I have a 0-145 MPH mechanical speedo that has a proportional offset in the MPH reading. In other words, it reads low, and it gets worse as the speed increases. The drive gearset in the trans is correct and my stock mechanical speedo reads within a MPH or two of true speed, so its the new speedo thats out of calibration.

Are there any companies out there that recalibrate these? Is a DIY recalibration possible?

 

It's fairly easy to recalibrate, in principle; all it is, is a whirling magnet inside a steel shell, such that the faster the magnet spins, the more it tries to drag the shell along with it.

All that has to be done, is magnetize the magnet some more.

Back in the 70s, when I worked in a meter lab, we had all the stuff to do that. Nowadays, with DMMs having supplanted moving-pointer meters, there's probably not many places left. Speedo shops maybe.

But... since speedos are all basically 1000 RPM = 60 mph, you can just get another speedo from anything that will fit, and stick it in your cluster, and nothing will be the wiser.

 

Yeah I was thinking of trying to find some small neodymium magnets of the right "sense" and seeing if I could trim it to read roughly what it should.

I have one that reads correct in my dash, but its only a 115 and not the 145 that Id like to fix.

Another thought I had was possibly switching the face and clock spring to the 115 one since it feels like the springs are what set the deflection vs. RPM. Going to have to find a way to measure the actual torque at full scale and see if the two springs are directly proportional to the differences in the two scales on the faces.

 

A small torque tester, something on the order of a few inch/ounces full scale, would be necessary. Even then, the assembly only balances the movement (indicator needle) against the torque of the magnetic coupling.

It sounds like the speedo error is rate, not incorrect zero In other words, the error is a consistent X % off across the range instead of a fixed X MPH through the range. That will be a span/rate and not zero error.

Can you close the gap of the magnetic coupling?

 

Take the scale off the one and put it on the other.

You'll notice that the span from say 0 to 100 is the same on both.

Main obstacle you might run into is the location of the trip odo reset stalk.

 

Speedometer error comes in 2 basic varieties. If the needle is clocked incorrectly, then it is possible to always be the same number of MPH high or low. But calibration error must always be a percentage. If you're 5% off, then at 100 MPH it will read wrong by 5 MPH, while at 60 MPH it will read wrong by 3 MPH in this example.

 

Ok I decided to test it properly and set up a test bench with a scope and spare optical VSS to actually measure the RPMs at the rear of the speedo. I could only get it up to 80 on the gauge, but the calculated RPMs from the optical VSS showed both the 145 and the 115 are reading the same and within a few % of the true speed.

Not sure why the 145 was so far off when I tried it in the car. I did notice the old speedo cable that I replaced had a lot of garbage in the end of it, so maybe it was slipping.

Long story short on the road test the speed matched the speed on the scan-tool, so its accurate for the most part. Only problem is that it doesn't read past 135 even though the car was going around 150 . Guess I cant expect everything from a 30 year old gauge.