Author Topic: Help Please - Cooling Fan Not Working (Read 3282 times)

Falcon 269 · « **on:** 02 March 2020, 05:58:08 pm »

As anyone who knows me knows, I hate troubleshooting electrical problems, so I'm appealing for help.

I’ve already hit the Search function which helped me with steps to investigate but a solution has eluded me so far.

Bike is a 2002 Gen 1, 53k miles. Yesterday I started her up just to circulate fluids and make sure all was well before advertising her for sale. Last time ridden was a couple of weeks ago and all was fine then.

On start-up, I’m 99% sure that the coolant fan came on as it usually does soon after starting. Went off with gear selection and I then proceeded to hold the revs around 3k rpm and blip the throttle to get the motor up to temperature. Usually when I do this I let the coolant get to around 90C and then turn off but this time I decided to let her run to check that the fan came on when the temp reached 98 - 99C. It didn’t. The Multigauge coolant temp continued to climb to 104C at which point the coolant overheat light came on - as it should - and I shut her down.

Today I started troubleshooting by applying 12v direct to the fan. It works fine.

The fan fuse is good.

Applying 12v to the relay contacts I got a ‘click’ suggesting it’s working correctly. I also substituted another relay and that made no difference to the fan operation. I’m calling the relay itself good.

Earthing the green/black terminal on the relay connector turned the fan on, suggesting the wiring from relay to the fan is OK.

The temperature readings on the Multigauge were exactly as they’ve been since I fitted the gauge. Coolant temp rose normally and the overtemp light came on as it should at an indicated 104C. This suggests that the coolant temp sensor in the thermostat housing is also good.

The coolant overtemp light comes on briefly at ignition on as it should.

However, the fan no longer comes on soon after start with the revs held at 1500rpm plus and the coolant fan does not kick in when the temp reaches 98 - 98C.

I’m stumped at the moment and the only thing I haven’t checked yet is the combination unit in the speedo which should be activating the fan relay when the temp gets too high. I believe it also has a timer function to run the fan at start-up.

Any ideas, experiences, hail Mary guesstimates would be much appreciated.

Mike

schlumpf · « **Reply #1 on:** 02 March 2020, 06:45:26 pm »

Hi Mike,
as you've ruled out all other potential fails, I suspect it's the control circuit inside the cockpit. I've detected and reapired such a defect a while ago. Maybe it's the same in your unit.

The picture shows the affected part of the pcb. Most likely transistor TR34 is not working. Ignore the red coloured parts, that's from another, older discussion about manually activating the fan.
According to the marking code (DF) it's a standard NPN transistor, maybe 2SC2463-F (55V, 100mA) (those markings are not clear without ambiguity). Any similar type should also do fine. Also replacing with a wired (non-SMD) type will work. I know, it's not an easy job..
For testing you could replace the temperature sensor by a wired resistor of less than 1700 ohm. That equals ~105°C. The voltage at anode of D32 should drop from 12V down to ~1.5V, the collector of TR34 accordingly (fan relay or equivalent load must be attached).

Hope that helps.
Regards,Christian

old son · « **Reply #2 on:** 02 March 2020, 07:56:44 pm »

Still sounds bloody complicated to me!!!

Falcon 269 · « **Reply #3 on:** 02 March 2020, 07:57:40 pm »

Hi Christian,

Many thanks for replying so quickly. I posted on the FZ1OA as well and grommet there was equally quick off the mark. He also mentioned you might be the man to have a chat with about this.

You both seem to conclude that the problem is most likely on the control board in the instrument cluster. I'll have a look inside tomorrow for obvious physical signs of failure but testing components isn't one of my talents. I could probably replace an SMD, though, provided I can get the correct component.

Please excuse my ignorance but could you simplify the process you mentioned for testing by replacing the temp sensor with a wired resistor? Where would I put the resistor in the circuit and how best to achieve this? If it can be hit with a hammer, I'm good to go but anything involving wiggly amps and squiggly volts melts my brain cell.

Cheers!

Mike

Falcon 269 · « **Reply #4 on:** 02 March 2020, 07:58:34 pm »

Quote from: old son on 02 March 2020, 07:56:44 pm

Still sounds bloody complicated to me!!!

Damn right, mate.

At least you don't have to try and wrap your head around it like I do, though.

schlumpf · « **Reply #5 on:** 02 March 2020, 08:26:08 pm »

Good to see that grommet thinks the same way.
For the dummy-sensor, just pull the connector from the sensor (at the thermostat) and insert the leads of a wired resistor. That mimiks a certain temperature.

The sensor itself is a NTC type and changes resistance over temperature. A few values:
80° = 3700 ohm105° = 1750 ohm120° = 1150 ohm
The dashboard switches to ON below 1750 Ohm, and OFF again around 2000 ohm. There is a small hysteresis (intended).Secondly it switches ON based on a timer (1 minute?, at idle, if you once revved above 1500 rpm). Whatever condition comes first.
For repairing you could "hammer the transistor out", means there is no harm if it gets destroyed. A big blob of soldering tin above all legs and pulling out. I guess it's also glued to the pcb, might need some persuasion. Take care to leave R33 and R34 in position. A good trick is to fix them before with kapton tape, or even mask the whole surrounding. Also very helpfull in case you use a hot air gun, if available.Be carefull with the heat, the copper striped come off easily.
For the transistor replacement you could use almost any NPN transistor. BC817 is a very common type in europe.

schlumpf · « **Reply #6 on:** 02 March 2020, 08:40:40 pm »

One more old picture.

Also I have to correct myself with the part. It's not a BC817 alike due to the package (it's not SOT-23). Must be a bigger type, maybe 2SD1898. That also fits better in terms of current capacity. Have to investigate further tomorrow..
Maybe others out here have a better clue.

Falcon 269 · « **Reply #7 on:** 02 March 2020, 09:51:36 pm »

Great stuff, Christian. Thanks so much for your time and knowledge.

Since I'm getting what appear to be reliable temp readings from the Multigauge, I presume the purpose of substituting various resistors is to see how the dashboard responds without having to run the motor up to excessive temperature?

Coincidentally, I bought a quality soldering station recently and saw Youtube videos on removing/replacing SMB components like the transistor you suspect of being faulty. If you could confirm the type I need, I'll be happy to give it a try. If it's broken, I can hardly do much more damage, can I ... ?

I look forward to your next post.

Mike

teecee90 · « **Reply #8 on:** 02 March 2020, 09:57:57 pm »

I had very similar symptoms with my fan. Turned out to be a problem with the fan cables - specifically the section of the cable where it doubles back on itself before entering the fan motor casing. I suspect the tight angle of the cable and the high temperatures in that location probably caused the cable to break down at that point. Initially I opened up the fan and replaced the cables, which fixed the problem, but I later decided to just replace the fan and it has been fine so far.

I had checked everything else like you (although not the instrument PCB). Eventually took the fan off and connected it direct to a 12 volt battery and noticed that went on and off as I moved the cable around.

Might be worth checking before venturing into the PCB?

schlumpf · « **Reply #9 on:** 03 March 2020, 06:58:40 am »

An overall test to locate the failure is quite simple.
- Remove the dashboard
- you see the 16 pin connector, loom side
- locate the green/black cable
- connect it to ground
Fan is running -> failure inside the dashboard
Fan is still dead -> failure in the relay, the fan itself, or the cabling

(edit: correct formatting restored)

Falcon 269 · « **Reply #10 on:** 03 March 2020, 07:55:25 am »

I'll do that this morning, Christian. Simple tests ... I like those.

teecee90, many thanks for the suggestion. I was aware of your recent fan trouble and checked the supply cable as one of my first moves. Fan works fine from 12v direct to blue/black connector and also by jumping the contacts in the relay connector. Visual inspection of the bend in the cable shows no stress or sign of failure, either. If only it had been that simple, eh?

schlumpf · « **Reply #11 on:** 03 March 2020, 09:12:15 am »

Some more results on searching for the "DF" marking of TR34.
Package seems to be SOT-89. Function assumed to be NPN bipolar transistor 80V 1A.
Matches:
2SD1898
TFM5213
2SD1623S

My best guess for replacements:
BCX56
or
BSR43

Falcon 269 · « **Reply #12 on:** 03 March 2020, 10:02:50 am »

OK ... well, not really ... grounding the green/black pin in the cluster connector brought the fan on, so the fault is in the dashboard. Deep joy.

All the components look so big in the photo above, so small on the PCB. Doh! Still, if I can source the TR34 replacement based on the info you've supplied this morning, I'll give it a go.

Then again, I might just wire a manual fan switch and leave it at that. I've hardly ever had the fan come on in normal riding, even on the hottest days in Spain. Only when spending ages in traffic and that's very rare for me. I used to have a manual fan for those days anyway, so that I could get the cooling started before the coolant got above 90C.

Thanks again, Christian. You're a star.

Falcon 269 · « **Reply #13 on:** 03 March 2020, 10:10:09 am »

Just confirming, is this the part required?

https://www.ebay.co.uk/itm/10x-BCX56-10-115-Transistor-NPN-bipolar-80V-1A-500mW-SOT89-NEXPERIA/202687225861?hash=item2f3119a005:g:yLEAAOSwHcBc5nb1

schlumpf · « **Reply #14 on:** 03 March 2020, 11:28:55 am »

That parts looks fine.
But hold on. I'll have a another look to the board this evening. Maybe it's possible to repair from the back side of the board, without removing the old component.

Falcon 269 · « **Reply #15 on:** 03 March 2020, 11:41:48 am »

That would be good. Otherwise, it looks like I have to remove the tacho and speedo from the circuit board to get at TR34.

schlumpf · « **Reply #16 on:** 03 March 2020, 03:56:44 pm »

This is my proposal on how to fix the broken fan activation w/o removing any old parts.
New parts are added on the lower side of the board, no need for complete disassembly, removing the lower compartment of the dash is sufficient.
The circuitry around TR34 is copied and put in parallel to the existing one. This should be fine as it only serves as a simple switch to ground for the relay activation.
Unfortunately there are a few more components to be added than simply TR34.
The replacement parts are best guesses by taking the marking codes, package type and estimated currents and voltages required for this application into consideration. All are non-SMD.

Replacement for TR34 is BC639, 1A/80V.
The proposed diode is a standard 2A/400 silicon type (most likely oversized).
Resistors are uncritical, smallest package with 200mW rating should be fine.

I'm open for discussions and opinions about this solution.

Falcon 269 · « **Reply #17 on:** 03 March 2020, 04:32:31 pm »

Deleted due to formatting issues.

Falcon 269 · « **Reply #18 on:** 03 March 2020, 04:38:17 pm »

Thanks, Christian. Unfortunately, you are dealing with an utter numpty when it comes to electrical matters. I thought Ohm's Law was a TV series.

I'm sure the diagrams and explanation you provided mean something to many here but to me it is still baffling. I can't speak Klingon, either.

I did manage to rig up a manual switch today, tapping in to the green/black at the connector under the left hand fairing infill. I have that running through a 3-pole, single throw LED illuminated switch. Green/black to one pole, earth to another. I can get either the LED to light or the fan to operate but not both together. I tried 576 permutations of wiring connections through the switch before deciding that life was too short to retrain as an electrical engineer. I shall focus my efforts instead on trying to become the oldest player on the PGA Tour.

Earlier, in a fit of enthusiasm and over-confidence, I ordered a ten-pack of transistors to replace TR34. I will spend a few days practising SMD component removal on some old PCBs I salvaged a while ago. There is a fine line between having enough alcohol to steady my hands while keeping my vision clear enough to see the infernal bits. I shall enjoy experimenting to find the perfect balance.

Cheers!

Mike

PS. Joking aside, I will try to crack the code on your latest suggestion and if successful, I'll add it to the list of things to try.

schlumpf · « **Reply #19 on:** 03 March 2020, 05:13:23 pm »

Sorry for my tech talk, I know it's not everybody's world.. But perhaps you find someone nearby to help with the repair?

I could also think about creating a small simple board that's carrying all the parts. As a drop-in module, attached with 3 solder wires?

Falcon 269 · « **Reply #20 on:** 03 March 2020, 05:39:15 pm »

I will certainly see if there's any electrical repair specialists near me but finding someone I can trust and explaining my needs in Spanish could be a bit tricky. I'm thinking this probably needs to be a DIY project.

If you could come up with a pre-assembled board that is virtually idiot proof (good luck with that!) I'd be happy to go that route and pay you for your time, parts and postage.

schlumpf · « **Reply #21 on:** 04 March 2020, 04:13:32 pm »

Ready, send me your address via PM please.

This is a quick and dirty buildup with parts I had in my box. Cable colours match the dots in the sketch from above.
I've tested it good on a salvaged dash (original fan circuit disabled). I had no relay at hand and took a big resistor instead, should not make a difference.
Some additional wrapping for isolation and it's good to go

Falcon 269 · « **Reply #22 on:** 04 March 2020, 05:09:22 pm »

Fantastic, Christian. Thanks so much! Even a dunce like me can fit that.

PM on its way soon.

Again, many thanks - really appreciate your knowledge and time on this.

Mike

agricola · « **Reply #23 on:** 04 March 2020, 07:30:13 pm »

Top man Mr Schlumpf

ogri48 · « **Reply #24 on:** 04 March 2020, 08:25:58 pm »

seconded. brilliant thread