It appears my speedo/odometer is not working. I removed the instrument cluster and took off the back. I checked the three pin connector which didn't show any corrosion, so put it back together. I unplugged and put back the connector next to the underseat tray. I removed the rear speed sensor from the final drive, cleaned it and replaced, but still no joy. I have noticed the indicators do not self cancel....does this point to a faulty sensor? Should I try the soldering iron trick?

The soldering iron trick will eliminate the sensor so well worth a try.

Either the soldering iron trick or try Kramba / KDiag.

Thanks for the pointers. I'll try soldering iron first as the lower tech option...

A run out on the A12 in London has thrown up a feature....the Speedo works once I get to about 40mph, and appears to be accurate. As my speed drops below this, it falls to zero once more 
So... I'll display my logic for your assessments!

If the Speedo works at all, that means the connector at the tool tray isn't at fault as at least something is getting through.
The Speedo is reading, although not under 40mph, so this means the Speedo internals are functioning.
The sensor could be at fault, hence the faulty reading under 40mph. This can be tested by applying my original plan, which was the soldering iron trick, to verify this. I'll clean it up again (in daylight this time) and see if anything improves. If not....replace sensor.
How does that sound? Reasonable? 

sort of sensible
if you have on going issues could it be that the vaned impeller inside the final drive has become loose  and is now free wheeling ?
I believe we have seen this a couple of times , or heard about it at least

I'm thinking the impeller would resist turning as rate of acceleration is changing (and sitting in diff oil ), but once a more constant speed is achieved it then catches up sort of thing ?

just a thought

I guess you could test it out by probing - observing the speedo sensor hole  whilst on the centre stand and selecting a gear and running the bike static ?
a good torch might help out ?

Thanks C99.....good call on the impeller. Will try checking that as well.

I'm kind of siding with Charlie. I can't imagine what could be wrong with a sensor that can cause it to function above a set frequency. The problem sounds more mechanical in nature, and might be related to the motor driving the speedometer. Remove the guts of the cluster and inspect the speedometer region for interference from dust and debris or from mechanical damage. There is a dampening mechanism on the speedometer needle that dampens its return to zero that might be acting up.

hi rob , currently working on a faulty speedo issue  on some new to me instruments

with Karumba only , I have to max out the volume output on the pc to get a good spread of speedo reading
which is different from this pc , than others that I have calibrated before  , usually only need to get about 30 % volume for a good consistent  drive

I have already fixed a burnt out track on the membrane circuit (an earth reference actually )
but think the input buffer of the op amp has been upset somewhat and I have a fair bit of dc on the input of the op amp

I have a replacement to hand (lm358 rather than lm2904 ) to install
 so hope this may resolve the issue

but it could be a similar issue has happened to others ?

time will tell I guess

Does the soldering iron have to be a particular spec in terms of the EMF it is omitting?
Stil investigating dodgy speedo. 
The final drive seems OK...removed the sensor, and with bike on stand, engaged 1st gear and observed the action in the drive:. which was engaging.

No, the soldering iron does not have to be a certain spec of EMI.  Even if it was, I wouldn't know what that spec would be nor how to measure it.

A soldering iron that uses mains electricity to heat the tip usually uses a tungsten heating element to do so.  This element is wound around a core and generates a magnetic field at the mains frequency.  A temperature controlled soldering iron won't work for this application because it generally doesn't generate a magnetic field.  Barring the availability of a main powered soldering iron, you could use a modified power transformer.  Some transformer cores, such as those used in door bells or microwave ovens, use "E" and "I" sections.  if you can break apart the "I" section from the transformer, you'll have the ideal magnetic field generator.  For example, see my Instructable to make a watch demagnetiser to understand how to do this.

The best tool for the job is Karamba ( or it's replacement Kdiag ), which uses a PC sound card to generate the proper signals for testing a K-bike speedometer.

I'll throw this into the mix... At the weekend I repaired a speedo that a few years ago had been showing similar symptoms and had been chucked in the 'I'll-get-round-to-it-sometime' box. The problem turned out to be that one of the two clockwork springs that act as needle dampers had come out of position and the spring wasn't re-coiling correctly. A slight push down with a screwdriver fixed it. The big problem is that the speedo needs to be stripped in its entirety to get at the clockwork springs .

Yes, the soldering iron does have to emit enough of an electromagentic field to register on the sensor. Some do, some don't. BTDT.

I have a couple of orange soldering irons that look pretty much identical. One registers, one doesn't. (So I wrote SPEEDO in Sharpie on the one that does.)

Blimey! So how do I know if my soldering iron is a good'un or not!?! A catch 22!
In the meantime I have ordered a second hand sensor from Motorworks for £25...not a bad deal, and at least that rules the sensor out of the process.

second hand sensor from Motorworks for £25...not a bad deal, and at least that rules the sensor out of the process.

So long as it is a working one!

Mike

I did think of that...but that would apply even to a new one!
I'll let you know what occurs. 

Twas the sensor after all! The second hand one works a treat....! Happy dayzz