Adjusting a Metron speedometer

[Mikenoangelo]

The speedometer on my Augusta, of the magnetic / hairspring type reads about 8-10km fast at all speeds. The distance reading is correct and the RPM (I fitted a really cheap digital revcounter which picks up a signal from the ignition lead) agrees very closely with the speed on a Satnav. 2000 rpm in top (25.7 kph at 1000 rpm according to tyre circumference and top gear ratio) shows about 52 kph on the satnav.

The over- reading seem to be the same at all speeds which suggests I could correct it by turning the indicator needle on its spindle. However I can't see any easy way of doing that accurately and I imagine there must have been some provision for calibration when it was made, but how?

I wonder whether turning the needle forward one full turn (which could easily be done by lifting it over the zero stop pin) would tighten up the hairspring and have the desired effect?

Any thoughts chaps?

Mike

[simonandjuliet]

My gut feel is "don't touch it". You know it over-reads so radar traps are not a problem !

But you risk breaking it if you do anything and then what do you do ?

Sorry not to be be more helpful

[DavidLaver]

The nature of springs is that they return to their original position...unless pushed into plastic deformation which is difficult to predict or control.

Is there the option of rotating so that it takes more effort to move it off its stop? 

Is there the option to rotate mechanism and stop relative to the display?

Are you willing to print a face to match the mechanism?

Is the spring removable to test and replace with another "correct" spring from stock?  At that point I expect you're off to an instrument specialist...

[lancialulu]

I seem to remember having this problem with an Aprilia with Metron (artdeco) speedo. I rigged up a drive to the naked speedo from an electric drill. If I remeber correctly I carefully pulled the neddle off after a drill activated speed was noted and by trial and error refixed it so the large fixed error was eliminated. Problem was as David said it took a while for the speedo to drive off the zero point i.e you were doing 20 mph before it registered.

I felt afterwards this was one of things I did not love about the car so we parted company. Sadly no regrets....

[Kari]

Just leave it as is to impress your passengers...

Move the pointer backward over the stop and note where it will come to rest. Then take the pointer off the shaft and replace it by the amount of degrees which it is over reading.

Please be aware that the dial near the shaft is not supported and very thin. Therefore some support should be used when using a levering tool(s) to take the pointer off. I use a piece of hard sheet metall.

I hope that will help

Karl

[Mikenoangelo]

That sounds good Karl thanks. The needle spring has almost no tension at the zero (10kph) resting point  and as David says, jumps to 20 kph as soon as the car moves. I can run it in the lathe before and after to check. If previous experience is anything to go by it should need 1000 cable turns per km so that 1000 rpm would read 60kph. We'll see.

I was stuck as to how to get a base line to re-orientate the needle but your idea solves that! As, unlike the Smiths/AT speedometer, there is no means of adjusting the spring tension, rotating the needle on the spindle must have been the original method of calibration.

Mike

[Mikenoangelo]

I ran the speedometer on my lathe, measuring the rpm with a digital tachometer and noting the distance clocked up over 5 minutes at a range of lathe speeds. This confirmed that the speedometer needs 1000 cable turns per kilometer as I guessed, and I could record the indicated speed at known “road” speeds. I proved to myself that the odometer reads a true distance by running the Augusta and my “modern” over the same route. I know the correction factor needed to find the true distance for the modern as I checked it over something like 100 miles on a motorway. The Augusta showed 34.9Km and the corrected equivalent for the “modern” (my 20 year old Isuzu Trooper) was 34.6).

The Augusta speed readings were consistently about 5 Kph high over a speed range up to 80 Kph which suggests that moving the needle on its spindle would put things right, rather than retensioning the hairspring. I have not done this yet as I noticed another so far unexplained oddity in that the speed shown by the needle when tested on the lathe, shows about 10 kph too fast for a few seconds when the instrument is connected to the lathe but soon falls back to the regular 5 Kph excess. Possibly some odd facet of the magnetic flux between the driving magnet and the driven element of the speedo, perhaps analogous to static and dynamic friction?

I also compared the speed reading of the instrument with that of a satnav which, most of the time suggested an error of plus 8-10 Kph rather than the 5 I had worked out, but which sometimes seemed to confirm my 5 Kph. It’s too hilly and twisty around here to get a long, straight, flat road comparison.

I’ll probably move the needle on the spindle after a bit more satnav work. It’s been a good diversion from decorating the bathroom!

[SanRemo78]

Can I suggest that you think about sending it to Speedy Cables? They're not always the fastest but will get it done for you. If I make it to the AGM this year with the Stratos replica come and ask to see the speedometer set up.... It's a tad complex but Speedy cables sorted most of it out for me!
Guy