h a l f b a k e r y"Put it on a plate, son. You'll enjoy it more."
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,
|
|
|
This product will not appeal (or make sense) to the majority of drivers. It will also disgust many car fundamentalists**. It is for a specific new-to-stick driver, and if you find yourself disliking it because it is not "old school", well, consider that you were learning once too.
Attach
(via OBD or other methods) a microcontroller to the following data in a stick shift car:
1) Engine RPM
2) Road MPH
3) Clutch engagement (in or out)
4) Current Gear selection (note: This is not in most OBD, and is the reason i have not already invented this product. This will have to be done with some prox switches (or similar) to detect gear position.)
Using the attached data, a few LEDs will indicate, while the clutch is depressed, whether or not your engine RPM is appropriate for the road speed given your current gear.
I'm thinking a Green for +/- 40rpm, a yellow on the right of the green for "engine rpm is 40-100rpm too high", an orange to the left of the green for "engine is 40 to 80 too low, and reds at the extremes for "don't let out the clutch now!"
This would be mounted around the outside of the tach (or even be integrated into cars tachs, say, a colored background?) so as to not require taking ones eyes off the road to interpret.
Since all car gear ratios are different, the microcontroller will have a bit of non volatile ram for gear ratio storage. It will "learn" gear ratios by checking MPH and engine RPM when the clutch is up.
So someone will have to drive the car in each gear for a second before this will work. Or it can come preprogrammed from the factory for your car for an extra $2.
My gf, who really should learn to drive my car, along with other new-to-stick people, could really use this to learn the basics of when to let off the clutch properly. Think about the benefits to seriously fast downshifting!
The advanced model will have a predictive algorithm to help with speed shifting. When you're in neutral with the clutch up, it will show you when it is okay to shift into the next highest gear from the one you were just in.
** If you hate the idea and love your car so much, go teach your spouse to drive your car without this. ha!
[link]
|
| |
I think my "Shift" indicator light works exactly this way and did help me teach my 19 year old son how to drive it. So... baked. |
|
| |
On board diagnostics, methinks...? |
|
| |
Actually, I read the title as "Red shift
indicator" and assumed it was a very
sophisticated speedometer. |
|
| |
[globaltourniquet] - this isn't a shift indicator. It's intended to make sure engine RPM is appropriate for the gear into which the transmission is being shifted. I think [ericscottf] is trying to avoid the lurching frequently experienced when someone is learning to drive a manual transmission. |
|
| |
I like it (+) and I don't think it's baked as shift indicators just sense tach signals. I would appreciate this on every different car I drive as it takes me some time to hear the engine rev and equate this to shift points. |
|
| |
So the mechanism actually physically prevents engagement at anticipated excessive RPM levels? |
|
| |
Hmmm.... I'm thinking of a word that starts with "M" and ends with "agic"... |
|
| |
OK, no nothing in the idea prevents engagement - it just says the light tells you not to let the clutch off because we will rev too high. |
|
| |
But wait - if we can turn a light red, can't we block the gear engagement? |
|
| |
You know how at certain points of the RPM scale, depending on what gear you are in, you can shift without using the clutch if you match the RPMs for that specific gear? |
|
| |
I think this light indicates these points. |
|
| |
It would help to learn heel/toe driving as well. [+] |
|
| |
I think this also could be acheived by wrting 1,2,3,4,5 etc on the tach and spedometer with a sharpie at the right points. |
|
| |
I would suggest having a tach with two needles on the same shaft (the way a clock does). The front needle would be shorter and would be light along its whole length. The rear needle would be longer, but the center part would be dark so as to blend in with the background. |
|
| |
One needle would show the speed on the engine side of the clutch; the other needle would show the speed on the gearbox side. For a smooth shift, both needles should be lined up. |
|
| |
Provided this is COMPLETELY PASSIVE it is a fine idea. Have no fear, [ericscottf]: this is coming from the most fundamentalist of vernacular-technology fundamentalists, hence the username. And, there is no need to bother with OBD (what [MaxwellBuchanan] would have said if his first anno were ironic): all this can be a done with simple soldering-iron-and-pc-board add-on gadgets. |
|
| |
(Related aside: Many years ago, I was a bus driver. The buses had semi-automatic transmissions whereby an epicyclic geartrain (like in automatics) was used with a manual selector. The driver needed to match the engine speed with the road speed when changing gear, just as though a 'crash' (non-synchromesh) box was fitted, so he would double-declutch only with no clutch. There was no issue of 'missing the gear' if the engine revs were wrong, just an unpleasant lurch as the new ratio was selected. One of the vehicles that the company looked into buying was fitted with a device called a prop-shaft monitor which prevented engagement unless the revs were correct. This was ~25 years ago.) |
|
| |
Hey folks
I'm so glad to see this idea so well received.
Global: This is not magic - It does not prevent gear engagement (though that could be done without magic), all it does is show you if your engine speed matches the road speed for the gear you're in (or it thinks you're about to be in, in the deluxe model)
Those of you with saturn cars + stick have a little orangy light that lights when you are supposed to upshift. That light is NOT this product - that light merely shows you when your engine RPMs are uneconomically high (For gas and wear on parts). This product says "hey, moron! if you let out the clutch now, you're gonna glaze it or lug the engine!"
UnaBubba: This product will not prevent/suggest against engine braking - recall that engine braking occurs while your clutch is fully engaged (unless you like repairing clutches) - This product wouldn't even be displaying any data during an "engine braking" process.
It is indeed a completely passive product. It will not prevent you from driving poorly, it will simply provide you with the visual indicators to drive stick properly.
Giblet: Those "shift" points change depending on how fast your car is going (mph)!. Please do not write on your tach. It will only be correct in a very narrow situation!
Supercat: I originally envisioned your method but abandoned it for the sake of ease of reading with peripheral vision.
Ned: I know it can be done by attaching a tach to the spark plug, and to the speed cable for road speed - but these data are already in OBD, so why not use it?
Anyone have any suggestions on how to pick up gear position? Like i said, reliable gear detection is the only reason i haven't baked this already. |
|
| |
No, I think [supercat] has it right. I think the easiest way to do this would be to have two coded-wheel (or other) frequency sensing devices, one on the engine output shaft, one on the gearbox input. Ie one either side of the clutch. |
|
| |
Have two colinear tachometers in the one unit, and use the output of these sensors to drive the two needles on the tach. You could have lights or some other very bright and clear indication on the inner needle on the tach (that is hidden by the outer needle when the clutch is meshed) - so when engine speed does not match gearbox input shaft speed, you get a very clear indication. |
|
| |
With a small ammount of circuitry you could change this to simple arrows ie rev up or down. |
|
| |
Premium models would be displayed on a HUD so you're not watching your dash as opposed to the road. |
|
| |
I like the 2 needle idea, but the problem is that the speed of the shafts is not important, it's the ratio between them which needs to match the gear that you are going in to. |
|
| |
It's very easy to learn the ratio. Before you change gear, look at the speed and the revs, that's roughly the ratio you will need to know. After you go into a gear look at the speed and the revs. |
|
| |
For practice I suggest a Mitsubishi Lancer, you can pop it out of gear without the clutch and then just push it towards the gear you want. As the ratios match it will just pop in, without any annoying grinding. This only works going up the gears, but once you've learned the ratios you can downshift too. |
|
| |
[ericscottf], because no car in my possession has, has had, or will have, ever, OBD. |
|
| |
[marklar], I used to have a BMW 2002 with which that could be done, with a bit of care, simply because the gearshift was slick and precise during those times when the bracket holding the mechanism to the back of the 'box was good and tight and the mechanism wasn't lying on the propshaft. I have also on occasion power-shifted almost like that in a Golf Mk I, despite the linkage being a bit clunky due to being composed of a mixture of 4-speed and 5-speed parts. The latter requires *thinking* of the clutch at the critical moment, though. |
|
| |
// the speed of the shafts is not important// Sorry [Marklar], the author didn't seem to be interested in clutchless gearshifting. He simply seems interested in knowing that, once the gear is selected, whether he could dump the clutch without jerking. In which case, cross-clutch speed differential is all that matters. |
|
| |
As for powershifting, yes, you'd need a speed sensor on each gear in the box, or a ratio calculator. |
|
| |
Seems like a good idea - and would be a useful tool to help people learn about changing gears smoothly. I learnt smooth gear-changing by driving a very old Morris, on which, after some practice, I could make all the gear changes without using the clutch. |
|
| |
Quite believable, [hippo]. I've got a 1000: another really good gearshift mechanism. I think that is important, one wants to be able to feel the shift fork detents very clearly. I wouldn't try it in a VW T2 Microbus, therefore. |
|
| |
I wouldn't try any form of power-shifting in a Minor, though. The Series II and early 1000 gearbox cases were made out of plaster of Paris, and the later ribbed cases weren't much better. |
|
| |
Custardguts: Any suggestion on how to get sensors into those rather inaccessible spots which aren't standard between transmission manufs? This is why i liked OBD to begin with, no transmission surgery necessary.
Ned: Without obd... well, you could hook up inputs to the speed cable and the tach cable. or forget the tach cable and go straight to the spark plug for engine rpm data.
but once again, the real trick is in figuring out what gear the car is in. A predictive algorithm could be used to "guess" what gear the user was about to go into, based on acceleration, but it wouldn't be good for those moments when this product would really shine - skipping a gear, downshifting by two or more, etc. (my car doesn't need its 4th gear. 3rd to 5th works quite well. YMMV.) |
|
| |
This may partially pay for itself in helping decrease clutch warranty costs. |
|
| |
We can prevent gear engagement by using solenoids, although it would be more expensive than what manufacturers are typically willing to do, as the business case is dodgy. |
|
| |
Although you might be able to match gears at specific RPM's, as a transmission engineer I do not recommend it, as you will most likely burn out your synchronizer rings. Generally expensive. |
|
| |
Here's an extension of this idea: |
|
| |
Have a series of closely spaced dual colour LED's (I think they exist) around the tachometer. |
|
| |
When you put the clutch in at 3000rpm in 2nd, the led closest to 6200 (or whatever) lights up blue (for downshifts), and the one at 1800 (or so) lights up green (for upshifts). |
|
| |
Any shift that ends up being above the engine redline is shown in red. |
|
| |
I was thinking you could have a dial showing the ratio between engine speed and road speed, from 0:1 through to 1:1 through to 1:0. Points that matched each gear ratio would be marked on the dial. Searched and found this idea, and now I'd add to my dial green, orange and red bands around each mark to indicate if it's close enough, and 2 or 3 color leds at each mark that will highlight the next gear ratio up and down, and maybe the current one when you press the clutch. |
|
| |
If you are thinking of a dial, then check out
"Synchroscope". It is used in the electrical
generation industry for getting speeds of
generators lined up with the national grid before
connection. It basically subtracts one speed from
the other, and the pointer rotates proportionally to
the difference. |
|
| |
These are fantastic ideas. I'd build one if I planned on ever
owning an internal combustion-powered vehicle. |
|
| |