28 replies to this topic

[FSJ]

The other possible solution is to stop making everything(exageration) Emergency Shutdowns. They are more appropriate for drive or control failures. Detected faults should trigger an appropriate action. For example, if a grip fault is detected, a regenerative drive may be able to make a more appropriately controlled stop than the emergency brake. There could be a concern about multiple failures.

Another problem is that as the standard becomes more performance orientated, we begin to worry more about numbers than actual perfomance. If a lift stops within the required distance without causing a dangerous condition, (i.e., excessive chair swing) do the actual decel numbers really matter? There can be a great variation in the dynamic response between lifts the same input. Short span heavily tensioned lifts respond much different than long spans with low tension.

[shoemaniii]

this is a great topic!

regarding the modulation of a solenoid valve, have often wondered how honda does it so inexpensively: the honda cbr1000rr has had a speed sensative steering damper for a coupla years now. pretty cool. the rate of dampening is very low at street speeds, higher at ludicrous speed.
bobp (bike nut)

[liftmech]

Lelec, on Aug 30 2007, 03:00 PM, said:

- For detachables mind the distance between the device(s) detecting improper grip attachment at the top terminal and the breakover point at the top tower. An emergency stop must bring the lift to a standstill in less than this distance because otherwise an improper grip attachment (typically an emergency stop) could get ugly. This bites me in the butt way more often than carrier spacing considerations do.

Yes, when you consider that the distance between the terminal end and the first breakover tower is often in the 30-60-foot range, this can get interesting. We've had inspectors note this in the past, and we now require that when downloading the lift cannot be run over a certain speed. This of course varies widely.

[lastchair_44]

liftmech, on Sep 1 2007, 02:19 PM, said:

Yes, when you consider that the distance between the terminal end and the first breakover tower is often in the 30-60-foot range, this can get interesting. We've had inspectors note this in the past, and we now require that when downloading the lift cannot be run over a certain speed. This of course varies widely.

800fpm for us, and we had to change our grip attach stop from an e stop to a normal stop so cabins will stop on the first sheave of tower 28. We run at 800 during the summer and between 1000, and 1100 during the winter. During winter ops we have an unload operator (sometimes two) and one downloader. We need the extra operator to facilitate the downloading of our non-skier/snowboarder restaurant patrons.

[Lelec]

shoemaniii, on Aug 31 2007, 05:08 PM, said:

this is a great topic!

regarding the modulation of a solenoid valve, have often wondered how honda does it so inexpensively: the honda cbr1000rr has had a speed sensative steering damper for a coupla years now. pretty cool. the rate of dampening is very low at street speeds, higher at ludicrous speed.
bobp (bike nut)

These systems work almost identically to a buck topology switching power supply. If you replace the reactive element (the inductor) in these supplies with the solenoid (also an inductor) you can basically turn a bistable valve into a proportional valve by fine control of the current through the solenoid. Ignoring switching and I^2 * R losses the voltage across the load in such system = Duty Cycle (%) * Vin. Since you can not instantaneously change the current through an inductor the load current will "ramp" (it is actually an exponential curve) up towards saturation when the switching element is "on" and ramp down towards zero when the switching element is "off". The "ramp" rate is dictated by the LR time constant with "R" being the inherent coil resistance in the on state and the inherent coil resistance + the on state resistance of load flyback diode in the "off" state. By careful selection of switching frequency and duty cycle range the result is a nearly DC current with "ripple" riding on it and with the ripple amplitude dictated by the switching frequency. i.e. the less time that you allow the current to "ramp" up or down the less it will deviate from the average. Add speed feedback to the equation and the braking force (or steering damping level) can be finely controlled based on this speed feedback by changing the duty cycle (or pulse width - this is where the term pulse width modulation or PWM comes from) of the switching element control signal.

With that as background info and assuming I haven't bored you to death, the widespread and increasing use of such PWM systems has lead to economies of scale and the associated dramatic drop in the price of the components necessary to realize such a system. For example a PWM routine in a typical PLC, using standard I/O can realize a switching frequency of a few hertz due to scan time requirements etc. This leads to the "clapping" effect (think of those little toy monkeys playing the cymbals) that you see on alot of older systems because the frequency is so low that the current goes discontinuous on each cycle. Contrast this with switching power supplies, that often need 1% requlation or better and therefore often have switching frequencies in excess of 400 kHz. There are now inexpensive PWM systems that easily switch at 100's or 1000's of hertz (the inductance value is high enough on most solenoid coils that 100 hertz gives a fine "analog" type response). I hope this was helpful and apologize if it is too "granular" as this is a subject to which I have applied a great deal of study. As a side note (since you are a self described "bike nut") BMW has a stability control system on some of their motorcycles that use the same principle and are pretty cool. It takes some getting used to when you apply the front brakes (hard) and there is no dip on the front end though. You can bring them to a skidding stop using the front brakes alone with only minute changes in the angle of the frame to the roadway.

[Emax]

Lelec, on Sep 4 2007, 11:02 AM, said:

These systems work almost identically to a buck topology switching power supply. If you replace the reactive element (the inductor) in these supplies with the solenoid (also an inductor) you can basically turn a bistable valve into a proportional valve by fine control of the current through the solenoid. Ignoring switching and I^2 * R losses the voltage across the load in such system = Duty Cycle (%) * Vin. Since you can not instantaneously change the current through an inductor the load current will "ramp" (it is actually an exponential curve) up towards saturation when the switching element is "on" and ramp down towards zero when the switching element is "off". The "ramp" rate is dictated by the LR time constant with "R" being the inherent coil resistance in the on state and the inherent coil resistance + the on state resistance of load flyback diode in the "off" state. By careful selection of switching frequency and duty cycle range the result is a nearly DC current with "ripple" riding on it and with the ripple amplitude dictated by the switching frequency. i.e. the less time that you allow the current to "ramp" up or down the less it will deviate from the average. Add speed feedback to the equation and the braking force (or steering damping level) can be finely controlled based on this speed feedback by changing the duty cycle (or pulse width - this is where the term pulse width modulation or PWM comes from) of the switching element control signal.

With that as background info and assuming I haven't bored you to death, the widespread and increasing use of such PWM systems has lead to economies of scale and the associated dramatic drop in the price of the components necessary to realize such a system. For example a PWM routine in a typical PLC, using standard I/O can realize a switching frequency of a few hertz due to scan time requirements etc. This leads to the "clapping" effect (think of those little toy monkeys playing the cymbals) that you see on alot of older systems because the frequency is so low that the current goes discontinuous on each cycle. Contrast this with switching power supplies, that often need 1% requlation or better and therefore often have switching frequencies in excess of 400 kHz. There are now inexpensive PWM systems that easily switch at 100's or 1000's of hertz (the inductance value is high enough on most solenoid coils that 100 hertz gives a fine "analog" type response). I hope this was helpful and apologize if it is too "granular" as this is a subject to which I have applied a great deal of study. As a side note (since you are a self described "bike nut") BMW has a stability control system on some of their motorcycles that use the same principle and are pretty cool. It takes some getting used to when you apply the front brakes (hard) and there is no dip on the front end though. You can bring them to a skidding stop using the front brakes alone with only minute changes in the angle of the frame to the roadway.

Granular? The Devil may be in the details, but so is the elegant solution.

I'd LOVE to work on a project with you. We could make beautiful music.

This post has been edited by Emax: 05 September 2007 - 05:23 PM

[liftmech]

I'd be very interested to see what you two could come up with...

[skisox34]

I believe it would be akin to two mad scientists working together muhaha

[shoemaniii]

i have a '78 cb750k (lightened/loud/black) that also doesn't dip when the front brakes are applied. nothing to do w/fancy/smancy PWM systems tho, just prehistoric brakes. think i'll get on my husky 610 to clear my head - a tad over 300lbs, 50hp, E-start, brembo brakes, etc - yeehaw. see ya!
bobp

This post has been edited by shoemaniii: 12 September 2007 - 12:12 PM