gday guys .....

for all of you that have had temperature issues i think i have the fix .....

rt showed me a temp sensor which he had pulled from his bike ....the diagnosis was by changing it he solved his temp sensor - fan operation issue .....there was more involved with the fan mechanism but we'll by pass that .

anyhow ...upon putting a mutimeter accross the temp sensor terminals i found there to be no issue ...both legs of the sensor mesured just right .

today i was exploring electricals and came accross an odd reading when probing the temp switch assembly .

(actually i was putting a bypass switch to ground out the temp sensor leg to activate the fan as an override )

i couldnt get anything from the sensor at all ...open circuit ...crikey !!!!

ok so doing some probing and removing the connector at the back of the sensor ( i thought the contacts on the connector was sure to be the issue )

i got conductivity from both legs to the right places ...efi and temp switch ...hmmm

i probed the connectors on the sensor mechanism and earth .....nada ....nothing at all ....what the ????

so i pulled the sensor ...(blocking the fluid release of corse) ... and measured it in my hand ....perfect ...

i reistalled it again ....dang nothing at all ....i retightened it ....still nothing

i measured between the stub and the engine in case of high resistance on the one bolt holding all together ....all ok

hmmm what to do ....

first thing i noticed on closer inspection was an aluminium crush washer and a brass housing on the sensor ... "click "

so with a little emory paper (wet and dry actually ) i cleaned up the brass collar of the sensor .....and the crush washer ...which was quite pited ...but looked only as surface corrosion and verry slight at that .there was no evidence of the white powdery corrosion that normally always accompanies this type of problem ...

any how the short version is ...dang the conductivity returned .....i would be guessing that most of us have been through the replacement process AND replaced the crush washer in this process .

my fix is ....clean up the surfaces where the sensor bonds ....

the threads i thought would create a good electrical contact ....and if you could clean them properly. probably would ...

so if you have a temp sensor related issue ....try this process first.



hope this helps someone

Brilliant deduction Charlie.

Have heard of that but never come across it my self. The copper oxide only conducts in one direction. You would remember the old copper oxide diodes Charlie.

I had never heard of copper oxide diodes. So this a natural property of copper oxide? What sort of currents could they carry?

Remembering that I did have a metal plate rectifier on my very first small bike, Suzuki 50cc `70.

wikipedia.

Inge K.

ah! ...yes i knew them a selanium rectifiers as well ....thanks for the link inge

Albyalbatross1 wrote:I had never heard of copper oxide diodes. So this a natural property of copper oxide? What sort of currents could they carry?

They were the bees knees at the time but were prone to overheating and had high internal resiatance but were much more portable than a vacuum tube diode.

Good to see you spending valuable off time for the good of mankind Charlie.
I need a heater here today, the world has gone crazy, and I think I better stiffen the suspension a bit till I can get back to normal food.
Cheers RT

oh ! just have to report ...back from a trip to day with other local members.

this is the first time i have acheived what seeemed to me to be about 25 kilometers per litre or 4 liters per 100

the sensor must have been working to some degree whether it is temperature related expansion of related components or lossy (high resistance ) possibly causing the temp sensor to read the engine at a lower temp than was actually occuring .

thats a massive 15 - 20 percent better running machine than normal

i have to say that the old "gerty von clickandshift" purred well today

I was impressed with Charlies improved performance yesterday. So have just spent the last hour doing the same with mine. I will go for a run on saturday and see if I can mprove on my 20k ltr.

I was discussing with a mech mate at work. he said charlies fix was spot on. The bike is no longer running CHOKED! I am going to check my old ute out tomorrow as well.

Ahh, the old smoke and mirrors (electricery) stuff. Well I can add my little bit from the mechanical perspective to help close the loop on this annoying issue. The temp sensor telling the EFI that things are cold (when really they're not) means a rich mixture is supplied. Hence black smoke and increased fuel consumption as well as reduced performance. Then the richened mixture washes the oil film from the cylinder walls and oil consumption increases. If this is prolonged, in extreme instances the oil can be smelt to contain fuel. Faulty thermostats can cause similar problems if they fail to close, which causes very slow engine warm-up and possibly not reaching correct operating temp at all. 1 way to check if the thermosat is functioning correctly is to observe engine temperature on a long downhill run. If the temperature drops excessively it is possible that the thermostat is not closing off fully and should be replaced. Remember, EFI engines are designed to operate within a narrow temperature range.

good advice dennis

Really nice work Chaz.


OG

My sons 1985 K100RS has over 100,000 miles on it and all of a sudden the bike will not start when hot. I did change the temp sensor and still have the same problem. I dont know if this has anything to do with the brass or aluminum crush washer that is on there. Or do any of you have another idea?

Measure the resistance of the sensor between pins 10 and 13 on the ECU plug. The bike has a safety that it can be run, but not restarted if it detects that its too hot. Look at Berrt's fault finding pages on this site related to the cooling water temperature.

as rene suggests it could be reading (resistance value ) verry low in ohms when it gets hot .......could it be a partial short to ground on the cables from the temp sensor .....from memory its really tight as it goes through the throttle bodies and onto the fuel rail let alone up the frame to under the tank (cable chaffing?) the under seat connections that rene suggest is ideal for measurement of this specific problem.

just a note ,,whilst several of us were away on a "tech " weekend at about 10.00 am (atmosphere was about 25 c and in the sun so could have been hotter ) we measured 3 bikes and found them within about 10 ohms of each other (around 1760 ohms )

rj ...i wonder if you have thought about the fuel regulator ,,,its tucked behind the tbs ...and we have seen some weird issues with them sticking open ... we havent established if this is temperatue related as well yet ...tank temperatures often get high also ...

could be thermostat partial operation as well ?does the fan operate at all ? durring or before this event ...or even if left running long enough ? we have seen issues there as well .

my initial investigation was due to lack of fan operation that made me suspect and start investigating .

food for thought ?

i had a wee giggle at "cable chaffing", it brought me to thought.
someone on this forum has , in passing , lauded the virtues of lanolin instead of wd40. Out of curiosity ( yes i am going through some cats!) i bought o can, and sam i am , converted. try please, its made from sheep smiles or something. i now dip my bikes ( 7 now ) in it at night !
I gonna start a new thread

Dennis wrote:Ahh, the old smoke and mirrors (electricery) stuff.  Well I can add my little bit from the mechanical perspective to help close the loop on this annoying issue.  The temp sensor telling the EFI that things are cold (when really they're not) means a rich mixture is supplied.

Mmm.. If you read my struggles with too fast idle in my posts on the subject you will understand why I'm interested in this. I think I must try cleaning that little sucker. I was wondering what the temp sensor/EFI connection was for.

Dennis, do you think it's possible that my fast idle(an understatement at 2000 rpm) could be due to EFI thinking it's cold outside?