them with these current ones. This misfiring issues wasn't occurring when I first had these plugs put in, but if it's a possible cause, I'll look in to a replacement.

The correct plug for a 16v engine either K100 or K1100 is an NGK DR7EA or the Bosch XR5DC.
I much prefer the NGK but I have had bikes that will get thousands of miles from a Champion and start missing and farting on an NGK at 500 kilometres.

Well then, plugs are next! I'll post here again when I've got replacement plugs in.

What do you guys think of these: http://www.amazon.com/Denso-IX22-Iridium-Spark-Plug/dp/B000M6SSDS

Any specific advantage to iridium plugs over regular plugs in our bikes?

Iridium plugs don't wear out as fast.
Iridium is not an earth element, it only comes from space via meteors etc.
Iridium plugs are not cheap.

your better off with platinum plugs ...last way longer not much difference in price but the electrode wont disappear as quick ....iridium are made for extreemly high temperatures which we don't have an issue with on the k bike
quite the reverse

New Bosch XR5DC plugs installed. Issue remains unchanged.

Cant be much more you can do. Im stumped. I suppose you are too. If I think of anything ill let you know. Good luck.

I could still replace the coils, hall sensor, and computers. All very pricey things. Is it possible for this to be caused by the computers?

I replaced the coils and for about 50miles thought it was fixed but that is the very first thing I thought of when it started to happen. As for it being the hall sensors I doubt very much as they tend to just stop working alltogether when they get hot. They are not as expensive as you think for aftermarket sensors alone and you put them on the backing plate yourself. Radio shack may have thembut here are two suppliers
http://www.jaycar.com.au/productView.asp?ID=ZD1900 this lot are australian but as I said Radio Shack may have the same thing
http://www.bbautomacao.com/home_hall_effect_sensor_cyhme56.html
It is very possible it is the ECU but as you say expensive just to suck it and see.
The coils are different to the orange tower 8v coils but are available from EME at a much better price than OEM
http://www.euromotoelectrics.com/SearchResults.asp?Search=coils
No need to buy two as they are identical and interchangable except for the sticker 1&4 or 2&3
I have all the parts you need to try but if the postage wasn't such a killer.

Hello again,

Good news!! The problem is definitely fuel delivery. I had about an hour to do some work today. I tested the resistance of the injectors and the voltage from the harness. Results for each injector were the same and are:

Resistance: 16.7 ohms
Harness voltage: 11.73v

I then took the fuel rail off with the injectors still attached. I pointed the injectors in to a bucket and pressed the starter button. There is almost NO fuel coming out of them. There is no semblance of any sort of injector spray pattern. I held the starter button a few different times for maybe 5-8 seconds each time. By the time I was done, there was maybe 1/4-1/2 of an ounce of fuel in the bucket. I'll post a video of it later tonight when I get home again.

So the problem is definitely fuel delivery. Fuel pressure tested fine, so it can ONLY be one of three things. The injectors, the wiring to the injectors, or the ECU. Each injector seemed to give an equally small amount of fuel with each burst. I'll take them each out and test them individually out of the bike later. Any input on these possibilities is greatly appreciated. Thanks!

The spray from the injectors, even when correct is remarkably small.
A better test is to hold a brown piece of cardboard about 6 inches below the nozzles when you spin the starter. You won't really see the fuel spraying out, but you should see a nice round damp area pattern on the cardboard.

Your test may not have been indicative of a failure.

Alright, I will do that tomorrow morning.  The spray should still be in a stereotypical injector cone shape though, correct?

Also, are the measurements I got from the injectors and the harness acceptable?

The 16.8 ohms is perfect and 11.9 volts is good and would rise slightly when running at road speed to about 12.2 v

You will probably see the cone shape of a good spray pattern but it is a very fine mist and takes 3-4 squirts to even wet your hand at 4-5 inches but dont hold your hand any closer as it can penetrate the skin when really close. The brown of a cardboard box is about the best color.

the cone shape again is very difficult to see......when you try the cardboard trick, the farther from the nozzle, the wider the spay circle will show. I hate to be a bearer of bad news, but I'm guessing you have not found the problem.

In order to see a significant amount of fuel accumulating....when testing on a flow bench they use four fairly small diameter graduated test tubes and then they still fire the injectors like thousands of times to get a measurable amount of fuel......then, they compare the amounts in each tube to the others for even ness

Well, poo. I'm learning a lot from this. When I finally figure this out, I'll have a great running bike after everything we've done to it.

The reason I question fuel delivery as well is because my mileage has gone up quite a bit in the past few days. It's gone from about 33 around town to about 38-40 around town.

Is it possible that a loose timing chain could cause this? Evey once in awhile I hear a questionable noise from the front of the engine that sounds like a chain rubbing on something.

After I take a look at the injectors and take measurements on the coils, it might just be time to take it to the BMW dealer :/

A loose timing chain wont cause that problem but the rattle could well mean it's due for at least an inspection.

My bike is now running better than any time since I bought it. I am begining to think that as similar as they were you don't have the same problem that I had.
I have spent most of today just doing measurements on various components, TPS CO2 etc and they all come up right and there is no problem. So I have concluded that the TPS was the problem because it comes up right when measured and the bike is great.

I am going to reread this whole thread and make lots of notes and see what happens (writers cramp most likely) and comes to mind.
92KKs problem dragged on like this and it finally got sorted so don't dispair.

The increase in the fuel mileage from poor to normal, could mean that it's running more correctly. 32-33 is not normal 40 is.

True, however I'm 6' 3" and 245 lbs and am not particularly nice on the throttle as much as I should be. I still thought the mileage was a little low though. Either way, it certainly isn't running correctly right now .

Off to work on the bike now, will post back later with results.

I've completed the cardboard test of the injectors.  I've taken two videos, one of it injecting with throttle closed, and one with the throttle quickly but gradually opened to full throttle.

Here it is with throttle CLOSED:
https://youtu.be/XUodqN-_Ixc

And here it is with throttle OPEN:
https://youtu.be/VWFrgi-PkpA


I also tested resistance on the coils:
Coil for 2/3 read as
----13.05k ohms at the orange towers
----0.9 ohms at the connector that the harness snaps on to (primary coil)

Coil for 1/4 read as
----12.92k ohms at the orange towers
----0.9 ohms at the connector that the harness snaps on to (primary coil)

Are these readings acceptable?  I remember seeing somewhere that the orange towers should read around 12.6k ohms and the primary coil should read around 2.6 ohms.

I think that all tests about correct. Yours is 16V correct? The 16V coils are a lower impedance across the primary circuit that are the 8V coils hence, you have a lower reading that the spec which may have been for the 8V set.

Maybe someone else has an opinion also. Mine could be wrong, but I think you have not found a failure.

Maybe it would be good for you to re-publish the exact symptoms again as the post has grown so long. I for one have lost track of what exactly we are trying to correct.

I think it is getting close to the time when you are going to have to find a "known good" Ignition ICU and Injection ECU to substitute in for a test.

One thing I might encourage you to do is retest all the values on all pins of the ECU harness with it disconnected from the ECU but connected to all of the other sensors.

Alright. I thought everything looked fairly normal as well. That is one thing I have not done is test any pins at the ignition computer. I've tested pins at the motronic connection, but don't know of tests I could do for the ignition computer connection. I also haven't been able to find a diagram for it anywhere.

perhaps Inge or Charlie99 will chime in on that question

Well, the original issue is still what was posted on the first page.  It involves awful misfiring in the 1k-4k rpm range, both in neutral and under load. It causes drops in power during acceleration as well.  It is the worst right at 2k RPM. I've taken a video of the bike where the misfiring is fairly audible.  I revved the bike slowly from 1k to ~4k a couple times.  The misfiring is quite audible in the video if you have any sort of bass in your speakers / headphones.  I figure with the length of this thread it's about time I actually showed everyone what's going on.

I warmed the bike up to operating temperature before I took this video because it does not miss like this when it's cold.  It starts misfiring after maybe about 5 minutes of around town riding.


https://youtu.be/BfVbiyWS5Tg

Thanks for the update.
Seems like that almost has to be hall sensors or ICU.....
Any chance you'll be able to find a forum member near you who has a like model & can swap out your ICU to his machine to see if it malfunctions????

I suppose you could check the pin connectors in the harness from the hall sensors to the ICU. On my bike it's a square 4 pin connector near the ICU with a single wire connector next to it.

As you describe the issue It sounds more like a component breakdown rather than a connector as it's very reliable time & time again depending on speed & temp. A connector problem might not be that consistent.

I think it was stated that you can disassemble the ICU and renew the heat-sink paste under the electronic module on some models....not sure if that's only the newer 1100's.
Again, logic though tells me that if the heat sink paste fails, the module burns out......so maybe not that.

I would have to guess you are suffering from an electronic component breakdown of some kind which could include the coils that you've pretty much already ruled out. But you are indeed getting closer to solving this.
Regards, Ibj...

Could anything at all have occurred that has for some reason changed the Ignition timing? It might be sounding too far advanced to me?   I'm just wondering & wandering here......maybe not for sure about anything.

Is it the exact same model engine that came out? Any chance we have an incorrect match between engine components and control components?

What is our recent history in the hall sensors & wiring to them......I know that there is one place on some bikes where the wiring between the ICU & the hall sensors makes a very tightly stretched turn around a somewhat sharp pinchy edge at the right rear lower corner of the radiator could chaffe through a wire.

I'm thinking about loud about some type of loose fastener/component on the hall sensor pate or the spinning parts that trigger it?

Maybe you could restate for me what's been done/checked in & around there?....And this was the machine that got marginally better my slightly over-adjusting the TPS?

Does this bike have an airflow meter?

Have a look at:

https://www.k100-forum.com/t4147-air-flow-sensor-readings?highlight=flow+meter


 You can put your ohm meter on both "sides" of that circuit while manually moving the air door slowly through it's arc of travel.......you should see a very smooth gradual increase in resistance on one leg while having a corresponding decrease on the opposite 2 pins........just like a TPS.........just the air dor moving the wiper instead of the throttle linkage moving the wiper.
I saw one post somewhere showing how the airflow meter wiper could be repositioned higher up on the wiper coil surface to get it off the worn area of the arc and on to a newer surface.

One could test the airflow meter resistance from the ecu plug end, while holding the meter itself in hand and slowly/steadily moving the air door through it's entire arc of movement. The mechanical door should move smoothly in the hand with no tight spots

So maybe you have a TPS only??????
If so here's an interesting video. I'm sure you do not have this tool, but a handheld meter like you've been using may show similar "Glitches" in the "sweep" of the TPS wiper.

https://www.youtube.com/watch?v=8KfKfImoIXg

Wow, tons of info. To answer a couple questions, this is a '92 bike with a '91 engine. I called the BMW motorrad service shop and asked if the motors were interchangeable and they said all the 16 valve motors for the k100s (years '90-'92) were interchangeable and the same. The bike ran great for quite awhile after I changed motors and got everything tuned at the dealer.

I don't think anything could have happened to change the timing. I do hear a little rattle sometimes from the front of the engine which is probably the timing chain, but I don't think that would actually change the timing if it were simply loose.

The misfiring was marginally better via slightly over adjusting the TPS from 0.375v to 0.5v This TPS was off of a k1100 and was tested good. It's the same part number and is compatible. It caused no problems when it was first put in.

The hall sensor that is in there was bought used off ebay. I bought it to put it in my '85 k100 (sold now), because that bike wouldn't start due to a bad hall sensor. The used ebay one fixed the '85 and it fired right up. Then, when I got the '92 about a week later I noticed the tachometer was a little jumpy (symptom of failing hall sensor). I took the used hall sensor out of the '85 and put it in the '92. The tachometer on the '92 stopped being jumpy. So the hall sensor that is in this current bike, 1992 16 valve k100, is used from ebay, but was known-good when I put it in about a year ago.

There is no air flow meter of any sort on the 16 valve k100s. There is an air temperature sensor, but that is it. Other than that, it's just a TPS.

I've gotta take off for a few hours, but I will make a complete list of what's been checked / replaced thus far and post it (going to be a long list). I will also try and answer any more questions I missed. I'll take a look at that TPS video when I get back as well.

Thank you for the update.

You can certainly check the air temp sensor with an ohm meter doing so at the CPU harness plug, comparing with the spec. I think it's unlikely but definitely possible since you can see a temp related change.

I'm somewhat suspect of the hall sensor set up,

you should be able to possibly create a failure of change with a hair dryer. Heat the hall sensors with the engine cold.......if it acts up immediately, you might have a clue.

might be worth it to obtain another set for tests or a spare for later.

I have to tell you......watching all these posts has got me paranoid! Wife & I are leaving for a 2500 mile HOT weather trip in 2 weeks on our 85rs. I've never done anything to hall sensors and I have no spares.

Hopefully, they don't pick any time in the next 3000 miles to quit! Also.....I have never gone through the starter

Well the good news is that I finally got my 1968 BSA running right.. Broken wire at the points ignition. So I have a bike to ride, woohoo!

I'll take a look at the hall sensor tomorrow. I've tried that hair dryer trick before and a household hair dryer just doesn't cut it. I'll see if there are any cheap heat guns around town.

In regards to your trip, I remember seeing a post a long time ago that the rider carried an extra hall sensor with him at all times, just incase. And in regards to the starter, if anything goes wrong it's not too hard to fix. And if it's real bad, hey, they're not all that hard to push start.

I just finished listining to the latest video and to me it sounds like mismatched throttle butterflys. Like one or maybe 2 are opening well ahead of the others. With a 16v this can cause bigger problems than with an 8v because the fuel metering is done by the opening of the throttle and air temp To me it sounds more like a stutter than a missfire. If it were one butterfly leading the other 3 then you will get a temporary lean or rich untill the butterflys are open enough to not really be affected but a lack of syncronism which would be around 15deg of opening.
You can get a good indication by feeling one and watching others to see if there is a lag. Make sure you operate the throttle by the handgrip and not the choke lever as the choke pulls at the back of the TBs and the handgrip at the center. (I cant remember who but someone on the K11OG forum did a writeup on how to sync the butterflys and used the choke to open and close because it was easier than the hand grip and it led to some monmentual stuff ups and a lot of laughs).
Also the big difference to my problem I now know is that yours is bad with or without load where mine had to be loaded
Have you had a set of vacuum gauges hooked up to see what the readings are at idle and just above where the stutter is?
Also looking at the video of the injector spray patterns and the injectors need cleaning as its more a dribble than a spray especially the #1 & #2.

@ibj The 16v engines the ignition and fuel management are all done in the one unit.

I'm on board with Rick.......a check of the throttle balance is indicated.

Before spending money on the injectors, try running a bottle of Chevron Techron through it. I have had VERY good success with this injector cleaner.

Alright, I've put a list together of everything I've done / checked / replaced.  I think I've got it all, I'm not sure.


Things that have been checked / replaced:

-New FPR installed
-New spark plug wires (NGK, non-metal caps)
-New / correct spark plugs installed (Bosch XR5DC)
-New fuel filter and in-tank hoses
-Cleaned fuel pump pre-filter
-TPS adjustment (currently at 0.5v to marginally help issue)
-CO potentiometer set correctly
-checked coils for jumping sparks (none, even when doused with water while idling)
-synced all TB air bypass screws via a vacuum gauge
-compression test (all cylinders normal)
-small bit of oil consumption not cause of misfire
-coolant temp sensor tests good
-measured resistances / voltage at Motronic connection (all normal)
-fuel pressure tests okay
-no noise associated with misfire aside from the stuttering sound from exhaust
-re-applied heatsink paste (liberally) to the ignition amplifier
-air filter is new (about 1k miles)
-no air leaks apparent
-sucked out some dirt from gas tank via a syringe
-tested resistance of all injectors (all passed)
-tested voltage to injectors from harness (all passed)
-tested fuel spray from injectors (passed, but could use some cleaning)
-tested coils resistances (both passed)


Things I have yet to do:

-test hall sensor via heating it before a cold start
-do some tests at the ICU connection (anyone know any?)
-balance the butterflies (I think.. I'll take a look around for a tutorial)



I have a feeling one of those last 3 things are going to be it.  I'll find a tutorial for the butterflies, and maybe find a heat gun for the hall sensor, and get that done tomorrow morning.

So I'm reading up on how to sync the butterflies, and I understand it now.
I've been reading this thread: https://www.k100-forum.com/t1526-how-to-balance-the-throttle-bodies-solved
In this thread, I came across this statement: "I would not adjust butterflies until I am sure that the compression is absolutely the same." I did a compression test and got 120 psi on two of the cylinders and 130 psi on the others. Is this something that would effect the process of balancing the butterflies?

I would expect a bit better but that difference should not make any difference. If one cylinder was down to 60psi then it would play havoc with getting it even.

Yeah, when I bought the engine from beemerboneyard, the site said this engine had 200 psi on all cylinders. My compression test is just a gauge with a metal tube and a rubber bit on the end of it that you press (as tightly as you can) in to the spark plug area and then turn the starter motor. They likely have a better way of testing compression. I also did the test while it was cold.

200psi from a K100 is not possible they run 10.3 - 1 compression so take 14psi of because the first atmosphere is zero as far as the gauge is concerned and then multiply 14 x 9.3 = 130.2psi so thats as far as it can go.
I have had readings of 135 but that is probably the gauge. If you take the MAF sensor out of the system and allow for carbon build up 140 might be a feasible number but 200psi they are talking crap.

You could not run a petrol engine at that compression, about 14 - 1 is as high as you can go without getting into big problems and that is 196psi.

Dont be concerned as 120-130 is quite acceptable and indicates a good engine.

I figured as much. I remember going through all that myself when I first bought it thinking 200 seemed a bit high. Alas, here it is: http://www.beemerboneyard.com/11001461839.html It's all lies... LIES!! I wonder if the mileage was faked too.