h a l f b a k e r yWe have a low common denominator: 2
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
*Most* people spend the lion's share of their time trying to keep their speed to (reasonable) city or highway speeds. Lets say 0-100 km/h or so. However, many cars these days have speedos that go into the mid-200s, and so the usuable portion of the speedo display is crammed into the lower left, and
hard to see.
The logarithmic speedo would gradually reduce the distance between the 10 km/h groups as the speed increases. (i.e. the difference between the 10-20 km/h lines is more than the spacing between 110-120, which is more than 200-210, etc.
No more squinting to see when you're going too fast in the town while in your sports car.
[link]
|
| |
This is a good idea. I'd really like an electronic display where I could dial the amount (range) of logarithmic-ness in. Or the range immediately surrounding the speed you are traveling is dilated, expanded. And I don't mean a numerical dispay--I hate those--but an electronic presentation of a analog style speedo but with your modification. |
|
| |
This sounds like a bathing suit that gets geometrically tighter and tighter the closer the fabric gets to your butt and genitals. |
|
| |
Kinky, a car in speedos, but you shouldn't generalize, most cars don't wear them, not even in the garage. |
|
| |
I think the log scale would give you a false sense of slowness. 200mph would still be fairly close to 50mph. I like the system on some tool bars on the Mac OS where a simulated magnifying glass enlarges what you point at. |
|
| |
Not sure I understand that, [reensure]; when I had an overdrive-equipped Triumph (two actually), the manual said not to disengage OD over 80mph in 3rd, or 95mph in 4th. Engaging was pretty much OK at any speed as it only operated on 3rd and 4th.
I may be misinterpreting what you're saying. |
|
| |
There is a problem with this. The requirement (certainly in the UK) for speedometer accuracy is surprisingly low (+-10% - same in Oz too, I googled). This means that clearly, if you are going over 100mph, your speedo could be as much as 10Mph out. If you made the scael logarithmic, they would be even more innacurate at high speeds, with the bars being close together. You need to improve speedo accuracy significantly to make this work.
As an aside, it is clear that some speedos are far more accurate than others. I've tested mine on occasion by timing a mile at 60mph (the easiest way to do it as the maths is then really easy). The Smiths speedo in my old MG was very nearly spot on, my old porshe was accurate at nearly twice that speed (don't ask...) but my Saab is out by my measurement almost precisly 10%, and my peugeot nearly as bad (about 7.5%). I have also read that BMWs tend to always overestimate speed, always by as much as the law will allow, to make people think their cars are faster than they are, but this is pure hearsay. |
|
| |
UK Fords did that as well, back in the day. One of my Triumphs referred to above under-reported the speed as it had the wrong differential fitted (the one from the automatic version). The speedo was 12.5% out, but I got so I could subconsciously translate to real MPH. Improved the fuel consumption too. |
|
| |
As [goff] said, the variability of the car's speedo would make this much less accurate than you'd want. Using GPS on a nice straight bit of road, I check mine regularly. The Saab is only 1-2% out usually across the range and the Pugeot 7% at low speeds and 8-9% at higher ones. Top spec cars, as most high performance ones are often heve navigation systems so the two could be linked with an 'old fashioned' speedo for those occaisions when you don't have a GPS signal. |
|
| |
I believe that some traffic devices such as road angel also use GPS to display your current speed... should be accurate - and yes in UK the law states that a speedo can be out by 10% plus 2 mph - i.e. 35 in a 30, 57 in a 50, or 101 in a 90. |
|
| |
When we design vehicle instrumentation, there is a deliberate intent to ensure that the spedometer reads more than it should. EEC and Federal legislation forces this because of an accuracy requirement (as distinct from tolerance) of V(act) +0/-10%. We therefore ensure that we meet this by aiming for 5-7.5% over actual speed with a tolerance of +/-3% as the fines / litigation costs for going outside homolated standards in a production vehicle are too great.
[TwoSheds] Speedometers must by law indicate the max velocity that the vehicle is capable of achieving. |
|
| |
As an aside, Citroen used a sort of rotating cylinder as a speedo that showed a speed range within a viewing window which was novel but I guess in the end was little better than a 7-segment display. |
|
| |
Another way is to fix the indicator needle at, say, the 12 o'clock position and make the dial move. This way the dial could be huge and show the speed in generous 1 or 2 m/kph increments but only show a small subset arc of the dial in the display window. |
|
| |
But "most people" (that famous test group) look at the direction the needle is pointing, not the number that it's pointing to, which is why digital speedos take longer to read. |
|
| |
I'm not convinced by Goff's argument: |
|
| |
(Goff says) "The requirement ... for speedometer accuracy is surprisingly low (+-10% ...). This means that clearly, if you are going over 100mph, your speedo could be as much as 10Mph out. If you made the scael logarithmic, they would be even more innacurate at high speeds, with the bars being close together." |
|
| |
Surely this is a benefit of a logarithmic scale, in that the reading error is a proportion of the reported value. |
|
| |
For a directly proportional scale the proportion of error will be greater the closer it is to zero, which is presumably why there is an additional 2mph error permitted as AJCrowley mentions. |
|
| |
Gotcha. In my day (aah!), the OD was selected manually via a toggle-switch on the dash or, as in the case of my Triumphs, a little slide switch in the gear-lever knob. |
|
| |
No, because as they are now, with some experience, one can note the general position of the needle and instantly know how fast he is going. With that system, one would have to concentrate to find the corresponding number; the irregular spacing will confuse. |
|
| |
Interesting idea. + (WTAGIPBAN) |
|
| |
My ’79 car displayed 120mph. I’m not sure about the early ‘80’s displays. |
|
| |
On an aside, I'd like a tach that indicates peak power when the needle is vertically up and markers for peak torque and peak economy. |
|
| |
Hmm, isn't peak economy the lowest speed you can comfortably do in the highest gear? |
|
| |
Anyhow, just get a bigger dial, or one that goes almost 360degrees for full speed. |
|
| |
Surely something could be done with an
electronic display - dials keep more or less in the same positions, but shuffle about getting larger or smaller as they become more important perhaps? |
|
| |
Maybe something watching where your eyes are pointing to decide which dial to enlarge? |
|
| |
Or something that talks to you... |
|
| |
I like the idea as stated. Sure I don't use the low speeds much, but it really annoys me that if I'm going less than about 10 mph, my speedometer is completely worthless. |
|
| |
On thought I had was that this might make people feel like their car really looses acceleration at higher speeds. If one were to accelerate at a constant speed, a logarithmic speedometer would move fast at first, then slow down. Since a car's ablity to accelerate decreases as speed rises, this is exagerated even more. So maybe this would subcontiously make drivers start accelerating slowly so they could increase the rate of acceleration some as they sped up. Or maybe it would annoy them as the car's ability to move the speedometer needle seems to just drop off at higher speeds. |
|
| |
Re: [Toffee] "Speedometers must by law indicate the max velocity that the vehicle is capable of achieving." |
|
| |
Ha! Not in my 1986 Buick Century. The speedo needle pivot point was behind a bar of the gauge bezel. In two separate narrow rectangular windows: above was the speedometer (which accurately indicated 0 to 85 mph over 170 degrees of rotation, 0.5 mph/degree) + odometer, below was the trip odometer. I believe that that mid-sized sedan with its 3.8L multi-port fuel injected engine and overdrive transmission was capable of 125+ mph. Holding the accelerator to the floor produced the somewhat comical result that the needle would travel past 85, then *disappear* behind the plastic, then *reappear* pointing at the trip odometer below - a total of 270 degrees of travel which, if its response remained linear, equates to 135 mph. Since mechanical speedometers are affected to some degree by the physics of springs, they typically move (slightly) fewer degrees per mph as speed increases, so my actual speed was probably even higher. |
|
| |
I'd like to see a display in which the actual position of the numbers remains unchanged, but a region around the pointer is magnified (much like what is seen around the mouse pointer in OSX, for you mac fans). This may only be feasible on fully electronic displays (replace the entire cluster with a high-res flat-panel display), but it would be sweet. |
|
| |