Speedo Calibration
using Desktop Computer
by Mike Walsted and Jeff Korn
Thanks to this post by Jeff K, I was finally
able to calibrate a couple of speedometers using my desktop computer, a length
of wire, and an old speaker extension cable from the 99cent store. I am not
recommending that anyone follow these procedures, since a wrong move could
result in a fried computer, or at least a fried sound card. Just in case anyone
wants to try, even after the disclaimer, here are the directions:
The first step is to download signal generation software. A version for windows
can be found here. I didn't search Freshmeat.net for a Linux version since I
didn't feel like compiling source code.
Next, procure a length of wire, and an old speaker or headset cable/wire.
The next step is to remove the speedometer from the instrument cluster.
Remove the cover from your computer. On my AT power supply, the yellow wires
going to power drives are 12 volts positive. Find a connector that is not
powering a drive, and make it accessible. Be sure to identify which connection
in the connector is 12 volts positive.
Connect the wires to your speedometer. If you look at the bottom of the card on
the back of the speedometer, on the side facing forward, the labels for the
connectors will be visible. The ground wire from the speaker cable goes to the -
connection, the + wires from the speaker go to the signal connection, and the
loose length of wire goes to the positive connection.
Put the connector for the speaker wire in the sound card's speaker output. Start
the signal generator, and turn up all of the volume controls. To test the
speedo, CAREFULLY touch the loose wire on the positive connection to your
computer's 12volt positive source. Depending upon which frequency is chosen, the
speedometer should display a specific speed. Test the speedo at 44.44444 (should
be 20mph,) 111.1111111 (should be 50mph) and 177.777777 (should be 80mph).
Remember the error amounts.
Take off the speedometer face plate, carefully removing the needle. Remove the
odometer assembly by removing the three Phillips screws. Now you can see the
workings of the speedometer. On the standard 85mph speedometer, 20mph is
straight to the left, 80mph is straight to the right, and 50mph is straight up.
Since the big weight is where the needle points, adjusting that will have little
effect on the 50mph calibration. Since the little weights are above each other
at 20 and 80mph, they have little effect on the calibration at those speeds.
There is a little tab near the upper right of the speedometer that determines
where zero is. If all of your speeds are a specific amount high or low, this tab
can be bent slightly to reset where zero is. For instance, if all speeds are
about 3mph high, the needle can be repositioned at -3mph for zero, and the tab
bent to bring it back to zero.
If the 20mph is low and the 80mph high, move the big weight inward. If 20mph is
high and 80mph low, move the big weight outward.
With the big weight at the top of the speedometer, you can adjust the 50mph
speed by shifting the little weights leftward if the speedometer reads too fast
at 50mph, and rightward if it reads to slow at 50mph.
After adjusting, retest. I had to test about 4 or five times, re-installing the
faceplate each time, then removing after the test. Did two speedometers in about
an hour, not counting removing the cluster from the car.
Hope this helps someone.
Addition Info/Post from the NATO Message Board, link
http://natomessageboard.com/Forum1/HTML/011100.html
How close were you able to get the speedos.
Best I could do on my home made 140 mph is +/- 2 mph over the range of 25 to 120
mph, with the 40 to 85 mph range being virtually exact.
Jeff Korn
------------------
I got the one that I ended up using to an error of less than 1mph, close to
zero, at almost all speeds, once the speedo warms up, compared to the trip
minder. Compared to the tach, there is a small error due to my tire/gear combo
requiring about a 16.3 tooth driven gear, and my actually using a 17 tooth since
that is what I had laying around and the local shops aren't likely to have a
16.3 tooth driven gear. I haven't tried the over 80mph calibration on the road,
so the higher speeds may be off more. The other one had the 50mph sliders maxed,
so I got it to within 1mph from 0 to 50 once warm, but then it got up to 3mph
error by 80mph, and to get the 80mph accurate, the 50mph would have to read low.
That one will be a spare, and I'll figure out some way of getting it more
accurate if I need it, since the local law enforcement folks frequent the 55mph
areas.
I'm glad this is useful, and I hope nobody fries anything. I was lucky - my
power supply resets when I do something stupid like accidentally short the 12V
10A + terminal, and my sound card took the abuse of the experimenting (such as
having a loose speaker ground wire touch the 12V 10A wire,) with no problems.
Mike Walsted - NATO member
------------------
I guess the next step would be to list the speeds that are straight up and
perfectly left and right on the speedometers other than the stock 85mph speedo
so we can list the frequencies needed to adjust the weights on those
speedometers when this gets added to the FAQ. I don't have the Canadian or
140mph speedos, so I don't know what speeds are in those positions. Of course,
anyone could figure the frequency themselves by multiplying those speeds by
2.222222222...
Glad this helps
Mike Walsted - NATO member
------------------
Looking at that overlay, it looks like
the numbers to use for calibrating a 140MPH speedo would be 24mph
(53.3333333Hz), 71mph (157.777777777Hz), and 118mph (262.22222222Hz). Or, for
the simplicity of having it right on a mark, 25mph (55.555555555Hz), 70mph
(155.5555555Hz), and 120mph (266.666666666Hz).
Mike Walsted - NATO member
------------------