I am fixing up the heating system on my Electra 1970, and while at have removed the instrument cluster. So, I thought I might as well fix some other things I have found but some questions came up. 1. It should not be possible to shift without pulling the lever towards you first, right? Because I can run through all gears by just just applying force vertically. Guess some notches inside there are worn. Thinking of if it is worth considering trying to repair, maybe it is possible to apply some welds here and there and grind up? 2. Is the speedometer adjustable somehow? The odometer is spot on but the speedometer shows too low. Thinking maybe the needle slipped down a few notches on its shaft or something like that. 3. Thinking about replacing the ignition switch. Many pages says their switch fits almost any GM in the 1970's. Is that to good to be true? I would like to have accessories and buzzer work like it once was stock, just because.
You can change the gears for the speedometer in the transmission. About the steering column, replacing the lock and key buzzer, download this handy PDF: Inside the GM Tilt Steering Column
If your odometer is exactly correct now, changing the speedo gear will change its accuracy of course. If you don't mind showing more miles on your odometer per year the speedo gear swap is easy.
I did calibrate the speedometer towards a handheld GPS (phone app). The data is plotted below. It seems that the measured speed is offset the true speed by a constant value. This is consistent with the fact that the speedometer needle is below zero when standing still. Pictures of other Buick speedometers shows the needle at zero at standstill. So a quick glance at this plot, moving the needle if possible and we could call it a day. But lets dig further to see if we have linearity errors as well, prompting gear changes etc. Not because I would bother with that anyway, but because I like graphs Focusing on the just the error in the plot below, we see the error is not quiet constant for all vehicle speeds, but ever so slightly decreases with vehicle speed. So we do in fact have linearity error it seems. But lets see how far we would get by just adding a constant offset to the measurements by moving the needle initial position. The mean error is 23.12 kph. If we where to add this constant to the speedometer (by moving the needle initial position) we could potentially get a mean error of 0.9 kph, or max 2 kph which is even way better then we need for this. So if it is possible to readjust the needle zero position practically, the problem should be solvable. The precision on this speedometer could potentially become better than I would ever have imagined.
