GT tacho / rev counter -- how accurate are they?

[Craig380]

I know every instrument will be a little different, and 25+ years of usage isn't gonna help any, but does anyone know roughly how accurate a GT tacho is?

I have a period test from the UK mag Motorcycle Mechanics, were they reported after a dyno test (33.4bhp @ 8200rpm at the wheel) that the tacho was under-reading by 10%.

Are they THAT bad?

[Suzukidave]

From what i have seen even on modern bikes that have a cable drive the speedo and tach can be all over the place wrong .

[pjmcburney]

I know every instrument will be a little different, and 25+ years of usage isn't gonna help any, but does anyone know roughly how accurate a GT tacho is?

I have a period test from the UK mag Motorcycle Mechanics, were they reported after a dyno test (33.4bhp @ 8200rpm at the wheel) that the tacho was under-reading by 10%.

Are they THAT bad?

Hi Craig,


Well... they're probably not very accurate at all after 35 or more years, but they can be made so again quite easily.

I had my gauges refurbished by a very reputable business here in Brisbane, and I was privileged enough to observe the process because I had accuracy 'issues' with my speedo.

For those not in the know...
The mechanism used in GT gauges is essentially the same as any modern cable driven gauge, so they can be calibrated quite accurately - and mine are.

The concept works on the eddy-current principle. The cable from the front wheel/engine spins a small cross-shaped magnet underneath an aluminium 'cup' which is then connected to the gauge needle. There is no direct mechanical connection between the cable and the needle - it's all done with magnets!

As the magnet spins, it sets up an electro-magnetic field which 'coerces' the cup to move. The faster the magnet spins, the larger the field and the more force put on the cup around it.
The cup/needle arrangement is sprung and damped, so the needle returns to zero (duh) and so it doesn't move too quickly or slowly - you want it to show an 'average' of the field affecting it, not an instantaneous representation.

The size of the air-gap and shape and size of the aluminium shell determines the calibration of the gauge. Make the air-gap larger (by cutting and/or reshaping the shell) and the needle won't deflect as much compared to the number of revs of the magnet. Move it closer and vice-versa.

The gauge restorer I watched used a calibrated motor to drive the speedo mechanism at a certain number of revs and then shaped the aluminium cup carefully until the gauge read right.
The calibration factor is printed on the face of the gauges, I think at 1600 (cable) revs the speedo is supposed to read 60 km/h (don't quote me on that), so it's a matter of trial and error to get it just right.

I've checked everything with a GPS since the resto and the speedo is astonishingly accurate - perhaps a little too accurate. When I'm doing 100 km/h, I'm reading 100 km/h.

Ideally you want the speedo to read a fraction fast i.e. you want to be traveling at say 58 MPH while your speedo reads 60 MPH. This gives you a tiny bit of leeway when the fuzz pull you over for speeding and gives you peace of mind that you're obeying the speed limit (at least when you're speedo tells you you are )

Take your gauges to a reputable repairer and you should be right.



Cheers
Paul

[Suzukidave]

Paul even after they have been set correctly cannot the speed reading still be thrown off by a simple tire change with a different profile tire ?

[pjmcburney]

Paul even after they have been set correctly cannot the speed reading still be thrown off by a simple tire change with a different profile tire ?

Dave,


Absolutely...

I was probably a bit glib with my use of the word 'astonishingly' when describing the accuracy of my gauges.

What I was really trying to get across is the fact that a GT gauge is pretty accurate (within itself) when setup correctly, even compared to more modern purely electronic gauges.

My explanation doesn't take into account other factors outside the calibration process, such as wheel and tyre combos which can throw out what you read on a speedo considerably.
Heck, two same-sized tyres from different manufacturers can be different enough to make a noticeable difference to what the speedo displays.

But, the gauge itself is still receiving and interpreting the information put into it pretty bloody well.


Cheers
Paul

[Suzukidave]

Thats great to know that they can be calibrated

[Craig380]

cheers guys

[Suzsmokeyallan]

Sorry i have to totally disagree with this part about the cup air gap adjustment. If the gauge is of the type using a magnet driven shaft unit moving the upper bell unit connected to the needle, the calibration is made by adjusting the return springs tension and the needles internal rest point.
Increasing the springs tension retards the needles movement against the magnetic field and slows the registered speed, while if you reduce its tension this increases its speed reading.
The needles rest stop inside the gauge can also be used to adjust the overall tension point, or in other words the initial range of tension to the needles movement.
Moving the airgap is not reasonable since the space is very limited already and no gauges ive ever worked on or calibrated of this style have ever shown any manipulation of the bell in anyways by the factory.
Ive worked on quite a few GT750 gauges and found them to be off in as much as 20 mph sometimes in either way.
These units i recalibrated were checked using a referenced set shaft rpm speed as well which is the correct way to calibrate an instrument.
The needle shaft damper chamber does nothing to effect the speed setting, however its primary purpose is to reduce or remove any slight wavering of the needle in its range of motion due to road conditions /vibrations and also to make its motion appear visually slowed and 'fluid'.
Having checked into this extensively with the gauges ive restored, the GT750 gauges of 72 models have shown the most errors compared to the 73 to 77 models.

[pjmcburney]

Hey Allan,


It seems there's more than one way to skin a cat...

I can certainly see merit in the method you use, but how can you be sure that the gauge will react, or read consistently throughout its entire range?

What I mean is, can you be absolutely sure that spring tension is uniform on the needle from it's rest point to it's full sweep point?

Using a mechanical watch as an example, the tension or torque provided by the wound mainspring decreases as it unwinds. It's not uncommon for mechanical watches to lose accuracy as they run and it's customary to keep them fairly much fully wound to keep them as accurate as possible.

Now, I'm just shootin' the breeze here, but wouldn't the tension on the needle increase as it moves from rest to full sweep, which (to my addled brain anyway) would mean that an exponential amount of force be required to move it?

Essentially, what I'm getting at is, if you vary the needle spring tension to recalibrate the gauge (at it's calibration point), aren't you running the risk of requiring differing levels of torque at different points on the gauge than what was set from the factory - 40MPH might be perfect, but the '80MPH point' is now in a different place on the face of the gauge.

I know this is getting a bit pedantic, but anyone else have thoughts on this?


Cheers
Paul

[Suzsmokeyallan]

Paul as you stated they measure the unit at a speed thats close to the required speed limits in most countries and then factor in a 'slight over' percentage so youre guaranteed to be legally under the indicated speed.
The spring regardless of how its tensioned will have a stronger reaction at higher tensions than lower ones, thats a given nature of any spring.
So if your unit with an untouched spring is accurate, and i apply a bit more tension to the spring setting, the gauge will still read the same way as before needle sweep wise, just that it will now be a percentage slower in registered speed throughout the whole range.
For the record, a movement on the springs angle of tension relating to about 0.20mm of deflection at the mount will alter the speed quite significantly (around 10 km/h) either way, so as you can see from this example the reading is VERY sensitive to the springs tension.
What they concentrate on in the factory is how accurate it is at the most legally required speeds which usually fall between 60 to 100 km/h.
In mass manufacturing on any items time is crucial, so its much easier to have a rest point and spring mount thats adjustable on the gauge to set it quickly and get them onto the next stage of manufacture/assembly.
From the large amount of gauges ive dealt with, i have seen many of them of the same model with the spring mount and rest stop locations slightly altered from their stock positions.
This tells me each gauge and its spring/rest stop has been adjusted in the factory to compensate for a speed reading at a set RPM on that particular model of gauge.

[H2RICK]

And, of course, worn/dry bearings on the output/indicating-needle shaft will affect everything too.....

Interesting discussion, gents. Thanks for the insights.

[ja-moo]

Spend the 20.00 and get a Sigma bicycle speedo. If you measure the tire correctly you get within .01% accuracy (I believe) That will show just how accurate your speedo is through the range. Don't be surprised if you lose 10 or 15 mph off your top speed........

[H2RICK]

Don't be surprised if you lose 10 or 15 mpoh off your top speed........ Shocked

John, I wouldn't think that those electronic bicycle speedos would weigh THAT much.......

Ahhhh......some nights I just crack myself up......

[ja-moo]

That was good Rick.......

[Suzsmokeyallan]

Rick you gotta stop that, however imagine if you put on two Sigmas for dual reference, youd lose even more top speed.