Modern Vespa

I am concerned that the digits on a digital tach will change too quickly to be useful vs a "conventional" analogue tach. where the graduations are quit close so the real minimum accuracy would be about 250 rpm

Mike, digital is always more accurate, but I take the point that it's easier to see the analogue type. Even the digitally driven analogue read-outs are not as accurate as the full digital screen type because of the introduction of a mechanical swing needle to indicate the revs. There is always some lag. We measured this in the workshop and found as much as 1/7th of a second! In engineering terms that's massive and means you can already be in the red zone for you engine before the tacho tells you. No good if it's a racing bike. But probably not important on a road bike.

Not necessarily... it depends on the nature of the circuit that drives the digital display.

I can see your point. I definitely prefer an analog tach for just that reason. On a speedometer, which doesn't change nearly as fast, I prefer a digital readout.

On another forum I frequent there is occasional near-hysteria about the readings of digital temperature gauges which show alarmingly big numbers, but nowhere near the same level of concern about a coolant temperature gauge which occasionally gets near the red.

For a critical measurement (e.g. speed where inattention can gain a speeding ticket) I'd rather have digital; for things where I'm looking for an indication of abnormality, I am happy with an analogue needle.

Being somewhat expert in throwing the proverbial spanner in a thread: I've long wondered about the relevance of a tach on/in any vehicle not possessing a manual transmission, besides possibly sprucing up an otherwise-dull instrument panel (e.g., Le Wife's Honda CR-V). As a reminder that the engine's running? Tell me what I'm missing here.

Last edited by amateriat on Sat, 01 Dec 2018 01:41:05 +0000; edited 1 time

I like an analog tach and a digital speedo.

Give me a traditional analogue display anytime. I've seen too many posts from members with failed LED's or the instrument display has suffered from sun damage.

I'm going to change the CVT weights so feel I should be aware of the difference rather than just by seat of pants.
Downside is that is might be more difficult to do the tach wiring than fit the weights.

But yes I see your point.

I don't think I'll ever get to the red line, given the traffic here!

On-line shopping sounds eay but as we are finding here most stuff "advertised" is not in stock.

The two tachs. $8.50 each i bought from a bricks and mortar motorcycle accessory shop LXV has proper handle bars so easy to mount clamped to the bars.

I was just going to ask what you'll use the tach for

For tuning purposes a digital tach with data transfer/recording capability would be super....but maybe a tad extreme just to test different weights.

I'll actually challenge the need of a tach also for the bikes with gears, if it's not for racing. If a new to me bike has a tach, I do use it. But only in the beginning to see where the peak torque figures are vs. speed & gears and also how this sounds. When I've learned this, I just play it by the ear and don't watch the tach.

I really like the set-up of my current bike: there's an electric tacho meter in a display, that can be switched to show other things too (like the remaining distance with fuel available). So it's there if I'll ever need it, but don't normally use it anymore.

The linear type of pseudo-analogue tacho is not much use either. The one on a KTM Duke was not readable when moving because it was small and jumped too quickly. You had to change gear using the engine noise as a guide but hitting the rev-limiter in bottom was all too easy.
The one on my present scooter is similar but arranged in a curve of progressively lit LEDs. Less important with CV transmission but it provides some information when cruising at steady speed. Some modern Triumph twins have no tacho whatsoever and this can cause difficulty if preparing to overtake or trying to extract the last ounce of performance during acceleration. I can't see much use for a fully numerical tacho except to check idle speed or cruising steady RPM.
If you want to change pulley weights I think a tacho is essential to avoid confusion about the actual effect.

You might think so Siver Streak, but we found that not to be the case. For use on cars and bikes, even very high revving bikes, all digitally driven types were far more accurate.

There are three principle types of digitally driven tacho's, leaving aside the old mechanically driven systems. Optically driven, electromagnetic, and electronic. Some combine elements of all three types, some elements of just two types. All have to take a reading from the crank sensor either electromagnetically, or optically. But anything that introduces a mechanical swing arm at the display to indicate the revs introduces quite some lag on a higher revving bike engine. On racing engines that I developed we used optical whenever possible. Easier and faster, and by far the most accurate. The engines revved out to 17,000rpm +. We displayed this in an analogue format but using an HD digital LCD screen. So no lag! It's the best of both worlds. When using digital with analogue readout, actual anologue readouts often could be out by 1k to 1.2k revs by which time the engine is in the red. Mechanical systems simply can't compete either. Typically they could be 2-3k revs out compared to even the most basic electronic digital system. There is not much built in engine strength redundancy due to lightness of components in most purpose built racing engines. Therefore avoiding the red zone is critical.

I'm not going to test different weights as such I don't have the facilities living in a company house. I'm just going 10% lighter and probably have the last set of 9g Dr. Pulley in Indonesia. $28

i am just curious what the difference will be. Not racing, unless you count my Karting instincts coming to the fore when negotiating slow moving traffic queues.

Max speed here only occasionally 80 kph. Most of time squirting between 40 and 60. there is a lot of slow moving traffic on narrow road interspersed with speed bumps.

That makes sense, thanks. I was thinking about this when considering (still considering, on account of the expense) Vespa's app-driven diagnostic module (the predecessor to the new thingie they have now, which is more attuned to TFT-equipped Vespas like the new Primaveras, Elecctrica, and hotrodded GTS300). That offered a tach as well as a host of other vital signs of the bike.

You said "always." That is what I take issue with.

Sure, the best digital tachs have virtually no latency, but I've seen plenty of cheapos with a very slow display refresh rate that even a mediocre analog tach will beat the pants off.

In reality though, for a scooter with a CVT, the only real utility of a tach is for diagnosing problems with the CVT and for tuning the CVT. For those purposes, just about anything will do. Overreving is not really a concern.

use which ever you want it makes no difference.

first install tach and get what ever readings your looking for at what ever speeds.

install your new items and go for another test ride. Compare your findings.

How do you know the digital refresh rate is slow? Did you actually test it against an analogue system? Bet not! Never seen any digital tacho of any kind that was slower than mechanical or digital/anologue. We tested every type available 5 years ago including many from China. All were amazingly accurate. That was a surprise to us too. Many had only a 50htz refresh rates, which is slow...but as we found, very effective. It's actually it turns out, not wholly the refresh rate that is the deciding factor. Aside from other electro mechanical issues, it's the actually type of engine it's being used on. The cheap digital China versions were great up to 10krpm, but above that were laggy but ahead of course of any mechanical or analogue readouts because they cannot cope with above 10k revs on fast engines especially when you get above 13krpm. So I do understand your reluctance to believe this. But as you say, on scooters such as ours it's not actually important. I was just pointing out a fact.

Analogue, though I have a feeling eventually we won't have the choice.

I appreciate that, but what kind of digital displays are you talking about? Mock "analog" displays using segments to emulate a sweeping needle?

When I am talking about display refresh rates on cheap digital tachs, I'm talking about the type that display actual digits. I have sitting on the bench in my shop right now a cheap Chinese digital tach of this type that I use to tune the CVTs on my scoots. It certainly appears to me to update its display only every 20ms or so. Perfectly adequate for this use, but slow for registering instantaneous over-revving.

Regardless, my point in principal is that I don't think your use of the term "always" is defensible. The accuracy of a measuring system is a function of many factors -- including calibration and resolution -- and when it comes to measuring a rapidly varying parameter, it is the overall time constant of the system from sensors to display that is critical, not just the display technology. When it comes to more slowly changing parameters, calibration and resolution are more critical.

Our brains seem to be coming at this from different points of view. You appear to be basing your determinations on empirical experience with specific commercial products. Coming from a metrology background, I'm basing mine on the general principals of measurement.

The only difference I have experienced was electrical (static, pulse. hf & vhf radio) interference with digital tachs.
I found them to be more accurate in the right environment, especially at high revs.

[quote="Silver Streak] our brains seem to be coming at this from different points of view. You appear to be basing your determinations on empirical experience with specific commercial products. Coming from a metrology background, I'm basing mine on the general principals of measurement.[/quote]

No, I've already covered what you are saying. As you say analogue covers different types of tachometer and their displays. Maybe I should have quantified those further at the start of my posts. Please bear in mind my statement in my first post was about accuracy which has two elemenst in this thread. Accuracy of information recorded and information displayed. Digital is always the most accurate. How it's displayed can be another matter.

Leaving aside the old mechanically driven analogue systems that are way too slow for modern engines, you are left with digital, or other electronically driven systems. The display can be of an analogue type with either a mechanical pointer needle, or LED/LCD digital readout, or an LCD Portrayal of an analogue display. The latter type was the fastest and most accurate at displaying information for everyday use on high revving engines. And of course, as you rightly say, digital tachos vary in latency as I previously pointed out in other posts. So of course you use the one that's right for the job.

Your tacho is clearly right for what you are using it for. But it clearly has latency issues for you, but will be near 100% accurate in the way it records it's information, even if it doesn't display all in totality in the blocks you would wish for. Thats because it uses a simple double cyclic algorithm on a cheap chip. It is very accurate for most road engines in lower powered bikes and better than analogue mechanical systems. Remember, accuracy and latency are in fact different issues as you will know. It's the information display that's important which is the point the OP is querying and that you are pointing out too.

I'm saying that all the digital tachos tested by us were near 100% accurate and far far better at displaying that information than plain old analogue tachos and far more accurate than any digitally drivenl mechanical analogue pointer system. Didn't mean to confuse you but I'm quite good at doing that when I'm tired!...lol.

I suspect we are killing this thread with competing pedantry, so I'm going to say one last thing and then shut up. We need to agree to disagree.

My point: There is no justification for your statement that digital tachs are ALWAYS more accurate than analog tachs. Your arguments are based on your empirical experience with a finite group of instruments that you have tested. It's fine to conclude that the digital tachs you tested were all better than the analog ones you tested, but you cannot logically extrapolate that conclusion to all tachs. It is certainly possible for a digital tach to be wildly inaccurate, and it is just as possible for an analog tach to be extremely accurate. It all depends on the quality of the design of the entire system and its calibration. True, it is easier to design a digital tach to be accurate, especially for rapidly varying revs where meter ballistics limit the dynamic response of an analog/mechanical system, but that doesn't mean that there aren't some digital designs that are worse.

I'm basing my viewpoint on my past 34-year career as a metrologist, culminating in 15 years as the chief of the research group at a major US government laboratory. My viewpoint is grounded in the basic principles of scientific measurement and logic.

OK, I've had my say. I'll stop beating this dead horse.