MOTOBRICK.COM
TECHNICAL MOTOBRICK WRENCHING In Remembrance of Inge K. => The Motobrick Workshop => Topic started by: moto949 on February 06, 2016, 09:36:30 PM
-
The otherwise pretty nice 1993 K75S ABS (25k miles) I bought recently has its fan running constantly, from start up to shut down. I rode it 700 miles home after a long distance craigslist purchase, and since then about 300 miles of short rides, and the fan always runs. I worry that the common problem of a burnt out fan will be my next problem. Now, I did purchase a coolant temp sensor and I understood this to be the cure, but before I dig into this hassle of a job (removing fairing, etc) I recently heard the cause could be a Temp sensing switch unit (in the fuse box area), and since this is easy to swap out, I wonder if anyone has experience with fixing this fan-never-off issue. I'll mention that the bike runs fine hot or cold otherwise, no stumbling at all. So is the fix the one minute $60 plug in switch or the use-up-my-Saturday-afternoon temp sensor? Thanks folks!
-
it is actually easier to replace the temp sensor than the switch...
taking the lower fairing off is a 30 second evolution. two bolts on bottom and slide off.
getting at the switch involves removing the airbox.
-
The otherwise pretty nice 1993 K75S ABS (25k miles) I bought recently has its fan running constantly, from start up to shut down. I rode it 700 miles home after a long distance craigslist purchase, and since then about 300 miles of short rides, and the fan always runs.
Did you check your coolant level?
-
it is actually easier to replace the temp sensor than the switch...
taking the lower fairing off is a 30 second evolution. two bolts on bottom and slide off.
It's not as easy as that, the sensor is located in the standpipe behind the radiator.
getting at the switch involves removing the airbox.
The fan relay is inside the relay box, you need to lift the fuel tank in the rear end and remove the cover/lid.
-
A popular mod is to add a switch to override the relay for the fan......maybe the PO forgot to tell you.
A phone call could save you some ............
-
A popular mod is to add a switch to override the relay for the fan......maybe the PO forgot to tell you.
A phone call could save you some ............
+1, that was my 1st thought, manual fan switch.
-
Thanks for the replies.
Re: Coolant -- thanks, but the level looks okay. BTW, would low-level trigger the fan instantly, even on cold start-ups? Not sure if BMW built this in.
Re: extra fan toggle switch. I'm hoping a new sensor or factory fuse box switch should solve this issue. Plus I don't have a temp gauge to help remind me to throw the switch. But I'll save your suggestion as a last resort, thanks. The PO (who I never met since his friend handled the sale) was not even a shade tree mechanic and had it serviced only by dealers (which won't touch any BMW from the 20th century any more). I figure this fan problem cropped up as the bike sat unridden much these last few years.
BTW I gotta agree with Inge K about the involved sensor replacement - it looks all tucked up in there, facing inward and requiring draining the system (I know, it probably needs it anyway).
What I believe I'll try is borrowing a friend's temp sensing switch (when he gets back next week). If that doesn't fix it, I'll just have to change the sensor. I think I read that replacing the sensor can also fix a rich running condition -- not sure if I have that problem (if I do it's minor), but I intend to tune the injection soon anyway so if I have to replace the sensor, at least I'll know I'll be tuning from a clean baseline.
I'll check back in a week or so, but in the mean time, if anyone else had this same problem and knows a fix, let me know. Thanks!
-
To test: remove the relay and measure the resistance between terminal 31 and E in the socket.
At ~20oC you should read ~2500 ohms, if you read a lot lower (fan starts at ~150 ohm)
you have a short in the sensor or the related wiring.
If the resistance is OK it's the relay that is the problem.......you could try to give it a few gentle knocks,
could be welded contacts.
-
We are getting the terms mixed up.
There is a sensor (sometimes called a switch) behind the radiator on top front of engine. Acessed by removing the air box. No fairing bits need to be removed on an S model.
There is a relay (sometimes called a switch) in the fuse box. Accessed by lifting tank.
The talk of a manual switch to override the computer is to tell you to look for one the PO may have already installed and not told you about.
-
The temperature sensor in the stand pipe on the left front of the block is not a switch with ON/OFF characteristics. It is a pair of negative temperature coefficient (NTC) thermistors joined together, with a non-linear analog response to temperature change. The unit consists of a threaded metal housing, with two terminals. Each terminal is connected to a thermistor with the housing itself being the common connection point for the pair. As water temperature changes, the resistance of the sensor changes -- goes higher with lower temperatures, lower with higher temperatures.
The temperature relay in the relay box is not a simple mechanical relay, but rather a complex electronic circuit with active integrated circuits like you'd find in your computer. There is a small relay on the circuit card that handles the (relatively) large fan current. Over time and with vibration, the solder joints on the circuit card can fracture, causing the relay to fail. It is possible to open the relay's enclosure and remelt fractured joints using a soldering iron, restoring the relay's function.
Inge's advice is aimed at isolating the failing component. We know the fan and wiring are intact and operational. We also know there are no aftermarket modifications to the electrics. The relay and temperature sensor are still suspect.
-
Thanks for the details re testing, Inge. I'll do so when I get the bike down to my buddy's shop next week (I just now tried the tapping bit, but no go. Oh well, worth a try).
Re 'terms', yes, I believe on other posts, members sometimes may say 'switch' for both the Coolant Temperature Sensor as well as the Temperature Sensing Switch Unit. Anyway, I just got my hands on a Haynes which says the radiator must be removed and drained to swap out the threaded Coolant Temp Sensor but I have a feeling you may be correct for a K75S if I wriggle the wrenches down there just right- I suppose I'll find out next week if I need to change it.
As for a hidden switch, as mentioned, this bike was only tended to by dealerships, so I doubt they rigged a switch... then again, this was the same dealership that way-back-when put the fork oil amount for an early K75 (I measured it as I drained it) into this same '93 bike's Showa forks causing crazy dive under braking.
Oh, just rec'd rbm's comments. Thanks for the details! So, being as they're both suspect, I will try to swap (or open/inspect/solder) the temp sensing switch unit first since that's easier. Then if needed, replace the threaded temp sensor. I guess you're saying the temp sensor has one terminal feeding the idiot light (when it reaches whatever predetermined temp) and the other feeding its signal to power the fan (at, hopefully, a slightly lower temp). Just to add confusion, I just checked my dash lights with the key turned to 'on' and my temp warning light, um, doesn't. The battery and oil etc do light up. Normal? Or, more confusion, bulb failure?
-
I guess you're saying the temp sensor has one terminal feeding the idiot light (when it reaches whatever predetermined temp) and the other feeding its signal to power the fan (at, hopefully, a slightly lower temp). Just to add confusion, I just checked my dash lights with the key turned to 'on' and my temp warning light, um, doesn't. The battery and oil etc do light up. Normal? Or, more confusion, bulb failure?
One of the elements in the coolant temp sensor is connected to the injection ECU for calculating the fuel mix.
The other element is connected to the relay which take care of both the fan and the warning light.
It's normal that the overtemp warning lamp don't show when you switch on the ignition.
-
Re: Coolant -- thanks, but the level looks okay.
Did you actually remove the fuel tank and check the coolant level by removing the coolant fill cap?
The coolant level in the coolant overflow tank is not an accurate measure of how much coolant is in the cooling system.
... fan running constantly, from start up to shut down
I guess if the fan is running as soon as you start the bike, it's probably not related to coolant level.
-
here is your sensor location with airbox removed. no need for fairing and radiator removal....
(http://i397.photobucket.com/albums/pp59/mblytle/K75S%20black/20160109_144618.jpg) (http://s397.photobucket.com/user/mblytle/media/K75S%20black/20160109_144618.jpg.html)
-
+ 1 Marshall It's the easiest way.
-
Great photo, mlytle! Thanks one and all. I'll be at it after this busy work week and I'll update results.
-
Update: I just swapped out the temperature sensing switch unit (the white, multi pronged, 2 inch box under the rear edge of the gas tank), and upon first start, the fan did not kick on and the bike idled fine - so far so good! Then something came up and I had to leave for awhile.
Anyway, the unit I swapped it out for is thought to be in working condition, though it is a used old part so I'll only know it's a 100% fix when I go on a ride and hear the fan come on in traffic as needed, fingers crossed (again, I wish BMW added an actual temp gauge along with the warning light so there's no guessing/worries). If the fan doesn't come on at all, that would mean this TSSU has a different issue, I suppose. Hopefully that long ride will be later today, time allowing, then I'll know for sure.
BTW, I did pry open my "suspect" temp sensing switch unit and got in real close with a jeweler's eye piece thingie but found no breaks in any soldering connections (it all looked pristine in there), so the failure had to be somewhere else, perhaps inside the tiny internal, black relay itself (?).
Thanks again, folks, and I'll update again when I can.
-
The only time my fan kicks in is when I'm doing the CO adjustment in my garage, so I fitted a override switch. It is comforting to be able to test the fan, and with the OEM fan they need to be spun or they seize up.
Regards Martin.
-
It works! Finally got back to the bike to give it a proper test. It started right up and idled smoothly as usual, even on a February morning (... okay, it's Florida and 60 degrees) Approx four and a half minutes later the fan kicked on. Good feeling. So thankfully, I didn't need to change the sensor in the coolant junction after all, but I do have a new back-up one which I may swap out next coolant flush as preventative maint' (and it's good to know it won't be as involved as I thought). Thanks to all those who chimed in!!
-
At 60º, four minutes seems early for the bike's fan to kick in at idle. Riding it will tell the story. Good luck!
-
Know what, you're probably right. What I think is affecting it is the idle speed; it's far too high at the moment, and that is probably building the heat too fast. It's up to 1500 rpm, even after settling down on the lowest idle switch level.
What happened was, when I recently got the bike it was running very rough during low speed acceleration. I changed the fuel filter (zero difference) and new plugs (noticeable difference since the PO - dealer installed - put in resistor plugs, which I heard is a no-no with resistor wires - I put in the correct ones (the ones I took out looked just right in color, etc) and it certainly got smoother, but far from perfect. I then had a local shop tune the injection and it only helped a little. Faced with either cleaning the injectors or swapping them, I went for the four-hole aftermarket green injectors. Soooo much better, like a new quick bike, but the idle crept up to the high level it's at now.
I'll be tuning the injection myself this time (which I think needs a re-do what with the new injectors and a valve shim change I just did) at my friend's shop with carb stix, and I'll be doing the ignition advance too. At that point I'll adjust the idle down. Hopefully I've covered all the bases.
Thanks!
-
I'll be tuning the injection myself this time (which I think needs a re-do what with the new injectors and a valve shim change I just did) at my friend's shop with carb stix, and I'll be doing the ignition advance too.
If it were my bike I'd balance the throttle bodies, like you intend to do, adjust the idle and stay away from the ignition advance if it hadn't been touched by the local shop—and I doubt that it was. They probably just went over the bike with a whisk broom and cleaned off the spilled coffee. :giggles
Take it for a ride to somewhere like Shark Valley Loop or out along the Myakka or to Loxahatchee then decide the next move.
-
... I went for the four-hole aftermarket green injectors. Soooo much better, like a new quick bike, but the idle crept up to the high level it's at now.
When I recently upgraded to the 4-hole injectors, it whacked my air-fuel mixture.
Once your bike is idling warm on the center stand (and the throttle bodies are balanced at about 1100 rpm) depress the green start button on the right hand grip for a few seconds. The engine RPM should increase a tiny bit. If it does anything else you need to adjust your air-fuel mixture. Easy to do using this method (http://skylands.ibmwr.org/tom/tech/co-setting.html). Basically, pop off the rubber plug in the top-front-right of the airbox, stick a short 5mm hex wrench into the hole and note the position. Turn clockwise to reduce the amount of air allowed into the idle mixture, turn counter-clockwise to add more air to the mixture. Adjust until depressing the green button acts as suggested above. (FYI - Turning the screw simply allows/restricts air to leak past the airflow meter.)
With my freshly installed 4-hole injectors, pressing the green start button slowed my idle considerably and almost stalled the bike. Too much fuel. I added air -- about 5 full turns of the screw.
AFAIK this adjustment only affects the idle / low-RPM mixture.
Next, adjust your throttle bodies again if altering the mixture changed the resting idle RPM.
I would not mess with the ignition advance now. Maybe in three months when you're sure everything else is perfect.
-
Tim it would be interesting to see what doing a plug chop at highway speed say 60-70 MPH looks like it. Are you doing any fuel comparisons. And I know you probably don't have a Dyno, but on gut feeling do you think the 4 holes are an improvement. I have been mulling changing over to them, but the lack of concrete data has me wary.
A curious Martin.
-
Like most I noticed improved smoothness and responsiveness with the new injectors. The upgrade happened about 300 miles ago.
I haven't put any real miles on the bike since figuring out the air-fuel mix discrepancy a week or so ago. The sun's coming out next week though so check with me later :)
-
:2thumbup: Keep on forgetting about the weather.
Thanks Martin.
-
I finally got around to tuning the injection (as I drift further off the original post topic of the ever-running fan, sorry) and the bike is running real well. Some notes though; The 5mm hex airbox plug adjustment made virtually no difference to idle speed either before or after the injection sync (Motion Pro non-mercury Sync Pro). I turned the adjuster from full-in to 6 turns out, and the idle remained static, then it dropped a tiny bit, but ran a little rough, at about 7 turns. Anyway, when pressing the starter button now after a full warm up, the idle rises by about 50 (fifty) RPM, so I guess it's right on - I left the hex at 3 1/2 turns out.
The current idle is at 1200 RPM (which I'm not unhappy with) but if I reach down at a stop light and press on the adjustment screw (between the 1st and 2nd brass screws) I can drop the RPM to 1000, since the screw rests above the stop with a gap of a couple mil (that screw is fully backed out now). BTW the TPS clicks just off the stop as it's supposed to, and there is a tad of play at the throttle grip, like the manual says.
Overall the bike is certainly quicker than it was. Okay, it wasn't running great then but I never thought it was too bad for a non sport bike. I'd go so far to say that it's faster now, even with the taller 2.81 gearing I swapped in when I had the splines lubed just days before this injection retuning / injector upgrade. The motor has good response at all RPMs and no hiccups anywhere, and no backfire popping. I didn't yet advance the ignition, as suggested, (thanks guys) but I may try that some day.
The bike is also smoother than before; I know this because right after I got the bike, the first mod I made was replacing the Mickey Mouse ear BMW mirrors with a set of Ducati Monster mirrors that actually showed more than my shoulders. Problem was, the new rear image was fuzzy, probably because the Ducati mirrors were plastic and lighter. I didn't really want to give them up because I liked the way they looked, and believe it or not, they cut turbulence that I could really detect (I rode around with one new one and one OEM for a couple days and noticed the difference side to side). But now, after tuning, there's no buzziness in the images at all.
And... the fan still kicks on and off perfectly. Thanks again everyone!
-
Sounds like it might be dialed-in nicely.
Curious exactly which mirrors you're using.
-
The current idle is at 1200 RPM (which I'm not unhappy with) but if I reach down at a stop light and press on the adjustment screw (between the 1st and 2nd brass screws) I can drop the RPM to 1000, since the screw rests above the stop with a gap of a couple mil (that screw is fully backed out now). BTW the TPS clicks just off the stop as it's supposed to, and there is a tad of play at the throttle grip, like the manual says.
1200 is a bit high. Sounds like the TPS is interfering with the TB given this description above. I'd experiment by completely slackening off the screws holding the TPS to see if the TB returns to its proper resting position and the idle drops to 1000. If successful, then try setting the TPS up but making very sure that you are not applying any side pressure that could cause binding. If unsuccessful in dropping the idle, then that's OK also if you're happy.
-
Thanks RBM, I'll try it soon as I get a chance.
Re mirrors: They were off a Ducati Monster 796, but that type is probably mounted on other Ducs, too. Wish I could attach a photo but my phone won't sync my latest shots and Verizon is working on a fix. I'll keep trying.
Please note that the mounting of OEM BMW mirrors is simple because of the threaded nut/bolt, but for the Ducati items I had to not only search through the many nut/bolt drawers of NAPA to find what fit as snugly as possible, I had to save some extra length for a few washers/spacers or my inner knuckles would brush up against the lower mirror stalk when shifting/braking. *I can't remember, but I may have also squirted a dab of silicone around the threads within the hollow mirror bolt housing to help quell any vibration I figured might occur. As I said, with the bike tuned and smooth, it's crystal clear, and I didn't have to tap any threads.
Beware that these mirrors are low slung. If I compare them to the bar ends on my old Guzzi Mille, I'd say the glass area is roughly in the same location, that is, the center point of the glass is approx 5 inches up at a low, 40 degree angle from either pinkie finger. The glass is also about 2 inches forward of said fingers, more or less in the same plane as the brake/clutch levers. So, like on my Guzzi, I have to dip my eyes down just a bit, but then again I do have an almost 100% clear view behind me now. At first I thought I made a mistake because of having to be looking further away and down, but I got used to it quickly and it's natural now -- it's just that the totally different location of the mirrors takes time to get used to. The OEM mirrors were 90% blocked when I wore my jacket (always), so cars sometimes seemed to come out of nowhere, and unless I consciously shifted my body over I couldn't see what was going on back there - (I'm 6'2" 225+) Anyway for me, the new mirrors are safer because my eyes always catch movement back there.
Not that I'm unhappy with them at all, but... I recall seeing a Triumph Street Triple (not sure what year) with cool looking mirrors with maybe longer stalks that rose a little higher. No clue if they'd fit our bikes. Remember, it's important to look closely study the handlebar angle of the donor bike vs your model K75.
-
Thanks. I see how those stick out further than my "short" OEM Mickey Mouse mirrors.
-
"When I recently upgraded to the 4-hole injectors, it whacked my air-fuel mixture"
I'm really glad I just read this. I just changed my 86 K75S injectors to the Bosch 4 hole green type. The bike runs better, much smoother and better pick-up but the idle is way too high at 1500-1600 rpm when properly warmed up. I tried to balance the throttle bodies using a Carbtune. I couldnt get #3 balanced close to #1 and #2 and I noticed that the FPR vacuum take-off hose on #3 was old and knackered so replaced it. That did it and I balanced them OK and adjusted the idle back to 1100 rpm. I followed the forum advice to set cylinder #1 to 1.5 turns out then balance #2 and #3 to that. I guess I need to start with #1 set to 0.5 turns out and balance the others to that?
That was about 300 miles ago and yesterday when I was out for a long ride I noticed the idle was back high again when properly warm. I just had a good look at the crankcase breather hose for cracks but that looks pretty decent, I tried the propane checks on the injectors - nothing, I can see that the idle adjustment screw is not resting on the base plate but can be pushed down a bit then comes back again, I checked the TPS wasnt holding it off - nothing there, and now I was just starting to wonder if the new 4 hole injectors may have affected the fuel-air mixture and thought I'd consult the Motobrick oracle for some guidance. Time to get back out in the garage.....CO setting and then rebalance throttle bodies to #1 set 0.5 turns out. Anyone got any other ideas?
-
I can see that the idle adjustment screw is not resting on the base plate but can be pushed down a bit then comes back again, I checked the TPS wasnt holding it off - nothing there
Do you mean the throttle stop screw?
Something is holding the throttle open. Check the action at the grip, the cable and the operation of the "choke" controls. Once you get the throttle closing properly, you will need to go back and rebalance the throttle bodies. But this time leave #1 open 1.5 turns. .5 turns is too rich, especially if those 4 hole injectors do indeed put more fuel into the mixture.
-
The bike runs better, much smoother and better pick-up but the idle is way too high at 1500-1600 rpm when properly warmed up.
Would a good strategy then be to do the tuning procedure when the bike is "properly warmed up" to the extent that it was when the idle was too high?
-
Once it's warmed up and idling, press the green start button for a couple of seconds and see if the RPM's increase a little.
-
Do you mean the throttle stop screw?
Something is holding the throttle open. Check the action at the grip, the cable and the operation of the "choke" controls. Once you get the throttle closing properly, you will need to go back and rebalance the throttle bodies. But this time leave #1 open 1.5 turns. .5 turns is too rich, especially if those 4 hole injectors do indeed put more fuel into the mixture.
Yes the throttle stop screw.
I checked all the usual things that may be preventing it returning to seat on the plate: throttle cable free play, fast idle (choke) free play, throttle position sensor. None of those are preventing it seating. Not much finger pressure is required to seat it so it feels more like the return springs are not strong enough and the butterflies are not seating fully.
I followed Tim T's advice. When I pressed the green starter button when engine is well warmed up (fan coming on and off) and the revs fell almost to the point where the engine died (as Tim described above) so I tried to adjust the mixture using the air flow meter mixture screw. It was originally set at 2 & 2/3rds turns out from fully seated. It was not possible to use the CO lean drop method because the idle still remains too fast (1600 rpm),even when resetting it by rebalancing the throttle bodies, however despite this and suspecting that the fast idle may relate to the idle mixture I set the mixture screw somewhere about 5.5 - 6 turns fully out and now it increases revs slightly when the green start button is pressed. The fast idle problem remains and I am not particularly happy to run the bike without knowing the mixture is not well sorted (having spent a lifetime running two strokes).
The fast idle is stopping me going further for now. But from what Moto 949 said he seems to have partially sorted this, and his problem sounds pretty similar to mine just not so bad. The strange thing is that the fast idle problem only really appeared after the upgraded Bosch green 4 hole injectors were fitted so I may pull these and check the o-rings and seating, although I expected that the propane check would have shown up any leaks there. If I can sort the fast idle I can then try to set the mixture with more confidence. Any thoughts would be welcome.
-
Now that the AFM adjustment is better, what happens to engine RPM's if you screw the three brass idle adjustment screws in all the way?
Have you ever cleaned those brass idle adjustment screws? Take them out, clean with carb cleaner, check the tiny O-rings, put a drop of clean motor oil on the O-rings, clean out the three holes in the throttle bodies, and then put the idle adjustment screws back in gently.
-
When I screw in the brass throttle body air bypass balance screws right in the engine speed drops but does not quite stall. Whilst that helps with the idle speed the trouble is the throttle body vacuum balance goes out of whack especially on #3. I suspected at the start that these screws might be the cause of air leakage, so I did remove them clean them and used a cotton bud to clean the holes and the screw seats. I did this before I tried anything else on the air leaks check. I am now re-checking all the stuff I initially did. Today I replaced the air screws with some newer second hand ones from a set of K1100 throttle bodies. The O-rings were much springier on these ones and offered some better resistance when screwed in so I felt they were sealing better. However the end result was the same as before. So I can have lower sensible idle speed, brass air bypass screws screwed right in but with throttle body balance way off OR I can have excellent throttle body balance but very high idle speed (>1600 RPM).
I was planning on changing the original throttle bodies for the K1100 ones (using 3 of the 4) but I dont really want to go any further until I sort out the idle speed and the air-fuel mixture. It was cold out in the garage today so after about 6 hours of this I was falling into Zen's 'gumption' trap. I've walked away for now for some beer and thinking time.
-
I know you said the TPS or throttle cable aren't keeping the throttle open but if I were you I'd remove the TPS from the throttle bodies. Two screws and it slides right off. I'd also disconnect the throttle cable from throttle bodies if removing the TPS doesn't lower the idle.
With a Carbtune attached I'd screw in the brass screws 98% of the way (not tight) and see where that puts your RPM's. That should get you under 1k RPM. If it does, look at the Carbtune and unscrew the brass idle screws to raise the RPM's to 1100 RPM. Don't worry about favoring any particular cylinder. Only unscrew the screws necessary to balance at 1100 RPM.
Those screws only affect the idle mixture. Once you open the throttle the air those screws leak becomes irrelevant.
Lift and hold the throttle lever on the throttle bodies and increase the RPM's to 3k. Do the cylinders looks close on the Carbtune?
-
Thanks for the second pair of eyes. I agree it would be sensible to eliminate the TPS and throttle cable by disconnection, that way it is 100% certain there is no interference with the idle speed.
I'll follow your advice when I next get out in the garage and let you know the results.
-
OK here is the final installment of this saga..
I decided that I needed to re-check again for air leaks. I had previously used propane around the obvious leak points but I was sceptical about how effective it was. I managed to borrow an exhaust gas analyser and decided to repeat the leak checks with propane and looking at the effect on the exhaust gas CO measurement instead of any speed increase. It clearly showed up a couple of leaks around the new injectors. I removed the injectors, cleaned the ports where they seated and used O-ring grease to reinstall the injectors. I restarted the engine, no excessively fast idle and no problem getting balance on all the throttle bodies with the Carbtune. I then set up the idle mixture at 2% CO. I have done a couple of hundred miles on it since and it runs like magic, certainly seems to have a much smoother throttle response than before from 60mph in 5th gear. :riding:
Thanks for the help and encouragement.