Accuracy of 'new' tachometer

After watching Ben Goodwin's videos of testing the Roadster 2.5 on icy roads, I found myself wondering:

If Traction Control and/or the wheels spin for any reason, then how accurate is the tachometer?

I assume that the speedometer is driven by the front wheels, so spinning the rears will not make the speed shoot up to an inaccurate reading. However, if the speed remains accurate when spinning the wheels then the tachometer surely does not. You can hear the motor revs as they increase, but there's no real readout. I have no doubt that the PEM will protect the motor from exceeding its maximum safe rpm, but it seems somewhat awkward to have a tach which doesn't reflect the actual rpm under all conditions.

Any thoughts?

doug said:

Any reason this shouldn't be merged with the other thread?

Click to expand...

Personally, I find long threads which ramble around through several only-slightly-related questions to be very difficult to read. It gets worse when they're merged with other threads. I read the entire thread about the traditional tachometer and it had nothing to say about my question above. If someone new comes here with the same question as mine, I'd really hate for this thread to be merged with the other because then someone with a very simple question would have pages and pages of unrelated information to wade through.

P.S. You'll note that I contributed on-topic to the other tachometer thread before starting a new one, so I'm at least trying to keep this site organized and useful for old-timers and newcomers alike.

doug said:

I'm guessing what is displayed on that needle can be changed with firmware, though clearly speed is more important in almost all cases.

Click to expand...

Has anyone looked behind the instrument cluster to see whether there is a mechanical or electronic input?

My Honda has an instrument cluster which is almost purely electronic with the exception of the odometer and speedometer. The latter are driven by a mechanical speedometer cable that connects to one of the axles or hubs. There is a rotary encoder that converts this mechanical sensor into digital pulses for the ignition computer, but the original mechanical sensor is linked directly to the speedo - presumably for the sake of simplicity. I guess it also depends upon the cost for a mechanical speedometer (where you still have the option of a custom legend for rpms) versus a voltage-controlled gauge.

The kW gauge is almost surely firmware controlled. It seems only remotely possible that it's a literal ammeter or watt-meter, but that seems unlikely since the scale is not linear. The negative (green) range is not to the same scale as the positive (white) range, which would not be trivial with a pure analog circuit.

Anyway, I'm not planning on spinning my wheels any time soon. The tires will wear out quickly enough as it is.

cinergi said:

The best picture I have of the speedo/tach unit is below (it's the black unit sitting face-down on top of the steering wheel column). As far as I can tell and remember, there's no mechanical input.

Click to expand...

Thanks, Bill. Note: You might think about picking up a polarizing filter (linear, not circular), because then you could cancel out the reflection in the glass and get a nice photo of what's behind - assuming you have a camera that accepts lens filters and such.

scott451 said:

The Needles themselves are stepper motors.

Click to expand...

Wow, for some reason that seems like overkill, but I guess it takes less current than a typical electronic gauge. Thanks for the info.

TEG said:

Why 'stepper motors' instead of 'voice coils'? (Using hard drive terminology.)

Click to expand...

"voice coils" or even the most ancient gauges require constant current to maintain a reading (above zero), if I understand correctly. A stepper motor can draw zero current when it's not moving, although it seems that these needles are always moving and I doubt the savings in current would be that significant.

What I want to know is whether there is any gearing between the stepper motor and the needle. It seems like there would need to be a serious reduction to allow invisible steps.

Doug_G said:

Stepper motors are more accurate - as long as they don't slip, you're guaranteed they'll be exactly where you want them.

Click to expand...

Hard drive 'voice coils' are certainly accurate enough to repeatedly seek to one thousandth of an inch, otherwise there would be significant data loss. I suspect that they're way more expensive than a stepper motor for the degree of accuracy needed. Both technologies are certainly more accurate than old-style ammeter gauges.

TEG said:

I realized that 'voice coils' can 'cheat' a bit. There could be track marks on the platter so the heads could watch for a data pattern and make the voice coil seek in and adjust heuristically to the correct position.

Click to expand...

A stepper motor can employ similar marks, provided there is a servo system (optical, mechanical, electrical) employed. Many subwoofers have dual voice coils so that the second one can be used for feedback and fine tuning of position.

On the other hand, old VOMs (Volt/Ohm meters) tended to have a coil to direct the indicator needle and they seemed reasonably accurate.

Click to expand...

Old meters are not accurate enough for disk drive seeking, but they're accurate enough for eyeballing a changing signal. Actually, speed and power output are values that do not really need to be so precise. Speed is often overestimated, as mentioned, and power could easily be 10% off and nobody would ever know (don't we get 215 kW out of our Roadsters even though the meter stops at 200 kW?).

I think that the only real requirements for the instrument cluster is that the needle movement not be visibly stepped (at least it would really annoy me if the speedo were not smooth), and there should not be any noticeable lag and/or latency.