Speedo accuracy.

[Peter Laidler]

As has been sort-of explained here. Spider is dead right regarding the odometer - or distance traveled. NOT too important on our roads, but where it is VERY important, hence the need for calibrated odometers, is in the deserts of Oz and the middle east. That's because of the requirement to travel an accurately set DISTANCE as you travel are on a compass bearing. And the prismatic compasses are all calibrated too. Yep, we calibrated prismatic compasses too

So far, so good. Further reading not for the feint hearted

As for the speed reading..... If yours isn't accurate..... It is easy to check the accuracy of the speedo. Just drive at, say, 55mph and check your speedo reading against your GPS reader. They are VERY accurate simply because, keeping it simple, that's the nature of the beast.

But correcting your speedo.

As has been said here before several times. If you are a few, say 10 or so mph over or under, the brave of you can simply (?) lever the pointer off the tapered spindle shaft and replace it at the correct place. That is telling you VERY simply. There is more to it than that, believe me.

In some cases, where the clock-spring cannot absorb the variation you want, that option is closed.
Are you STILL awake at the back of the class.......

The clock-spring might not be able to absorb the required change because a) when it returns to zero or 0, there is insuffiicient spring remaining to return the pointer to zero. At the top end of the scale, in our case, where we might want to use a 130mph scale on an original 90 mph clock, the magnetic DRIVER drum might not have sufficient energy to drive the DRIVEN plate to overcome the clock spring tension.

In these cases, with our facilities we COULD increase the tension on the clock spring - but don't try this at home - BUT WE COULDN'T DECREASE THE CLOCK SPRING TENSION.

Are you still with me? In those cases, with the instrument set on the driven table we could gently move the magnetised DRIVEN plate closer to or further away from the driven plate (or vice verca, depending). We could come up with a dead accurate speedo and certify it.

The clock-spring is just a good compromise in that it pretty-well retains the same loading or tension throughout its range so in theory - and practice - it is good for the whole 340 degree speedo range. So if you are adjusting the readout by lifting and replacing the needle to get a more accurate reading, do it slightly over the half way range.

That's how we did it in an air conditioned, clean-air instrument workshop.

[Kiwi Craig]

Peter
The speedo in my 65 Cooper S only reads up to approx 50-55mph no matter how fast you go.
What is the problem there? Is it to do with the magnetic drive you mention?

[Peter Laidler]

I would say that in your case, we KNOW that the magnetic driver and driven plates are working, because they've rotated the pointer to show 55mph on the screen. So far, so good. That the pointer will not read any further suggests to me that something behind the screen is preventing the spring loaded driven wheel/plate from rotating any further. It won't take much to stop it, believe me

If it was mine, I'd take the speedo out, strip the bezel and glass off, unscrew the 2 BA screws from behind and release the mechanical drive train. Be very careful with it and slowly rotate the pointer to the 'stop' point and look at what the problem is. Gently blow across with an aerosol can type air blower.

[DUF2 ‘65 Elf]

Having looked at the Guesswork’s spreadsheet, it looks like I need to change the speedo head from 1376tpm to a one with 1248tpm. This should then bring the speedo in line with the 12” tyres and the 3.4:1 diff ratio.
I still need to change the speedo cable to stop the needle wandering. I realise this will be an awkward task with the engine in situ.
In the meantime I will keep my eyes open for a speedo with 1248 tpm.

[surfblue63]

https://www.ebay.co.uk/itm/275523136844?

https://www.ebay.co.uk/itm/195443158822?

https://www.ebay.co.uk/itm/225234071683?

[DUF2 ‘65 Elf]

I have looked at Guesswork’s spreadsheet and it would appear that I need to find a speedo with 1248 tpm.
I was chatting to a mate, perhaps overthinking things and the thought came up, if the existing speedo 1376 tpm is reading slow, surely substituting a speedo with LESS tpm would make the speedo read even slower?

[surfblue63]

If the tpm is less, then it needs less turns to register a mile.

Say you have a speedo with a tpm of 1376, the wheels need to turn 1376 times for it to register 1 mile.

A speedo with a tpm of 1248 would register 1.10 miles for the same 1376 turns. (1376/1248=1.10)

Now if you convert to time. If your 1376 speedo did that mile in one minute, then the speedo would indicate 60mph.

For the same 1376 turns in one minute the 1248 tpm speedo would indicate approximately 66mph. (1.1/1x60=66)

[DUF2 ‘65 Elf]

If the tpm is less, then it needs less turns to register a mile.

Say you have a speedo with a tpm of 1376, the wheels need to turn 1376 times for it to register 1 mile.

A speedo with a tpm of 1248 would register 1.10 miles for the same 1376 turns. (1376/1248=1.10)

Now if you convert to time. If your 1376 speedo did that mile in one minute, then the speedo would indicate 60mph.

For the same 1376 turns in one minute the 1248 tpm speedo would indicate approximately 66mph. (1.1/1x60=66)

That is great, I did think maybe.I was going the wrong way. thanks for clearing that up for me. So now I just need to keep my eyes open for a 1248 speedo.

[DUF2 ‘65 Elf]

I am still struggling in my mind with the TPI.
I have now got a 1248 tpi speedo, when I tested it the cable input was not turning. The red odometer gear shaft was stuck. I have removed this and freed it off. So the input is now free and turning.
But then I started thinking (yes I know I shouldn’t!) if a speedo has 1248tpi it must turn at a certain rpm to indicate say 60 mph. So if a speedo has 1376tpi, surely the cable is turning faster at 60 mph that a 1248tpi? So surely more TPI means the cable is turning quicker at 1376 that 1248?

[surfblue63]

The speedo cable will turn at the same rate regardless of what speedo you have. The speedo will indicate a different distance because of the amount of turns required per mile (TPM).

As I showed before the 1248 speedo will give a 10% increase in indicated speed over the 1376 speedo if nothing else is changed (wheel size, diff ratio, speedo drive gears).

[DUF2 ‘65 Elf]

Thanks Surf blue. I thought that was only relating to the odometer gearing.
Obviously there is no direct connection from the speedo cable to the needle, as it relies on magnetism pulling the drum around.
I just can’t see how the speedo needle/ drum works when (imho) the cable will has less RPM on 1248tpi as opposed to 1376.
I have looked at a chart listing all BMC speedos, hoping to see some sort of correlation between an 850 and a Cooper S but it does not make things easy for me!!

[timmy201]

The speedo tpm is critical to get the odometer working correctly

As you say the speed reading is related to the magnetic pickup and this can be tweaked independently of the odometer

I found this out when the speedo shop calibrated the speed on my kph speedo in mph, so doing 100kph it was showing 60kph. The odometer still read perfectly in kph

You really need to know your diff ratio and speedo drive gears in the gearbox to get it reading correctly

[DUF2 ‘65 Elf]

My original post gives details of my car. I have also used the Guess works calculator and 1248tpm seems to be the correct one.
My problem is I cannot get my head around the 1248 would appear to be correct, but the 1376 speedo would appear to need the speedo cable rpm to be higher on the 1376, so how come the speedo will read correctly (maybe!) at lower rpm?
I may be completely wrong about this and perhaps over thinking it!
The speedo currently reads slow by about 10% showing 38ish whilst GPS shows the car doing 43. I just hope the replacement speedo reads closer to the actual speed.
The other thing is it is a bit of a pain to get the speedo out on an Elf!

[1071 S]

TPM figures are pretty irrelevant to speedo (note..not odo) calibration. My S started with a 3.76 before graduating to a 3.4 and eventually 3.1 diff all with the same speedo.

Each time it was calibrated by a local specialist. I believe they do it by changing the spacing between the drive and driven magnetic plates. At one stage the speedo was a bit wobbly so a mate (an ex-instrument fitter) fixed it by making a special tool (a sort of stiletto blade from an old hacksaw) and wiping out the collected dust and whatever from between the plates.

A long time ago during a boring trip across the Hay Plains my co-pilot (mates young son) and I amused ourselves by recording the odo reading (miles) against a long string of km posts. After feeding the data into a spreadsheet the answer to life, the universe and everything turned out to be roughly 20%. So, if I ever need to know how far I’ve been, its just a case of reading the odo and adding 20%. Probably not good enough to win an international rally but good enough for sailing ships (as we used to say in my old profession).

Cheers, Ian

[DUF2 ‘65 Elf]

I am getting more & more confused by this speedo situation.
I have the 1376tpm speedo still fitted to the car, I have the 1248tpm on the bench, (I have freed off the seized odometer gear & shaft)
I have run both speedos off my battery powered drill.
At max revs in reverse on the drill, the 1376tpm reads 55mph.
The 1248tpm reads 55mph
So the speedo reads the same on both instruments!! WTF.

So changing the speedo head with differing tpm, may affect the odometer reading, it does NOT change the reading on the speedometer.

[timmy201]

A drill on full speed isn’t a calibrated test rig for a speedo. Unless you know the exact rpm of the drill at that point you might have another factor at play like one speedo is a bit gummed up

A gauge repair shop can measure them accurately

[DUF2 ‘65 Elf]

Thanks Timmy. I know my drill,is not “calibrated” I just wanted to see what difference there would be between the 2 speedos. I was surprised to see the speed was the same on both clocks

[Tornado99]

I have a similar issue...MK3 car with non original engine. Speedo reads about 15% lower than actual speedo (via GPS). No idea what diff ratio nor what cable drive gear is in there so I can't really use the GreenWorks web tool.

Thinking a sure fire way to solve this is to replace speedo with the electronic type that takes input off a Hall effect pickup off a drive shaft etc or uses a GPS input signal. Are the e-Speedo units on MiniSpares any good?

[1071 S]

Or.... as explained several times above.... have the speedo calibrated by an instrument specialist. This is not a time or $$ expensive task and does not involve tampering with the speedo drive gears.

The odometer of course is mechanically driven so does require the correct drive gears or some other solution (intermediate gears) to read correctly.

The speedo calibration (not odometer) is determined not by the revs of the speed cable but by the space between the magnetic plates.

Once the instrument is calibrated, then the speed reading will rise and fall in accordance with the speed of rotation of the driven plate ...which is driven directly by the speedo cable and the associated gears.....

There are lots of alternative electronic speed measuring systems. A mate recently installed a GPS driven unit with a small HUD that projects a digital readout onto the bottom of the windscreen just above the speedo. Quite unobtrusive in his 50s Swallow.

Cheers, Ian

[Spider]

I am getting more & more confused by this speedo situation.
I have the 1376tpm speedo still fitted to the car, I have the 1248tpm on the bench, (I have freed off the seized odometer gear & shaft)
I have run both speedos off my battery powered drill.
At max revs in reverse on the drill, the 1376tpm reads 55mph.
The 1248tpm reads 55mph
So the speedo reads the same on both instruments!! WTF.

So changing the speedo head with differing tpm, may affect the odometer reading, it does NOT change the reading on the speedometer.

Either one or both of these speedos have calibration errors or the face plate with the TPM for one (or both) numbers do not match the speedo head.

This is where you need to first check the odometer to actually see what the speedo is. Check everything, assume nothing.