SiteLock
Magnetic Speedometer Correction - KLR650.NET Forums - Your Kawasaki KLR650 Resource! - The Original KLR650 Forum!
KLR650.NET Forums - Your Kawasaki KLR650 Resource! - The Original KLR650 Forum!
Go Back   KLR650.NET Forums - Your Kawasaki KLR650 Resource! - The Original KLR650 Forum! > KLR650 Discussion Forums > Procedures & How To Articles
Register FAQ Members List Calendar Mark Forums Read
Procedures & How To Articles KLR650 Modifications, Procedures, Tech Info


Reply
 
Thread Tools Display Modes
Old 11-12-2008, 07:56 PM   #1
Len P
Active Member
 
Len P's Avatar
 
Super Cool Since: Nov 2007
Locale: Hollywood, Florida
Postings: 645
Len P is on a distinguished road
Default Magnetic Speedometer Correction

Before I start I want to say my inspiration for this thread comes from another in this forum where a KLR 250 speedo was calibrated the same way. I tried it and wanted to share the details so anyone brave enough to try it can avoid the mistakes I made and have easy success.

Wattman's fix is much more straightforward and easier and I wish to take nothing away from his method. I personally did not want to move the error to a different part of the scale, however insignificant the error is.

I recently recalibrated my speedometer using Neodymiun Iron Boron magnets. This is a very easy procedure to do and also very easy to screw up, but you can recover, I did. I want to share the details so that anyone willing to try can avoid the mistakes I did. While this can be done, please read everything I write and ask any questions you might have before trying this. I do not want anyone to be disappointed or angry if this fails for you.

To start with the KLR 650 odometeter is very accurate. While disassembled I did a tooth count on the odometer gears and calculated the input ratio as 2240 revolutions per mile. The front wheel gear is 1 wheel turn to three cable turns yielding a ratio of 746 2/3 wheel revolutions per indicated mile. Given the diameter of the stock tire the odometer is perfectly calibrated when about half worn. If you divide 63360 (inches in a mile) by 746 2/3 you get 84.857 inches which is how many inches the odometer assumes you go per wheel revolution. If you mark the wheel and the road and roll the bike you can determine how far you travel per wheel revolution and compare it to the 84.857 for true odometer accuracy with your given tire.

Given how close the odometer is I chose to calibrate the speedometer to the odometer. To do this you need a way to drive the speedometer. By running it at a steady speed, use a stopwatch to determine time in seconds per mile. 3600 divided by seconds per mile is the speed that should be reflected by the needle. Mine showed 65 when the odometer timing showed 61. That fits right in with the average of 7 to 8 percent high.

While demaging the speedo is real easy, it is nearly instant and regulated more by distance than time, meaning if you get too close for a brief instant you then will have a negative error. The magnetic field can be restored, however. I had mine reading 34 for 61 since I originally mistaked the field as two poles, it is actually four. using the NIB magnet I was able to restore the field until I once again read 65 for 61 leading me to believe the ring magnet is originally at saturation.

Enough babble, on to procedure and pics.

Take the fairing off. Three screws and the turn signal removal is all that is necessary. Maybe you can do the procedure without removing the fairing but my hands could not fit in to get the speedo harness plug loose. Having the extra room is worth the few minutes it takes for removal.

There is a cover below the instrument cluster. It is shown below in the red box, the blue circles show where the attaching screws are.

Once the cover is off separate the plug shown below in the blue box. When separated both sides fall free of the bracket, allowing the rear side to come off with the instrument cluster. The speedometer cable and the two attachment nuts are shown in the red boxes. When removing the nuts watch out for the washers as they are not attached, they will fall free. The rubber grommets remain after the cluster is removed.



Below is the rear of the instrument cluster. Before disassembling the unit you must remove the tripmeter thumbscrew. It has a small phillips screw that must be removed before pulling it off. After that removing the screws below circled in red will allow the rear cover to be removed, gauges still attached. The screws circled in blue allow the speedometer to separate from the rear cover.


Below is what the cluster looks like before removing the speedo.


After removing the speedometer it is a good idea to mark the ring for polarity to make remagnetization easier if it becomes necessary. Below shows how I did it. The poles are not very easy to determine since the outer ring is steel, effectively shunting a lot of the magnetism around the ring. Aim the compass north, then place the compass as shown below and identify the two points located 180 degrees apart that are attract the compass north, mark those points "S". Then turn around, point the compass south, place the compass as shown below and identify the two points that attract the compass south and mark those "N". These are the poles of the ring magnet inside. The marks should encircle the outer ring in a N-S-N-S pattern every 90 degrees. If not, recheck. If you overdemagnetize these points are where you will need to remagnetize, so make sure they are right. You can use a regular magnet to determine the poles as well, just do not use the NIB magnet as it will disturb and weaken the field.





Once at this stage I set up a high tech driving mechanism shown below. Be careful not to overtighten the chuck on the cable end as it is a blade connection and could be damaged. When on high in reverse my speedometer read 65 and timing the odometer revealed 61. The input ratio of the odometer as mentioned above is 2240 revs per mile, so a drill turning 2240 revs/min will show 60 when the speedo is right. I would try to pick a drill in that range. Other types of drives, especially synchronous motors might be better but connecting them gets complicated.


The magnets I used were bought on Fleabay, 1/4 by 1/4 by 1 inch long, N50 Neodymium Iron Boron, magnetized through the length, meaning the poles were on the 1/4 by 1/4 points. Most sold are magnetized through the thickness, poles on the 1/4 by 1 side. I recommend through length magnetization since this is what I used. I would try attaching the magnet to a popsicle stick or pencil with about 1/4 inch of wood protruding beyond the pole to avoid the magnet pulling into the ring. This is what I did as I wrongfully assumed the outer ring was aluminum, IT IS STEEL. Once it grabbed the magnet it was too late.

Get the drill running at a steady speed and clock the odometer. I clocked one mile. Accuracy increases as you clock increased distances, but so does error if you use a less stable motor (drill vs synchronous motor).

3600/(seconds per mile) is true speed. Once you know true speed you gradually approach the rotating ring from the side with either pole pointing directly at the rotating ring while watching the needle. As you get within 1/2 inch you will begin to see the needle drop. DO NOT LOSE CONTROL OF THE MAGNET. Approach very gradually and draw away. If it is still too high get the tiniest bit closer the next time and draw away. Once you think you have it clock the odometer again and compare it. This is really easy to do. Just maintain control over the magnet.

If you go too far remagnetization becomes necessary. One of the magnets I used was not strong enough to restore what I had removed. I stacked four magnets to make one strong one. Placed then side by side to make a 1/2 by 1/2 by 1. You need to stack them with like poles in the same direction and this is extremely difficult. I stuffed mine into a plastic pipe and hose clamped them together. You could buy a N50 magnet with a 1/2 by 1/2 pole, I just worked with what I had. Determine the poles by suspending it or using a compass. Suspended the north pole will point north, and this pole will draw the south mark on a compass. Use the north pole to recharge the south marked poles on the ring, and the south pole to recharge the north poles. Use a swirling motion over each mark. Run the speedo up to verify that the magnets are once again too strong, then try the process again.

I have used a Schwinn Bicycle computer before and after to check calibration. I set the wheel diameter at 2155, as this is the mm equivalent to the 84.857 inches that the odometer assumes for each wheel revolution. Before the correction my speedo consistently read about 7 percent high compared the bicycle computer. Now from 10 to 80 they always read the same, under 1 mph difference.

Wattman's fix is much more simple and straightforward and keeps you out of trouble. I did not want to move my error to a different part of the scale so I tried this. I have tried to make this as clear as I can. I will add more as it comes to mind.

One more piece of important info. The inside of the speedometer is sensitive to contamination, especially the magnetic type. You would not want to attempt this on a bench near metallic grindings or shavings. The magnet in the cup would likely capture the particles rendering the unit worthless. Try to keep it as clean as you can.

Len




Len
__________________
'06 KLR 650 Red/Silver

Last edited by Len P; 11-12-2008 at 09:52 PM. Reason: More info
Len P is offline   Reply With Quote
Old 11-13-2008, 06:46 PM   #2
Len P
Active Member
 
Len P's Avatar
 
Super Cool Since: Nov 2007
Locale: Hollywood, Florida
Postings: 645
Len P is on a distinguished road
Default Re: Magnetic Speedometer Correction

If anyone wanting to try this suspects the speedometer pointer has been relocated in the past it will need to be moved into its original position. The following example shows how to detect and fix the error.

A clocking test is needed at two different points on the speedo, the further the points are apart the more accurate the results will be. Suppose at low your drill causes the speedometer to indicate 33(S1IND) and clocking the odometer indicates 30(S1ACT). You then put your drill on high and get 66(S2IND) indicated and a odometer clock speed of 60(S2ACT). Use the following formula to solve for spring rate error. The * means to multiply.

((S2IND-S2ACT-S1IND+S1ACT)/(S2ACT-S1ACT))*100=Spring Rate error in percent

((66-60-33+30)/(60-30)*100=3/30*100=10 percent

In a properly indexed speedo this would result in 1mph error per 10(11=10, 22=20, 33=30, etc.)

For the following formula ERR is the percentage as a decimal, 0.10 in this case.

S1IND-(S1ACT*(1+ERR)=the needle position error

33-(30*1.1)=33-33=0, in this case the needle is in the right place.

Lets suppose the needle has been moved 5 mph lower than it should be. S1IND is now 28, S1ACT is still 30. S2IND is now 61, S2ACT is still 60. All I did was subtract 5 from the two indicated speeds. Using the same formula we get the following results.

((S2IND-S2ACT-S1IND+S1ACT)/(S2ACT-S1ACT))*100=Spring Rate error in percent

((61-60-28+30)/(60-30))*100
3/30*100=10 percent The spring rate error is still the same.

Now using the second formula we get the needle position error.

S1IND-(S1ACT*(1+ERR)=the needle position error
28-(30*1.1)=28-33=(-5) The negative 5 indicates the needle is 5mph lower than it should be.

If correcting the needle is necessary I would retry the test to assure the needle is in the correct position before trying to recalibrate the magnet.

I had all this figured out, thought it might help somebody out one day if I posted it. I know I am nuts. Who else ever figures this stuff out.

Len
__________________
'06 KLR 650 Red/Silver
Len P is offline   Reply With Quote
Sponsored links
Old 11-13-2008, 06:50 PM   #3
Scott_Rides_Again
Been here a while
 
Super Cool Since: Apr 2007
Locale: 410m N, a smidgen W of Spoonbooty
Postings: 15,633
Scott_Rides_Again is a glorious beacon of lightScott_Rides_Again is a glorious beacon of lightScott_Rides_Again is a glorious beacon of lightScott_Rides_Again is a glorious beacon of lightScott_Rides_Again is a glorious beacon of light
Default Re: Magnetic Speedometer Correction

Thanks Len! Great job!
Scott_Rides_Again is offline   Reply With Quote
Old 11-13-2008, 09:39 PM   #4
Len P
Active Member
 
Len P's Avatar
 
Super Cool Since: Nov 2007
Locale: Hollywood, Florida
Postings: 645
Len P is on a distinguished road
Default Re: Magnetic Speedometer Correction

Many thanks Scott.

I need to clarify the remagnetization process. I mention a swirling motion to restore the field. This is correct. What I did not clarify is that contact is necessary. Let the poles touch while swirling the magnet over the marked poles. Just do not stray too far from where you mark the pole.

This is the remagnetizing magnet I made. These four really do not want to be together, but they are very strong this way.


A demonstration of strength as the one above holds the other six through over 3/4 inch of wood. This is not a trick. These magnets are super strong.


Another pic of them in a chain. They are suspended in the center by the bottle. They are joined lengthwise by magnetism alone. The four on the left are stubbornly pointing north. If you place one of these magnets on a smooth enough surface they will orient in a north/south line.


Len
__________________
'06 KLR 650 Red/Silver

Last edited by Len P; 11-13-2008 at 09:46 PM.
Len P is offline   Reply With Quote
Old 11-15-2008, 08:45 AM   #5
Swiftbow
Been here a while
 
Swiftbow's Avatar
 
Super Cool Since: Oct 2003
Locale: Santa Fe, NM: North of Spoonbooty
Postings: 4,111
Swiftbow is on a distinguished road
Default Re: Magnetic Speedometer Correction

Thanks for putting this together Len!

I haven't worked on this mehod but thought of it many times after Moat posted about the magnetic trick with his 250.

Doug
__________________
"Accelerating into ... neutral"
Swiftbow is offline   Reply With Quote
Old 11-15-2008, 09:24 AM   #6
bentmettle
A Regular
 
Super Cool Since: Sep 2008
Locale: Ann Arbor, MI
Postings: 75
bentmettle is on a distinguished road
Default Re: Magnetic Speedometer Correction

fantastic writeup.

I don't recall reading it in the directions, but keep rare earth magnets away from your computers, laptops, portable hard drives, and CRTs (including old tube TVs).

Magnetism scares me
bentmettle is offline   Reply With Quote
Old 11-15-2008, 10:56 AM   #7
Len P
Active Member
 
Len P's Avatar
 
Super Cool Since: Nov 2007
Locale: Hollywood, Florida
Postings: 645
Len P is on a distinguished road
Default Re: Magnetic Speedometer Correction

Many thanks guys.

Yes, Moat was the one that did the 250. Glad you mentioned him. I mentioned his post and I wanted to show the link, he has a lot of good info, but I got distracted trying to remember how to post pics. He gave me the nerve and inspiration to try this. His pics were able to show the magnetic ring. The 650 speedo cannot be disassembled like that without risking ruination. It is entirely crimped together for tamper resistance, I suppose, although I believe mileage could be reset by removing the faceplate. The crimping might just be easier for mass production.

I believe mass production is to blame for the error these speedos have. I read somewhere that the US requires 5 percent accuracy on production models, 10 percent on per model approval (?KLR?), but in neither case can there be an underread. Seems easy enough to design a unit that falls in the 7 percent overread category without requiring further calibration.

And, Bentmeddle, I am glad you mentioned some of the hazards of the rare earth magnets. I meant to put it in there and forgot. Note the time and date of the original post, 8:56pm Wed night. I spent nearly two hours putting the post together and one of my favorite shows, Mythbusters, was about to air. I got in a rush. I went back later to edit the end and the disclaimer slipped my mind.

The damage that occurs to any tv or crt screen can be permanent as the field can be strong enough to deform the shadow mask inside.

Finger and eye injuries are also possible as these magnets can slam together from at least several inches apart, although I am not sure if the chips can escape the field. There are several good videos on youtube showing their power. Saw a slideshow showing a seriously injured finger with the fingernail sticking out from between the magnets. Never know if it was real or a setup but I assure anyone the bigger ones are capable of taking your fingertip off. The small ones I have shown will snap together from around 4 inches apart.

Anything that holds magnetic data (floppys, vhs tapes, security badges and credit/debit cards, etc) can surely be ruined as well. I would bet even memory chips and cell phones could fall victim.

Watch these magnets, they might be more hazardous than learning how to wheelie.

If I can find a sacraficial unit around I might try a how-to-video and post it here. The others vids posted in this sub-forum go much further in helping than pics and print ever can.

Len
__________________
'06 KLR 650 Red/Silver
Len P is offline   Reply With Quote
Old 05-02-2010, 06:59 AM   #8
Len P
Active Member
 
Len P's Avatar
 
Super Cool Since: Nov 2007
Locale: Hollywood, Florida
Postings: 645
Len P is on a distinguished road
Default Re: Magnetic Speedometer Correction

An update 18 months after correcting my speedometer with this method.

My speedometer is still holding the same calibration 5800 miles later. No long term side effects from changing the magnetic field.

While I never expected any trouble I knew only time would tell.

Len
__________________
'06 KLR 650 Red/Silver
Len P is offline   Reply With Quote
Old 04-22-2011, 12:11 AM   #9
Swinada
A Regular
 
Swinada's Avatar
 
Super Cool Since: Apr 2008
Locale: Peace River Area, British Columbia
Postings: 61
Swinada is on a distinguished road
Default Re: Magnetic Speedometer Correction

I know this is a very old thread but could the pictures be updated so they are visible again. I only get red X`s.

never mind, must have been my computer after right clicking on the pictures and selecting show pictures they appeared.

thanks for the write up.

Last edited by Swinada; 04-22-2011 at 12:14 AM.
Swinada is offline   Reply With Quote
Sponsored links
Reply

Tags
correction, magnetic, speedometer

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump

All times are GMT -5. The time now is 03:25 PM.


Powered by vBulletin®
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
© Copyright KLR650.NETAd Management by RedTyger