[Mark in MT]

Getting a speedometer to report accurately over the whole range is a challenge. I have done a couple thousand, some are a lot easier than others. It is quite common for them to be right on at 20 mph, and at 60 mph, but a couple mph high at 40 mph. I have not figured that one out yet, though it may be the hairspring rubbing on itself at different speeds.

I have a Stewart Warner "Speed Analyzer" circa 1940's that I have calibrated to check my repaired and restored units on. Some I can get in 2 minutes, some take a 1/2 hour to get into spec. Below 20 mph, they are very hard to get accurate, ditto 65 plus. I concentrate the accuracy in the 20-60 mph range. About 10% of the units I see require the magnets to be recharged, as they will not get up to speed with the adjustment at maximum. You can't just take tension off the spring cause it won't return to zero.

I had one unit that gave me fits, couldn't get to calibrate. I noticed that there was oil on the hairspring, the viscosity between the coils was enough drag to make it inaccurate. Reading low on the way up, high on the way down.

Calibration of 90+ year old instruments is a lot of fun with the proper equipment, a real challenge with out.