Hi everyone,

Here is a link to a vid of my idle issues.  

https://youtu.be/knJWesu7u6g

Here is what I have checked:

1985 BMW K100

 Valves are in spec
Compression is in spec 
All electrical connections have been cleaned and dielectric grass applied
New spark plugs
No air leaks, I had previous found a leak in the throttle shafts and replaced the throttle bodies (off a known running bike with steady idle and still has the blue marking on throttle linkage)  and also new rubber and also sealed the harder rubber plastic intake to the manifold with Honda bond semi drying.
The fuel injectors were professionally cleaned and new o rings installed.
Hall sensors working fine.  
Fuel pressure tested good on both lines as per spec
I have also replaced the coolant sensor with a used one.  This helped the situation out to where you now see where I am at in the. Idea.

Now when I was trying to see if the coolant temp sensor was an issue, I unplugged the plug to the sensor that's screwed into the engine.  I then plugged in the other coolant sensor while the engine was running.  I had it grounded with an alligator clip.  Now when it was running and I would ground the clip to the sensor the rpms would rise sometimes up to 3k.  If I grounded the sensor directly to the engine the same thing would occur. 

I have one question that I could simply answer but it's snowing today here in NY.  I swapped out the sensors and when taking the one sensor out not a lot of coolant came out.  So at the moment I'm thinking that the idle is hunting around as a result of not enough coolant in the engine so that when the coolant which is a different temp hits the sensor it's sending the correct info to the computer and then immediately sending the wrong info when the coolant isn't surrounding the sensor.  Either that or I have some sort of issue with the wiring going to the computer.  

I know the afm could be an issue along with the computer itself but I also have replaced the engine (compression was low on the original engine and was cheaper to buy one with good valve clearance and compression than having the heads done. Orig engine had 95k on it.  It didn't look like I could get a 3 angle grind on the valves themselves and they are quite expensive.)   When I originally started the new used engine for the first time it idled perfectly.  This was also done with no coolant in the system at all.  Just to see if it would fire up.   The old engine had the same problems with idling as the new one has now so it definitely seems to be electrical.  

Im wondering if fixing the vacuum leaks fixed all the original problems and it's just the coolant problem now.  I'm going to try and check a little later in the day as it's an easy thing to do but I just want to start my own thread and have everything clearly stated as this has been a problem the entire time of owning the bike since August. 

Thanks for your help,

James

Have you changed the fuel filter?  If not, I would.  Get a NAPA Gold 3032. 

Temperature sensor could be causing a problem, but I have a hard time believing it could respond fast enough to cause the cyclic surging you are experiencing.

How does the engine respond when you rev it up?  Is it starving for fuel?

I've replaced the fuel filter as well.  It's not starved for fuel when revved.

Have you adjusted the air bypass in the MAF sensor. If it is not flowing enough air at idle the barn door will flap a touch and really mess with the idle.

No I haven't.  But the spring should adjust theflapper and not the screw correct ?

I totally missed that you had put that video up and after listening to it I totally agree it is electrical.
Probably the insulation breaking down on the leads or coils. Check all the connectors at the coils, ICU and ECU plugs maybe even the ignition switch itself with an intermittent connection or the kill switch.

hmm  not much coolant coming out , could it be that the seal on the filler cap  is good and holding the water by vacuume ?

you can test the coolant sensor by measuring pins 10 and 13 at the under seat ecu
if you look through the efi trouble shooting guide you will find the correct locations and a pic that explains it

I also had an issue with the coolant sensor not grounding correctly.

after cleaning the threads of the mount on the water stub , and finding a particularly corroded aluminium seal washer on the sender
I replaced that with a copper belville washer the right size , and soldered the copper washer so that there was some tinning all over the washer , i figure copper and aluminium are not good together ,, but tinned all over has made a great seal and good conductivity to both pieces ..just in case I took a small  flexible conductor cable up to the main "electronic " earth point on the frame just above the icu . the tinned washer made a good point to solder to
I did this in case the bolts through the water stub to the engine block were also corroded ,and was causing some form of potential difference  as the case of the sensor is one end of the sensor and needs a  good ground for everything to work properly

a trip after wards proved that my fuel economy became very different and much more economical

good luck

yes I agree test all ignition components and clean everything (connector wise ) under the tank whilst you have it off

don't forget the ferrules on the plugs

cheers

Well it looks like I'm going to have to wait till this blizzard passes through tomorrow here in NY before I even attempt to get back in the garage

Quick question.  The wires to the coils from the harness...  should the blue wires be closest to the engine or the green wires?

The green wires should be nearest the engine, with black/blue on the outside of the left coil and black/red (I think?) on the outside of the right coil.

Pic of my bike for reference.

Ok that's what I thought.  

I just bought a used mass airflow unit used on eBay from maxbmw.  I'm hoping this does the trick.  I hate throwing parts at it but it certain would help at the very least just check it off the list.

The MAF sensor did not do the trick.  I also double checked the fuel pressure.

In December, I helped a member troubleshooting a FI problem.  For no reason, the bike was shutting down. LEDs where installed on the fuel injectors to monitor their power and a strobe light was installed on the ignition and a gauge on the fuel line. Just before the bike was ready to die, the LEDs on the fuel injectors started to fade. The first thought was a defective fuel injection computer, but after replacing it, the problem persisted.
 
The culprit was the ignition module (even if the ignition was always strong without a miss).
The electronic ignition module pin #8 is sending engine rpm info to pin #1 of FI computer and this was part of the circuit failing.
What if your ignition module is failing and the output signal at pin 8 is not consistent? It took us a long time to pinpoint this problem and it's worth investigating on your bike.

CF

Ok so I tried doing diagnostic work on the computer female pins going to the computer and the ignition.  

I was able to do a continuity test on the coolant temp sensor wire to pin 10 going to the efincomputer and it checked out good.  I didn't get continuity when testing the ground from the other pin on the sensor plug...  I'm not sure if I'm supposed to or not. 

Then I tried doing what worked originally and plugged another coolant temp sensor that I have in to the plug (without it being installed into the engine) and got this https://youtu.be/f-a2lmsYN_0 

Exactly what I'm looking for but when I turned the bike off and then back on again it would not start.

I am going to replace the sensor with a brand new one.  Is there an aftermarket coolant temp sensor that can be used off a BMW car?  Or do I need to pony up the bucks?

Hi,

I replaced the sensor and the bike was idling fine and rev beautifully.  

Once the coolant system started to take in more coolant from the reservoir the idle started changing but would stay constant.  Still would rev great.

The idle went to about 700 rpm and then finally stalled out which could be expected as when I put on the throttle bodies the idle adjust screw wasn't set.

The problem now is that it then stalled out and now it won't start.  I unplugged the plug from the sensor and plugged in the other sensor to see if that could be the issue but it still would not start.  Now I'm really starting to scratch my head.  Lol

It would be interesting to see if the bike restart when totally cooled down. For me, it seems to be an electronic component acting up when it heats.

If a K overheats it doesn't cut out but if I remember correctly if it is overheated and stopped it is not allowed to start again until it has cooled down? Temperature sensor on the radiator controls this? Does the correct signal arrive at the plug under the seat as opposed to leaving the sensor?

92KK 84WW Olaf wrote:If a K overheats it doesn't cut out but if I remember correctly if it is overheated and stopped it is not allowed to start again until it has cooled down? Temperature sensor on the radiator controls this? Does the correct signal arrive at the plug under the seat as opposed to leaving the sensor?

You are totally correct Olaf. I never thought of this. But if he replaced the sensor with a cold external one, it should temporarily fix the problem.

If its the sensor yes then it should but if its after the sensor connection, in the loom or faulty at the plug to the FICU/Motronic then that may not sort it?

Ok so I put more fuel in it (I have the fuel pressure gauge still attached and it had the proper pressure leading me to believe there is enough gas) and it started up and ran fine

I am going to check tomorrow and see if it happens again. I'm going to finally take it for a drive and see if running it helps.  

I bought the coolant sensor at autozone for $18.   I read somewhere that it's interchangeable with an e30 car as they shared the same basic Jetronic efi system.

When the bike shut down I plugged in the other sensor and there was no change... so for now it looks like that new sensor will work but if I have any changed in idle etc I will buy the OEM sensor.  I will report back.

I am going to end the thread as being solved.  The combination of everything replaced seems to have added up to what you now see in this video...  the new coolant sensor being the last peice to the puzzle.    



I now have a good running bike with good compression and valve clearance and a stable idle.  (In the video above the noise is the baffle inside the muffler)

Replacing the engine with one that has good compression and valves in spec saved a big hassle and extra money.

Replacing the throttle bodies helped as well after seeing that the shafts were leaking.  I was taking a chance on used throttle bodies but I did see a video of it running on the bike before disassembly. 

Replacing the rubber in the intake plenum to throttle bodies and re sealing the intake manifold to the engine head im sure helped as well.

Having the fuel injectors professionally serviced wasn't necessary but at least I know they are clean and flowing properly.  

The MAF sensor was replaced with a used one from maxbmw...  I'm not sure if that made a difference or not and I can fit the old unit but for $50 I'm ok with perhaps buying it unnecessarily.  I'm not happy buying the wrong throttle cable for $50 and all the valve shims I bought for use on the old engine then buying another engine but that's just part of fixing it.  

Im trying to sum up for posterity what to do for anyone else encountering similar problems...   I basically opened up a can of worms and can't precisely say which one thing cured it.  It's more of a culmination of it all.   I had a used temp  sensor from the previous owner and saw an improvement when fitted which led me to believe the used sensors was functioning properly before any of this replace this and that happened... what I would recommend now is just buying a temp sensor off an BMW e30 car ( any mid 80s 3 series as they used the same jetronic system which is very dependent on the temp sensor.  It's $18 here in NY.  Everything else is your basic check for leaks, check compression, check valve clearance, check fuel pressure and just be persistent.   I was having  little small backfiring through the throttle body which sounds like bad timing and was even more frustrating when it occurred again after replacing the engine...  this was cured by replacing the coolant sensor as I was still experiencing those symptoms after puting on the MAF sensor.  Also make sure your coolant system is "burped" and doesn't have air in the system.  I hope this helps someone  and I wish I had a more definitive answer.

Thank you for the sum up.
Last year, at the end of the season, I started to experience irregular idle, and twice the engine cut on me. As the bike is still under 5 tarps and 15cm of snow in the back yard, I haven't look at it yet. I was too busy repainting the fairing and sidecar. As soon as the snow melt (another 15 cms of snow is expected today ), I will look at my problem. This morning I ordered a temperature sensor (just in case). They are cheap and it's always nice to have a new reliable one to test.

Looks like I spoke too soon...

Although I did ride the bike around the block and it was super smooth...  I was experiencing the rpm being too low at stop signs and the bike would stall out but would start right back up...

Then I started experiencing the bike dying out when the rpms were low and then it would stall out and not start back up like I had experienced in a previous post and the bike would not start back up again...   if I would come back a time later and try to start it, it would start....  then it was wait for a while... then try and coax it to start by playing with the throttle position...

Then I was experiencing the bike having higher rpms after blipping the throttle when at idle which may have to do with the throttle cable and throttle gear at the handlebar...

Today I wanted to have another look as I'm as stubborn as this bike.  I removed the tps to double check it's not interfering with the closure of the  butterfly's...   I then started experiencing the not wanting to start up thing again...   I checked the spark of all the spark plugs (pulled the plug and put a spare plug in and ground it to the engine) and could verify good spark on every cylinder...  the reason for checking is when it doesn't start I smell fuel like it's flooding.  As much as I scratched my head before I'm really starting to scratch it now....  

I had good continuity when checking the wire from the temp sensor to the ecu computer female plug...  maybe there is a poor ground?   I'm not sure where the negative side of the temp sensor plug goes to...

I'm running out of ideas and any help is greatly appreciated.  This is where my head is at now...

Possible faulty wiring in the efi somewhere, possible faulty efi or bad connection that needs to be resoldered, possible intermittent bad ground from temp sensor.     That combined with a possible interference from the throttle cable not fully closing the throttle bodies.  

Could the temp sensor relay possibly be at fault?

Bekman wrote:......
I had good continuity when checking the wire from the temp sensor to the ecu computer female plug...  maybe there is a poor ground?   I'm not sure where the negative side of the temp sensor plug goes to...
There are 2 thermistors in the sensor. Both are earthed thru the threaded body. Corrosion can occur here. Of the plug connections; one goes to the FICU #10, the other goes to the temp controller.
.....

Could the temp sensor relay possibly be at fault? No

The temp sensor relay only controls the radiator fan switching. Nothing to do with the fueling or ignition.

Re checking the sensor at the FICU plug. The best way to check this is not by how you've described above but by measuring resistance between FICU plug #10 and #13 counting from the lead end of the plug. This not only tells you that you have continuity to the sensor, but that the sensor is operating within spec and that the sensor is earthing correctly. You should be getting 2.5kΩ @20ºc and <300Ω @80ºc and above.

Another possibility could be your exhaust valve clearances may be too tight. Worth checking if you haven't already

New sensor has been fitted and grounding has been checked of the sensor.  Exhaust valves are in spec.  

I believe I have to check the harness, the ignition control module and the efi box.   

I pulled the spark plugs on the bike after it died out again and the spark plugs are soaking wet.  I put on another set of spark plugs but the bike still would not start.  This is leading me to believe I either have a weak spark or a very rich running condition...  should I be checking the alternator output at idle as well?   

If I wait a while or come back the next day the bike will start.

If anyone could help me with tips on checking the harness, possible taking the cover off the efi box and ignition module to resolder connections to fix a possible hairline crack in the solder would be greatly appreciated.

I apologize for my writing as it's been a pain writing anything more than a few sentences on my iPhone and double check everything makes sense.

Which temp sensor did you end up with. The cheapie from Autozone (re your post #22) or did you buy an oem part?

Cheapie...  but if I plug in another sensor it still will not start.

I cant see that a wiring fault would cause that, I would change the ECU for a known good one and see what goes.
Usually they don't even start if that sensor is unplugged.

Have you read post #13 on this thread? Exactly the same problem. The bike stopped and the next hour or next day, it worked fine again for a short time.
The original 1985 ignition modules have been replaced under warranty because they were not waterproof. Is your an original?
You can find ignition modules on Ebay for as low a $39. ICU on Ebay

Read em all...  going to get both ecu and icu

Got the ICU and ECU in the mail.  Let's see what the culprit is!  Could still be something else

Apparently the icu and ecu are not the problem

I have a spare set of injectors still mounted on the fuel rail from the used engine I bought.  I fitted it to the fuel and electrical lines to see if it would squirt fuel (I tested for spark and confirmed I had spark) and it is not squirting any fuel at all.  It does have fuel pressure though as when I went to put my fuel pressure gauge on again the fuel line was pressurized when taking off the fuel line.  The ecu is off a known running bike

Holy wiring harness madness!  I believe I found the issue after diving into the wiring harness...  The green and red striped wire coming from the ecu was connected to all green and red stripe wires by what I can only guess is a little glue and tied together and that part of the wires were zip tied to the frame to what I can only guess as to act as a way to hold the connection together

Same thing goes for the yellow with purple stripe wires going to the fuel injection...  was this factory?  It doesn't seem like BMW to me but I could be wrong.

Just a quick update...  I'm waiting for coils in the mail and the used wiring harness has been fitted.

Did you disassemble your ignition switch and clean its contacts? RicK G mentioned that in post #6. They can look innocent enough but defects in them can be insidious, often having sporadic affects just like these.

Ok I can now say for real that the problem has been fixed!  I'm idling at 1k rpm and was able to use my vacuum gauges to sync the idle air screws.

I had fitted the new coils but had experienced very rich running.  I ordered and fitted a new coolant temp sensor instead of the BMW car one, after it was fitted I then ran into idle issues.  

I could not sync the idle air screws...  cylinders 3 & 4 had lower vacuum than 1 & 2.  I wasnt sure if the fuel injectors were seated fully so I began to check for leaks with starting fluid.  I confirmed they were not fully seated and I also found the intakes had a leak where the hard plastic meets the cylinder head.

Now after I found this out, I realized again that I did not have this problem when I originally fitted the new used engine.  What had happened is i ended up switching those hard plastic intake manifolds with the ones from the old engine and I had used Honda bond semi drying.  When I pulled the intakes you could see the seal from the Honda bond was compromised on cylinders 3 & 4 and 1 & 2 were close to having an issue too.  I removed all the Honda bond and sanded the mating surface on the head and switched out the plastic intakes, more importantly I used the drei bond I bought to do the water/oil pump in the past.

Success

It's a long drawn out story of making sure there are no intake leaks, making sure the valves are within spec (ended up with new engine) and using a oem temp sensor.  

I haven't tested the old coils, MAF, ICU, ECU and after digging into the fuel injection harness, I'm not sure if that was necessary as well...  but necessary after I was through with it.  Either way the bike is fixed and running great and I was able to obtain those parts relatively cheap and plan on selling them cheap too.  Replacing the throttle bodies helped as well as I found a small leak in the throttle shafts.

great to hear ....woohoo !

Good feedback on the problems found. It will no doubt help others with similar issues.

Mike