Speedometer Calibration Tale

[mountaineerxii]

So on the maiden test ride around the housing development after rebuilding a 20yr old XII, I discovered a "lazy" speedometer. The needle responded, but very slowly up and down, more like a fuel gage, than a crisp speedo. Going thru the twisty's, one needs to know quickly how fast they are going before either leaning in or getting on the brake, so having to wait on the speedo to catch up with me wasn't going to work.

XII's are great because you can tear nearly everything apart to get to what's going on. I checked the wheel sending unit and cable first, but they were fine. I then proceeded to tear the instrument cluster down and extract the speedo for further inspection. Flicking the needle by hand, it definitely was lethargic, especially compared to the tach needle. As I looked at it, I surmised that the oil in the small little needle shaft bushing at the front was either dried up, or turned into gooey gunk. In my enthusiasm for finding root cause, I popped the needle off the shaft and applied liberal amounts of penetrating oil to the shaft bushing, which freed up everything quite nicely, resulting in a very flickable needle. After my "successful repair" euphoria wore off, I then developed a sinking feeling, realizing that the speedo needle shaft was not splined or keyed in anyway; meaning that the calibration relationship between the shaft, clock-spring, needle and home peg were all set, and then lost, when I removed the needle.

I have since re-calibrated the speedo, verified it with a GPS, and it's spot-on, without having to tinker with it further.

Here's what I did, you can follow it too.
- I accurately measured the outside circumference of the front tire within 1/16", with a brand new tire (properly inflated).
- With the speedo cable unplugged from the speedo, I rotated the front tire 10 times and counted the number of revolutions the speedo cable made (4 cable "flats" = 1 cable rev). I did this several times to verify.
- You can then do the math to turn road mph into wheel rpm, then into cable rpm, but to save some bytes here, 3000rpm of the cable is right at 80mph (which seemed way too much like an intentional design point to be a random coincidence)
- My variable speed drill just happened to be rated for 3000rpm. I chucked a mechanical tachometer into the drill and verified. After a few minutes of running, my drill topped out at 2990, close enough.
- I placed the needle back onto the speedo shaft such that the needle just kissed the home peg when at rest, basically a neutral clock-spring position.
- With my speedo unit and cable (laid out very straight to minimize drag) on the bench, I then chucked the drill to the speedo cable.
- I had my assistant carefully hold the drill and cable and start increasing speed (be careful of the drill rotational direction) until the speedo started gaining MPH.
- After a few seconds, I had the assistant give the drill about half speed, the speedo read about 65mph (should have indicated about 40mph).
- I then stopped the drill, grasped the speedo drum (at the back of the needle shaft), rotated the drum and held it at 25mph, then twisted the needle on the shaft back to zero.
- I continued starting/stopping the drill, and adjusting the needle on the shaft, until when the drill was at full speed, the speedo read 80mph perfectly. It took about 3 adjustments to get it right.
- I then put the speedo back on the bike, took it out for a test ride and confirmed with a satellite-based GPS it was on, using long straight stretches and cruise.

One word of caution. You will likely find a honey-like clear substance glop on the needle spring. This is silicone. It is used as damping to keep the needle from bouncing due to the irregularities of the mechanicals driving thru the cable. It is common on many mechanical instruments & gages. Don't wipe this fluid off or allow any penetrating oil to wash it away.

Surprisingly, considering only the one high-end calibration point, the speedo appears to be pretty close at low-end speeds too. I have found these mechanical units aren't perfect and appear to have some forgiveness built in. Playing around with it on the bench and doing the math, if the drill was off even a couple hundred rpm one way or the other, it meant only a 1-2mph difference, not much considering the resolution of the speedo display is only in 5mph increments, and not very big to boot.

Hope this helps someone out.

[cushman eagle]

Thank you for sharing your experience I don't think I would have been as calm as you were if I did an unindexed disassembly .I have learned the hard way over the years , to index components as I disassemble them,so I do not have to spend un billed time to get a customers tractor back to them.Today I used a camera to take pictures of a maze of lines and wires on a transmission shift tower before I got to the point of,how does this go back together? .Your method of calibration was brilliant!