Had something interesting happen today.

Quick backstory: I've been having issues with my '92 k100 running hot. I have been running cooling system flush for a couple weeks now (replacing it after every 3-6 hours of cumulative riding per the instructions). The bike has definitely started to run in the more normal heat range consistently, and cools down quickly when the fan comes on.

So today, I'm riding around town and the heat gauge is in the normal range, just below the half way mark. Then the idiot light for heat comes on. What? My first thought was "liar!!", but I don't know what to think of this. Any input is appreciated.

Thanks!

Either the temp gauge is cactus or the light circuit is. The light should come on at 111deg c.
They run on totoally different electrical circuits.

Is it possible it could be the sensor(s) as well? Do they each run off of a different sensor, or is it just the one behind the radiator on that metal boot that the water goes through? (can't think of part name )

the way I see it the temp sensor on the oil/water pump goes to the gauge...the one hidden behind the radiator does all the interesting stuff like info for fan on off and overheat light and to the computer for fuel /air mixture etc...

yep your right klompy

if the temp sensor ...behind the radiator was going faulty ....fan and warning lamp comes on far too early

this has been noted a few times in here ....but if the light came on ....why didnt the fan ....?

does the fan actually come on after a while idling ?

you may have found a couple of issues ...( the fan electrical connectors often become corroded ) and there has been noted problems with the temp relay (under the tank electrics box )

worthwile chasing that one i recon

there ahve been quite a few post in here about this ......try using the "search" feature to find out more ?

Okay, so after reading that the temp sensor behind the radiator also controls mixture and such, I think that might be the problem.

Awhile back, the fan stopped coming on (possibly due to the temp sensor now?), so I wired it directly to a switch on the dash. I'm good about making sure to turn it on and such the only thing with this is that it didn't come on too early, but just didn't at all, or, way too late, but I didn't let it get that far. Also, when I was diagnosing all this, the temp relay checked out fine and with the help of a member here, we figured it was the lack of a ground signal from the motronic unit to the fan, which seems as if that could be caused by a faulty temp sensor.

Recently, I noticed that on start up and while the engine is still cold, there is a blue-ish black / dark colored smoke coming from the exhaust. I wrote it off as regular ol' K-bike smokey starting because it went away after the bike warmed up. But now I'm thinking that if the temp sensor is going, it could be causing a mixture problem and that is, indeed, gasoline smoke I'm seeing.

So with that fan issue, the temp light coming on when te gauge says its okay, and the seemingly rich mixture on start up, what do you guys think?



I usually do try the search box first, sorry. Totally skipped it on this one, but I'll check around still.

I know it might sound simple, but I had some electrical problems which I eventually eliminated when I repalced the ECU under the headstock.

I did this because it was €75 plus VAT to simply test it as against €50 for a secondhand one.

I cleaned off all the contacts on the replacement unit until they looked nice and shiny and put it in. Problem solved. I still dont know if it was the old unit or a faulty connection but I suspect it was a faulty connection. I just wasnt spending the testing fee to be told I needed a new one anyway.

I guess check all the connections and clean them up as it seems they are as likely to be a problem as the sensors or components themselves. I used small pieces of 1200 wet and dry paper, used dry and then cleaned off all the bits. As far as I know a bad sensor connection at the unit can make it look like a faulty sensor.

While you are at it replace your fuel hoses with longer ones and you can get at things under the petrol tank without depressurising the system and undoing connections....

Fix this first and then see what happens with the smoke. A very rich mixture will smoke and also wash away/burn any oil.

Sounds good. I'm going to drop by a RadioShack today and pick up some deoxit for the computer connections and give it another shot. I'm also planning on flushing the system once more, so maybe I'll pop the radiator off and take a look at that sensor. I might have another one lying around somewhere as well, so maybe I'll just swap it.

Also, I just let the bike idle and heat up to just before the red on the heat gauge. The idiot light never came on... It could be possible that since I'm flushing the cooling system, there are chunks of stuff going around in there and that could have possibly shorted the sensor somehow making the light come on like it did? Just thinking out loud here.

Is there a way to test the temp sensor behind the radiator? Either on the bike or off.

it should read 4 k round about at 10 dec c

for many of us out here w dont get that cold so we see ussually 2.4 k round about at 20 c

you could take it off and out it in boiling water with some alligator clips attached ...it should get close to 100 ohms ...but not sure exaactly

you have to do the test to both the sensors in the brass casing as one feeds the injection control the other feeds the fan relay box

pin 10 on the injection control under the seat (and earth )

have a look in the efi ignition troubleshooting data ....on the portal page ...its listed there

Since it's a Motronic equipped model things looks a bit different.
Sensor ohm values is similar, but it's a single element sensor and connected
to the ECU through contact # 21 and 32.

ECU connector config. is found in this earlier thread about the fan issue,
where the conclusion was an internal problem in the Motronic unit...........
which I still believe is the problem (if both all of the contacts and pins at the
Motronic unit connection is clean and free for corrosion).

BTW. The incorrect signature have been mentioned earlier......................

I assume the test procedures remain the same though, just with the different values.

Yeah, I tried multiple times to change that signature, but it doesn't want to save and stay. I'll click save and it just reverts.

makes sence inge ....i was still thinking 8 valve ....doh !

Rabidchiwawa007 wrote:I assume the test procedures remain the same though, just with the different values.

Inge K. wrote:Sensor ohm values is similar

But ECU connectors and sensors is different.

Rabidchiwawa007 wrote:Yeah, I tried multiple times to change that signature, but it doesn't want to save and stay. I'll click save and it just reverts.

I'll guess you have tried to change it at the information tab, instead of the signature tab?

Hah! Got it.

Well, I took the radiator off, cleaned it out a bit, and it looks to be in great shape. Water flows freely through it, as well as air.

I also took the temp sensor out and tested it in boiling water. I also tested the extra one I have lying around. One is from a 96k mile engine and the other is from a 29k mile engine.
When I put the sensor in rapidly boiling water, the 96k mile one measured down to ~196 ohms. The 29k mile one measured down to. ~202 ohms. Both were clean and free of corrosion / build-up. A previous post above stated they should get around 100 ohms in a pot of boiling water. Should I look in to a new sensor?

After I put it all back together with regular coolant (no flush this time), it actually runs a little bit warmer for some reason. It's still within normality while moving, but boooo.

I also purchased a new fan for $24 USD. It should pull up to 3x as much air as the stock 20yr old fan. It's rated at 1550 cubic feet per minute. I'm not sure what the stock one is rated at, but it's probably around 300 or 400 CF/M.
http://item.mobileweb.ebay.com/viewitem?itemId=260885562107&index=6&nav=SEARCH&nid=10354940937

the 100 ohms ish ....is for the 8 valve mate .

dont know the specs for the 16 valve k100 - k1100 ...sorry about that ...

but sounds like you have a bit of a result so far

Rabidchiwawa007 wrote:I assume the test procedures remain the same though, just with the different values.

Inge K. wrote:Sensor ohm values is similar

At 100^oC you should read ~183 ohms +/- 10%, so your sensors seems to be OK.

Where do you get that 100 ohm value from, Charlie?....it would correspond to
around 125^oC.

right at this moment .... i cant think where i got that figure inge /...perhaps too much water round here sending me cabin crazy ...and too much turborg in the system ?

grin !!!

i do seem to remember 180 odd now that you mention it ...