Triumph Rat Motorcycle Forums banner

Broken Speedo Sender Fix / Also mod for 05+ wheels on Early model

184K views 247 replies 56 participants last post by  Big Bopper  
#1 ·
This alternative Speedo Sender could be employed by those who have a broken speedo magnet

Triumph won't provide the magnet - which is frequently broken after re-assembly of front wheel after service, tire-change etc - as a separate line-item and you have no option but to purchase a complete speed sender assembly - not cheap!

Now - instead of paying $175 for a new Triumph sender, execute this mod for ~ $25!
The best part is it will never break again!

Also, for anyone who would like to use the late model (05+) wheels on their early models,
And this will also work for those doing fork/wheel swaps for example, where no provision for the OEM speedo drive exists on the replacement wheel.

Note: easier for Sprint - all years use 25mm axle;
02-04 S3 & 02-06 Daytona use 20mm axle so more involved for wheel swap


you have been previously prohibited since the new model has no speedo drive detent in the wheel.

Well, I have the solution for you!

I used a very similar solution on my GSXR fork retrofit to my Mrs SV650 and felt pretty sure that the same method could be employed on the Triumph.

The simple fact is that the sender is nothing more than a Hall-Effect sensor - a magnet creates a field induced current as it sweeps past a detector.
Most bike share a common process but the mechanics of the device and where it is located varies.

Now on the 99-04 series of Sprints (or similar era Speed Triple, Daytona, TT600 etc), you have an electronic speedo that receives its input from the ECM which in turn receives a signal which is generated by a front wheel sender.
On the 05+, the principle is the same although the sender is based on the output shaft and of course the ratio of pulses generated is quite different. The output signal to the instument is CANBUS and quite, quite different, but all we need to concern ourself with is the input to the ECM.

So how to create a signal on an early bike using late model wheels that has neither a device from the transmission or the wheel?
The simplest method is to replicate the action of the wheel driven sender if possible - by using a frequency based on the same rotation diameter (i.e. the wheel!) no additional calibration should be required.

Back to the SV for a moment - we'll use this method as a basis for the swap. The SV happens to use a very similar wheel/axle based sender on the front wheel as the 99-04 Sprints.
When upgrading to GSXR front end, those wheels pose similar problems in that the donor bikes utilize a final output drive sender.

I was able to take a GSXR sender - there are after-market Hall Effect Trigger available, but GSXR sensors are plentiful & cheap off EBay (less than $20) .
I mounted this to a simple bracket using the inside of the fender stanchions and it is triggered off magnets simply epoxied on the rotor.
It works perfectly on the SV!

So - would that work on the Triumph?

Since I had all the bits at my disposal I thought I'd check this out - it's actually been on my agenda to validate this for some time anyway.

I actually have on-hand an old GSXR harness which not only donated the plug that obviously fit the GSXR sender, but also conveniently gave up another which fit the Daytona harness speedo input!
So I made a simple extender cable with each of those connectors on either end.

Then I took the speedo sender off my TL (TL/Hayabusa/GSXR/SV1K all have used the same sender for about last 10 years!) and plugged it in the extender, then the extender into the Daytona harness.
Turned on the Triumph ignition, and per the previously suggested test, just took a magnet and osciallated it back & forth across the sensor face, while monitoring the instrument display.

Bingo! Good reading on the speedo commensurate with the rate at which I was moving the magnet.

So this validates that the process works!

There are 4 magnets used for the Triumphs.

Here's how it came together on the SV - you would use similar for the Triumphs.

This is the GSXR/TL/Hayabusa/SV1K sensor
(common to all from about 99 - current):

Image


This is the sender mounted on a simple bracket on my SV
- the bracket is mounted to the fender stanchions with modified (very short) screws so that they dont interfere with the actual fender mount screws on the outside of same stanchions:

Image


Image


.

The mounting of the sender will clearly be a little different on the Sprint forks or early Speed Triple/Daytona forks, but should not be very challenging.
The important thing to know is that it will work electronically


Magnets mounted on the rotor are the triggers.
They just need to be positioned on a radius that sweeps past the sensor location.
This is the source I used for the rare-earth magnets
They are also available on EBay if you do a search

Here is a convenient mounting location in the rotor buttons!

Image


Image



Important Note: You should only use the rotor buttons as a mounting location if you have an equal multiple of 4 and thus can space the magnets at exactly 90 deg intervals.
Spacing the magnets non-symmetrically will result in speedo error.

.
.
.
Here is wiring relationship between the OEM Triumph sender & the Suzuki one

.................Daytona Harness .......Triumph Speed Sensor .........GSXR sender

Power .... Black/Yellow.................. Red/Orange ......................... Black/Red .
Ground.... Black ............................ Black/White ......................... Black/White
Signal ..... White/Blue ...................Pink ...................................... Black

You can use crimps if you're not comfortable soldering - but because the GSXR cable is really short, it will give you bulky cable just up from the sensor.
Best if you can solder and use heat-shrink sleeve over each individual splice, then another heatshrink over whole joint section.

When you splice the sensor cables together you can ignore the first block above (Daytona harness)

From the GSXR sender to the donor Triumph sender cable:

Connect Black/Red to Red/Orange
Connect Black/white to Black/White
Connect Black to Pink


Figure out how you're going to mount sensor first - once you have optimum position for that then the magnets will take care of themselves.
You can make this on either side of course, doesn't matter which leg/rotor you set it up on.

One thing to note is that the magnets MUST be applied with poles in same direction.
Easiest way to do this is to stack them then mark the top of each one with a marker as you peel it off the stack.
Then assemble with the dots all facing out.

If you can get rare-earth (Neodymium) magnets so much the better - they are small and have really strong field, so should be detected over quite a gap - on my SV set-up they are 1/4" and maybe 5mm away from sensor - with the rare-earth I'm sure they would trigger from further away.

You can actually use any kind of Hall Effect sensor - here is another option -> Cherry GS100701

Image


Another mounting method:
See this really cool mounting concept from Jen A in post # 133 later in the thread - http://www.triumphrat.net/speed-tri...oken-speedo-sender-fix-also-mod-for-05-wheels-on-early-model-4.html#post2005378
EXTREMELY well executed mod!

Image


See the actual post from link to get the full details.

.
.
.
 
#2 ·
Your the brains here mate, I am going to try this on my 06 Daytona, (see pm):) will the distance from the hub make a huge difference, I am going to mount the magnets on the dics rotor near as possible to the spindle, can I simply remove the Suzuki plug and fit bullet conectors to hook it into the triumph harness?
 
#3 ·
Thanks for the info, as posted elsewhere the magnet is in four pieces, pulled the wheel this morning, funny how it can do that while in position, the wheels not been out since a new tyre last october time, looks like it got trapped somehow as the breaks are to the rear edge of the lugs? will do the job this coming weekend as I will have to locate a magnet supplier and also waiting for the sender to arrive, will post piccies of the whole job and such, am looking at mounting the whole thing on the rhs as that will be a direct route for the cable, will be soldering the joins, and mounting the magnets on the disc roter I think.
 
#4 ·
I'm not sure where you would find in the UK, but try looking in hardware store - or cabinet supply?

Here are the ones I used -

http://www.leevalley.com/hardware/page.aspx?c=2&p=58750&cat=3,42363,42348&ap=1

These are intended as kitchen cabinet door latch devices. They work well for this application because you can fasten the retainer cup (whether by nut & bolt or by epoxy); then just press in the magnet.
Again, these are tiny, but have amazing field strength. Neodymium is what you're looking for.

I fastened mine with quick setting (90 sec) epoxy - it's not like there's any load on them other than centrifugal force. If you have to remove them, you can still snap off with pair of pliers.

Here a place I found in the UK
http://www.magnetukonline.com/materials.asp

Also, Here's some on EBay UK
 
#6 · (Edited)
Ordered some from a craft site on ebay ÂŁ2.98 for twenty 10mm magnets delivered, whole job will cost ÂŁ12 and some time, have the metal and the bracket will also attach to the fender bolts, they extend through plastic clamps on the daytona so will fit the bracket with nuts on the inside,light ally right angle bar will extend about five inches from the mount down the forkleg to the hub for the sender then epoxy the magnets onto the disc roter, the wiring will be as the diagramme, looking at about two hours work then some time playing around to find out how many magnets it will need going for two first then ride out with the misses on her varadero to check calibration.
 
#7 ·
... then some time playing around to find out how many magnets it will need going for two first then ride out with the misses on her varadero to check calibration.
You shouldn't need a calibration factor - it's going to be pretty obviously on-speed, or maybe 2x - I really don't think 1/2 speed.
The speedo will calibrate exactly as it was before, since rolling circumference is same and can only vary by the multiple difference of # of poles rotating past the OEM sensor.

Were those neodymium magnets you got? Or regular ferrous type?
10mm sounds 'big' so I'm thinking the latter.
The EBay ones I linked would definitely work well.
 
#9 ·
Quick tip before you actually install the hardware

Once you complete the wiring integration, plug it in, turn on ignition and holding sensor in one hand, oscillate a magnet past the sensor face with your other - you should see the speedo indication and vary with the frequency that you move the magnet. That will validate for you that your sensor is actually wired up correctly & functional (just save scratching head later if you find you're not reading anything)
This essentially replicates the test that I did on my own bike (the Triumph I mean)
Once you validate that, then you know your system will work and just a case of mechanical installation.

Then, after you install, before you take it out, just spin wheel and see that you have reading on the gauge - this will again validate for you that the magnet proximity appears to be close enough to trigger detector.


I was looking for a good front fork/wheel pic of a gen2 S3 or Daytona:

It looks like a good mounting point for the sender would be from the lower fender mount - then the magnets could go on/in the rotor buttons (this is a common mounting location for the magnets with the SV guys)
With the sender just below the fender mount, it would be fairly unobtrusive and I think work quite well there.

.
.
 
#10 ·
ITS ALIVE ITS ALIVE, works as described, I mounted the bracket on the fender mounting bolts as I said, the sensor lines up with the buttons and I started with two poles result, exactly half MPH, four poles bang on the money, very tight for space though as there is only 4mm between buttons and fork stantion, used 3x10x1mm magnets on four opposing buttons, they wont go 12, 6, 3, 9, but close and it seems fine, tested against my wifes varadero through all speeds up to 100MPH. Pictures will follow when wife sorts them and sends them to my album. total cost ÂŁ7 for second hand sensor on EBAY and ÂŁ2.98 for 20 10x1mm magnets from same source.
 
#12 · (Edited)
...You probably don't need three magnets at each location - will probably work fine with one at that spacing I would have thought?.
Possible but they are so small that a stack of three works and still blends in well, will sort out the routing of the wire a bit better at some point but very happy with the result, especially the cost, problems whilst doing the job were, as warned polarity is crucial as I found out when after finding two poles read half, so stuck on two more the wrong way and still read half, DOHH, sensor height in regard to magnets is pretty important, slot your mounting holes to give some up and down adjustment mine would not read the magnets untill directly in line for height, also hence three on each pole, apart from that quite straight forward have not soldered the wires either tinned the bare triumph wires and plugged them in as did not want to butcher the suzuki harness untill I was happy the job worked, first pic, bits needed, second lining bracket to get hieght, third, close front view, two nuts hold it on the inside of the fender mounting bolts, four, front veiw of set up with sensor and magnet fitted, five this is what it looks like from the normal viewpoint.

Image
Image


Image
Image
 

Attachments

#13 ·
Re painted bracket today and tried putting the sensor in the vertical, no speedo, seems it is really distance critical and the act of putting it vertical moves it too far from the poles! re routed the cable to the inside of the forks and round to the standard clips, all good, while the wheel was out cleaned the calipers/pistons and fitted new brake pads, oh and cleaned the blooby thing:eek:
 
#16 ·
Where were you last summer..

nice, I like a schmuck payed my monies to bike bandit to replace my dead spedo sensor last year, I may have some really small hall effect sensors lying around, (1/4" by 1" reads from face with a threaded body. 3 wires ) they seem to like 5 volts, may work real well for the application, as I recall they would pick up anything ferrous within a few millimeters may not have the resolution at speed though..came from some scraped medical supply manufacturing equipment, quite nasty caustic environment I suspect. I'll have to look for them.
 
#19 ·
... do you leave the old trimph sensor in place or is there something to replace it and cover up the void?...
You can either just leave it in place (remove the rotating magnet - or what's left of it - and the drive ring from behind the grease seal) or have a spacer bush machined to replace it.
Former is probably just fine but if you have access to lathe then simple spacer to make.
 
#20 ·
It's speedo time!

Changed my tires last week and today the speedo was acting a fool. I removed the front tire and sure enough the magnet is two pieces! :eek: I would like to attempt this mod. My question DEcosse is, how many magnets do you think I would need for my TT600. Also, after visiting the link you provided, which magnet would you suggest, (A,B,C,D or E?) thanks.
 
#21 ·
TT600 will be the same - 4 magnets. One magnet in 4 locations is what is required.

It depends on where/how you want to mount them - 1/4" or 3/8" is plenty big enough to trigger the sender.

If you search 'Neodymium magnet' on EBay you will come up with no end of shapes & sizes!

Note that on the S3/Daytona rotor it has 10 mounting 'buttons' - not conducive to symmetrical spacing for 4 magnets using those locations.

doodledog suggested that it does not matter if they are symmetrical - in his case he reported that the assymetrical layout still works.

Edit - the asymmetric spacing may cause an instability of the speedo at higher road speeds
I would advise to keep the spacing symmetric at 90 degree intervals to eliminate this from being a problem


Image
 
#24 ·
Got the sensor on Saturday and I'm still waiting on the magnets. They are coming from hong kong...so that'll take another week or so.

In the mean time, I decided to work on the electrical aspect of the mod since both pieces are triumph, they used the same color coding for wires. Pretty straight forward.

I then decided to test it, to see if it would work, by using any magnet laying around the house (Magnet on the back of a speaker, a magnet used to remove security locks and even the original broken sensor magnet) and It just wouldn't work. I was able to determine which side of the sensor to use but still no cigar. However, whenever I would tap the sensor on the magnet, the speedo would begin to work. going as high as 22mph!

Is it because I'm using some cheapo magnet or that maybe the sensor itself is defective?
 
#25 ·
Don't know what to tell you - it might have been better to stick with tried/tested sender
I would say that you shouldn't necessarily assume wires are same just because colours are same - you would need to check the wiring diagram to be sure.
Any magnet should work, so something wrong with sender or wiring I would have to assume
 
#30 ·
tried/tested and still nothing


Alright, I tried it differently originally for sakes of using triumph parts and also because the bolt opening is located at a 90 degree difference from the GSXR. On a TT600 the stantions are located way higher than other bikes. Check this pic out. So I was trying to see if using the two bolts at the bottom of the fork was an option. Anywho that didn't work....for reasons which I explained previously.

So I then decided to do it the tried/tested way. Acquired a 2003 GSXR sensor and unfortunately, I get the EXACT same results!! The sensor does nothing when the rare earth magnet is passed by the flat surface yet if I tap the sensor.....say on my chest, it then does. This is the same thing the 675 sensor was doing!

I checked the wiring connection a dozen times and they are as you posted it. I have no idea what else to check. Keep in the mind that my speedo was perfectly ok until the magnet cracked. Any other ideas? thanks!
 
#28 ·
..what did u use to mount the sensor to the fender??
..... Further what kind of bracked are u using?
Simple piece of angle aluminum will work

You can see it in this pic of doodledog's:

Image


Long enough to mount to both fender bolts and extending just far enough below the fender to align with the magnet fixing locations (the rotor buttons in this case)

.

You can see similar mounting piece of hardware - angle bracket - in my pic but in this case the angled bracket is orientated 90 different.

Image


Both are essentially the same even though slightly different fender mount.

.
 
#27 · (Edited)
Hello guy's..

This is a great topic. I think mine sensor/magnetic is also broken Last week i bought a 02 955i and the speedo meter looks likes faulty. low readings (15 Km/h at high speed). I wanna try this mod, im searching at: speedo sensor susuki/gsxr but cant find any sensor on ebay?.. Further what kind of bracked are u using?
 
#31 ·
Fair enough trying something new - I would have expected the 675 one to work actually, there are all pretty much same principle.

Since the GSXR one doesn't work either, there must be common denominator

Wiring between original and GSXR is

From the GSXR sender to the donor Triumph sender cable:

Connect Black/Red to Red/Orange
Connect Black/white to Black/White
Connect Black to Pink

The Pink Wire is actually the signal wire.
Red/Orange is the power, originating from the cluster back out to the sensor - you could check that wire wrt the black/white to see that you hav epower with ignition on

I downloaded a manual for TT600 but the scan is poor and hard to make out the colour code - looks to be same though I believe.

There is a subtle difference between the TT600 and Daytona wiring systems - the TT600 has the input directly off the speed sender to the cluster; the Daytona is buffered by the ECM.
But should make no difference - the OEM sender is still just a Hall Effect sender - same as the GSXR - and most likely the 675 sensor too. So should have an adequate input buffer amplifier in the cluster to process the signal directly.

Is it perhaps how you are sweeping the magnet by the sensor?
This is going to appear condescending but not my intent at all ....
not like a piston going towards a cylinder-head & back - oscillate it back & forth across the face (simulating the magnets on the rotors scanning past it) - clear the magnet completely off the sensor on either side as you sweep across its face - not short strokes maintaining proximity of sensor.


On mounting on those forks - yes, that one is tough unfortunately - you may be able to do something with the caliper mounts?
Not the most atractive solution, but at least for validation purposes, you could also zip-tie it directly to the fork leg.
But of course need to get some basic functionality first ....
 
#34 ·
The Pink Wire is actually the signal wire.
Red/Orange is the power, originating from the cluster back out to the sensor - you could check that wire wrt the black/white to see that you hav epower with ignition on
What should I look for? When all the wires are connected the speedo will give me a reading. Just not an accurate one.

Is it perhaps how you are sweeping the magnet by the sensor?
This is going to appear condescending but not my intent at all ....
not like a piston going towards a cylinder-head & back - oscillate it back & forth across the face (simulating the magnets on the rotors scanning past it) - clear the magnet completely off the sensor on either side as you sweep across its face - not short strokes maintaining proximity of sensor.
Yes, I was sweeping it correctly but just to make sure, I ran to home depot, made a custom bracket (to mimic yours) and installed it all on the bike as it should. The sensor is less than 1'' from the magnets and still no go!

When I stop the bike and keep the engine running, the speedo gives me all these crazy readings from zero to 88mph. Just from the vibrations of the bike!!

I'm really starting to think that it has something to do with the wiring! :(
 
#33 ·
...The magnetic of mine is broken in 4 pieces, I thought maybe the magnetic can be glued or something?
Some people have had success with this - you can try it, nothing to lose really.
You need to be sure there is no epoxy interfering the free rotation. You may need to 'dress' it after it's cured.
 
#36 ·
Ok well, since your last post, I've decided to switch back to the 675 sensor (since I'm getting the same result no matter which sensor I use) and I closed the gap between the sensor and magnets to less than 1/4''.

I also re-checked the wiring (see pics) and still nothing.

I used my multimeter (with setting on 9v Batt) and connected the black prong to the black/white wires and connected the red prong to the orange/pink wire and the reading (with the bike on) fluctuates between .90 and 1.05.....not sure if I'm doing this right though!?!?


Here's the bracket I created.....
Image


Here is one of the many crazy readings I'm getting (Bike is NOT moving!)
Image


Here's the gap between the sensor and magnets....has been reduced to less than 1/4''.
Image


And lastly, here's the wiring (I removed the electrical tape for illustration purposes)
Image



Thoughts?