|
Speedometer
calibration
|
|
|
Adam S-J
|
|
|
|
| If you are not going to use the speedometer and
the drive train from your donor car, or if you have changed your
differential, you will need to check that the speedometer reads
accurately before you take your car for its SVA test. This is not
difficult - no need to check gearbox or differential ratios. |
| 1) You need to know how far your car moves for
one revolution of the engine, and since this will vary from gear to
gear it makes most sense to carry out the check using 4th gear, which
is usually a direct (1:1) drive. However, the method will work even if
4th is not direct. (But when you come to parts 8 to 10 you must use the
same gear, and this isn't going to be feasible if you do the
measurements in 1st or 2nd, for example, much less in reverse!) |
| 2) So: put the car on level ground and make a
mark on the ground directly by the centre of one of the rear wheels. |
| 3) Rotate the crankshaft pulley until the
timing notch lines up with the Top Dead Centre mark. |
| 4) Put the car in 4th gear (and take the brake
off). Push the car (it doesn't matter forwards or backwards) and count
exactly four revolutions of the pulley. (You can do more if you want,
and this will give you a greater degree of accuracy - but 4 should be
enough) |
| 5) Measure carefully the distance the rear
wheel has travelled. Since we are working in MPH, I suggest you measure
in inches. The distance ought to be in the region of 70 to 80 inches,
depending on your wheel diameter and final drive ratio. |
| 6) Work out the following sum: distance in
inches divided by 4 (to give distance in inches per one engine
revolution) divided by 36 (to convert to yards) divided by 1760 (to
give miles per engine revolution). Then multiply this figure by 1500,
then by 2000 etc up to your engine rev limit and note each answer: the
figures you get should be an accurate reading of how fast your car is
going for those Tacho readings. If you don't feel like doing all of
this, you can download a spreadsheet (MS Excel
or Lotus 123) to do the work for you. |
| 7) Jack up the back of the car so that the
rear wheels are clear of the ground. Use axle stands. Make sure the car
can't move!! |
| 8) Connect up an accurate tacho (or use the
dashboard one if you are sure it is accurate), start up and gently
engage 4th gear. Don't cough, or you might go through the back of the
garage. (Seriously, I have done this and it is not really dangerous) |
| 9) Make a note of the speedo reading at 500rpm
intervals from 1500rpm. |
| 10) Either consult the spreadsheet or else get
out your calculator and work out how fast you ought to be travelling
for each 500rpm interval. If your speedo is under-estimating your true
speed (e.g it says 25mph at 2000rpm but you have calculated that you
are doing 30), it won't pass the SVA. If you are travelling no more
than about 9mph slower on your speedo than you calculated, then you
ought to be OK. (The exact error margin is 10% of the true speed plus
6.25 mph, i.e. at a true speed of 30 your speedo may read anything
between 30 (NO less!) and 3 (10% of 30) plus 6.25, that is 9.25, giving
39.25 mph). The true speed figure you get should be accurate so long as your tachometer is accurate. To set up for SVA I would recommend you aim for half way between the lower and upper limits allowed. If you have the space to run your car (such as an airfield) you can cross-check by measuring a distance (which will need to be considerable, in hundreds of yards) and timing the car over this distance at a set speedo reading. You can then work out your actual speed with about the same degree of accuracy. Note - depending on your wheel diameter (the larger the wheel the greater the error) with a diameter of around 71 inches a tacho reading error of 250rpm will affect the true speed reading by about 4mph, an error of an inch in measuring the distance in part 5) above will affect the reading by about .5 mph at 30 and .9mph at 70 - it should be possible to get both readings well within these limits. |
| If your setup does not read within the limits, you can try a different speedo. If you get one from a car which has the same overall rear axle ratio as your Locost then it ought to be OK. Or, you can splash out on a programmable speedo (around £140). Or, you can get your speedo recalibrated (see Links page for a company called Speedograph Richfield). This will cost at least £50. Or, you can try tweaking the speedo you have - one idea is to take the needle off and move it in relation to its shaft, another would be to move the dial relative to the speedo. Be careful - speedometers are pretty fragile inside. |
| Footnote: my car was SVA tested on 11th April 2001 at Glasgow, and the speedometer (ex-Triumph Dolomite, re-calibrated in the way described above) passed the test, reading consistently 3mph faster than actual speed up to 70 mph. |