I would like to calibrate my speedo using a signal generator before I put it back in the gauge cluster. Is this possible?  If it is can anyone tell me which hole is for what?  I have a list of the frequency for each speed from the forum so I know what the values should be.  I also have an oscilloscope so i can generate any kind of wave I need and adjust the duration of the pulse.  Is there any particular type of wave I need to generate?  I.E. Square wave, Sine wave etc.?

I'd be interested is reading a detailed thread on how you get it done. Good luck with your calibration.

Regards

There is a program developed by one or more members of the German Flyingbrick forum. IIRC its named Karamba, which runs on a computer. By connecting the sound card (headphone connection) to some contacts on the speedo you can simulate any speed you like, whilst considering tyre size etc. Have a look around/do a search. Worthwhile little program!.

I did download it but it keept shutting itself down after a few minutes or if I changed the bike type on my computer.  I can do the same thing as that program using a signal generator.  In the tech section I found a list of frequencies for each speed in 10 MPH increments up to 100 mph.  All the signal generator does is give you the ability to generate the pulse and the scope lets you adjust the duration of the pulse in micro seconds.  I discussed this with my dad yesterday (he is a physicist who worked in micro electronic R&D at 3M for a looong time)  It's great having someone who is way smarter than I help me with this.  I will definitely have a write up and pictures for those who are interested.  I just want to make sure I have the specifics correct before I see him.  What I don't know is what each of the 3 holes are for.  I am sure one is + other is - and the third is signal but I need to know which is which.  Here is the table from the write up

Speed, MPH	Frequency, Hz
10	13.33
20	26.67
30	40.00
40	53.33
45	60.00
50	66.67
55	73.33
60	80.00
70	93.33
80	106.67
90	120.00
100	133.33
110	146.67
120	160.00
140	186.67
150	200.00

These values are for a k1100, are these valid for a K100?

I borrowed this picture from the thread link below. It's one that Inge K. posted. I think you will find it helpful.

https://www.k100-forum.com/t7575-can-i-use-this-tach

On the cluster harness, Pin 22 is the yellow from the speedo sensor and pin 23 is the ground from it.

Having done quite a few of these I can tell you that turning the potentiometer roughly 45 degrees counterclockwise will get you close.

Yes thank you, that is exactly what I need, but it is showing the connections for the tach not the spedo.  That is what he put on my other thread when I was inquiring about if I could swap my tach's around.  He probably has not seen this post yet.  I wasn't planning on putting the spedo back on the circuit board until I get it calibrated and I have just ordered a new set of gears for it.  So i will have to wait for those to arrive at any rate.

If you do as Duck says you also test all connections between the cluster connector and speedo connector, the connectors you want to connect to is a known problem source.

And it also a circuit between the wheel sensor and the speedo, which amplifies the signal and change the vaweform from sinus to square.
The pcb is at the bottom in your picture,...BE is input, BA output.

Ok so my best bet to check all the connections is to install the spedo and pull it to tweak the pot if the speeds are off then?  So you have a picture of where the connectors are on the large connectors on the back like you did for the tach?  So I should input a sine wave and it will convert it to a square wave?  Sorry for so many questions, I just want to make sure I understand and can tell my dad the correct information.  Can I expect the same speed values for those freq from table above?  The spedo motor is a 12V motor correct, there is not some kind of step down in the voltage to it? Or at least the input should be 12V as I will be using a external power supply.

Thanks again for all your help

I suspect the frequency/speed table is dependent on the rear axle ratio and rear tire size. Or, it would depend on at least the rear tire size.

Duck allready said where to connect, but add 12V+ at pin #6 to that.


This could also be helpful:



About the table....miles is a unknown language to me.

50 km/h - 20,4 hz
100  "    -  39,4 "
200  "    -  76,0 "

tgolden wrote:I suspect the frequency/speed table is dependent on the rear axle ratio and rear tire size.

Rear axle ratio changes the engine speed, not the wheel speed.

Inge K. wrote:

tgolden wrote:I suspect the frequency/speed table is dependent on the rear axle ratio and rear tire size.

Rear axle ratio changes the engine speed, not the wheel speed.

Thanks for that clarification. I wasn't sure if there was a reduction between the wheel speed and where the rear speed sensor picks up the signal.

JM02:

Your chances of getting it "perfect" the first time are what they are.

Get it where you think it's close, install it and then go for a ride with a GPS to validate it.

Then tweak it accordingly.

I've used Karamba a few times and the end result is has usually been that it still needed some additional minor tweaking to get it just right.

For example, I used the K1100LT settings of Karamba on my K75 turbo (because that's the drive train and wheels/tires I'm using) and Karamba didn't yield perfect results - it reads a tad lower than it should. It's pretty close now but I still need to go back in and tweak it a bit.

My bottom line:  Trial and error may be required to fine tune it but, in the end, it's much better than wasting your time with all sorts of math trying to make sure it's "perfect" the first time.

My approach is to put a sine wave into the cluster on pins 22 & 23, be it Karamba or a soldering iron to the speedo sensor - because that's what your speedo sees.

Personally I've always found that messing about with computers is a lot more trouble than sticking a jeweller's screwdriver into the slot on the speedo and giving it about 1mm turn is quicker and gets it as near as I'll ever need.

The rear wheel speed is very much dependent on the weight of the rider as the heavier rider will make the tyre flatten out more than a light rider and so increase the number of rotations per mile or kilometer or furlong if you measure speed in furlongs per week.

RicK G wrote:furlongs per week.

  I like the way you think.

If you're still wondering about the wave form - it's a 'square wave'
The sensor in the rear hub is just another 'Hall' sensor, which, IIRC, will give 6 pulses per rear wheel rotation. Distance travelled is dependent on the wheel/tyre radius where it touches the road (ie dependent on the bike's load).

The signal that comes from the speedo sensor is a sine wave.  It is a magnetic analog device.

The Hall Effect sensor is different and is composed of "switches" that do generate square waves.

Hi there, the issue regarding the distance vs load seems interesting, here is how I see it.
Distance is measured based on the perimeter of the wheel, which is dictated by the rim´s diameter, for each turn you´ll advance one perimeter. Now consider the effect of the load, this consist in flattening the contact patch until the wheigth divided by the flatened area (ie contact pressure) equals the tyre pressure.
The tyre pressure remains constant as long as the load do not exceeds the permissible one, this means that the tyre´s volume also remains constant and hence the perimeter, regardless of the flattened contact patch.
Being the perimeter independant of the load, the speed measure would not be affected.
What do you think?
Best regards
JAP

The distance traveled is dictated by the circumference of the TIRE, not the wheel/rim.

http://www.csgnetwork.com/tiresizescalc.html

The link below will give you the equations to compute vehicle speed based on tire/wheel, gear ratios and engine RPM.

http://www.clarks-garage.com/shop-manual/trans-02.htm

duck wrote:The distance traveled is dictated by the circumference of the TIRE, not the wheel/rim.

http://www.csgnetwork.com/tiresizescalc.html

Yes, you can change the circumference by changing the tyre but  once you set a tyre, the circumference won´t change with the load.
Best regards
JAP

japuentes wrote:

duck wrote:The distance traveled is dictated by the circumference of the TIRE, not the wheel/rim.

http://www.csgnetwork.com/tiresizescalc.html

Yes, you can change the circumference by changing the tyre but  once you set a tyre, the circumference won´t change with the load.
Best regards
JAP

I think that depends on the level of precision used in the equations. A specific tire's rolling diameter is variable and a function of vehicle weight, tire pressure, lean angle, temperature, tire wear, and tire rotational speed. Some of the variables are considered negligible for the typical level of precision used in these types of computations.

I think that if you intend to calibrate your speedo by generating a specific frequency in Hz that corresponds to a specific MPH indication and tuning the speedo potentiometer so that the indicated speed is correct, you need to generate your own speed/frequency table that corresponds to your rear tire size. The idealized correct frequency for a specific speed in MPH is based on rear tire size alone. The idealized correct frequency would need a correction factor to determine actual correct frequency to compensate for weight of the bike while riding and tire wear.

Regards

tgolden wrote:I think that if you intend to calibrate your speedo by generating a specific frequency in Hz that corresponds to a specific MPH indication and tuning the speedo potentiometer so that the indicated speed is correct, you need to generate your own speed/frequency table that corresponds to your rear tire size. The idealized correct frequency for a specific speed in MPH is based on rear tire size alone. The idealized correct frequency would need a correction factor to determine actual correct frequency to compensate for weight of the bike while riding and tire wear.

Regards

Agreed. This should help:

http://classickbikes.com/ckb.tech/0.ckb.tech.files/K.Bike.Tire.Size.and.Speedo.Cal.xls

Thanks duck. I was going to post the equations to compute the frequency/speed table, buy why reinvent the wheel.

Nice work.

Regards

Yes thanks, most helpful.  I really am ok with the spedo being a few mph off.  My car has read 3 mph slow since the day I got it and I drove it off the show room floor.  Once I get it reasonably close I will probably just mark the correct speeds with a grease pencil.  It would just be nice if the mileage would indicate correctly though.  I always use the trip odometer for my gas/mileage calculations on trips.  Idiot lights suck, they never go on until you are to far gone to fix a situation.

I used the Karamba programme a few years ago after repairing and replacing the broken cog set inside my speedo on my '87 K100RS. The odometre is now dead nut spot on and the speedo, with the coordinates of a 160/60ZR18 tyre measured round the centre and those measurements plugged into Karamba, with incremental adjustments made at the back of the speedo head, reads on average about one point five kilometres per hour higher than actual, as calculated using GPS, at motorway speeds. That'll do me.

Some computer's settings upset Karamba and make it unreliable. I mucked with it for hours on one until I gave up. It now works on one of my laptops, como Dios manda, but not the other.

I tried it on my netbook.  The netbook is designed for long battery life and the amplitude of the headphone jack signal it puts out wasn't enough to work with Karamba.

bomber737 wrote:Yes thanks, most helpful.  I really am ok with the spedo being a few mph off.  My car has read 3 mph slow since the day I got it and I drove it off the show room floor.  Once I get it reasonably close I will probably just mark the correct speeds with a grease pencil.  It would just be nice if the mileage would indicate correctly though.  I always use the trip odometer for my gas/mileage calculations on trips.  Idiot lights suck, they never go on until you are to far gone to fix a situation.

I have a suspicion that all the car manufacturers purposely set their instruments to read fast by a couple of points .....no good in saying to the officer with the speeding ticket in his hand ....honest, "it was on cruise control " and set for the right speed ...
don't work in a court of law ...

I once over heard a senior officer as the question of a senior age driver (who admitted he was over by a few points ...but was depending on the percentage over - under law to give him grace )...and when exactly did you have your last calibration of the speedo done ????

we do take too much for granted sometimes ...including some sales guys trying to offload some nice and expensive over sized tyres ...as "the best thing since sliced bread" and "will remarkably change things and the way it feels on the road "

nothing like doing the maths first....and being "aware"

duck wrote:I tried it on my netbook.  The netbook is designed for long battery life and the amplitude of the headphone jack signal it puts out wasn't enough to work with Karamba.

yep pretty hard to drive a signal past 3.3 volts on a netbook ...maybe lucky to get 5 ...they all have high current audio drivers and poor current reserves (read nice big juicy current dumping capacitors) on chip these days ...its all demand and switch it faster stuff ......erk !

I might add that our "in house"  old style power supplies (that we have sold thousands of) have about 200,000 microfarads ....per rail,  and most are over 15 years old ...and still going fine ..."quite remarkable " in this day and age ...all these new fangled switchmode power supplies only seem to last at best 7 years ...before they die horribly , many of these old supplies hold the power up during the now common ...intermittent power outages ....a blessing for us service guys . phew !!!

The speedo calibration will change as the rear tire wears. I measured the tread depth on a new Metzler 130/90-17 at 0.257". That tread depth equates to a 2% change in wheel/tire circumference as the tire wears from new to no tread (worn) at center.

The frequency change from new to no tread for the tire described above, using duck's very helpful spreadsheet program, yields the following:

Speed (MPH)     Frequency-new     Frequency-worn
       60                   76.94                    78.48
       80                   102.59                 104.64

The pulses/mile changes from 4614 (new) to 4709 (worn).

I would recommend measuring the actual wheel/tire circumference and comparing it to the computed value for a new tire. It may be better in the interest of speedo accuracy to use the mean diameter (measured) of a new tire and completely worn tire.

Regards

You always can refute the ticket by demanding to see the radar gun calibration records. This goes both ways

If you get a speeding ticket, crank the speedo calibration pot to the extreme and go get a shop to give you a speedo check certificate. Take it to court and have the case reduced to faulty equipment.