How are the actuation cables in multi-stage telescoping lifts tensioned?

[Robot360]

For multi-stage telescoping lift designs where the actuation motor actively drives the stages up and down, how are the actuation cables typically kept tensioned? I have searched throughout the forums and white papers, and have not seen much on specifically how the cables kept tensioned.

For instance, with a continuous-style telescoping lift design, the up-feed cable and down-feed cable should ideally feed in and out at the same rate, but since they wrap around drive pulleys, their effective feed rates will differ from each other as the lift extends/retracts since their pulley-wrap radius differs depending on how much cable is on each drive pulley (assuming the cable wraps onto itself). Thus, in that case, an active tensioner would be needed to take up the slack. A possible exception could be if the cables spiraled around the drive pulleys in a single layer (so the wrap radius remains constant), but in that case, is something done to ensure the single-layer spiral wrapping?

A similar cable-tensioning issue could potentially exist with cascade-style telescoping lift designs, too.

Are there schemes in place to ensure the cables always remained tensioned without an active tensioner?

For this thread, I am only referring to designs where one motor (or a group of coupled motors) drive all of the stages, not designs where each stage has its own motor.

I would appreciate your insight. Thanks!

[EricH]

Running that sort of system... I don't think 330 used anything more than a turnbuckle while running wire rope through their lifts. (Up to 10' long lifts, for anybody keeping track at home.) Generally, though, we were able to keep the wraps flat, just by having enough space. I don't think we did anything else special. (Turnbuckle, if used, was a between-match tensioning if things were getting loose.)

Remember that the cable is often only about 1/4" thick, or maybe less, counting the coating, so if you have two wraps, the effective diameter only increased by 1/2" or so; most likely, the other direction decreased by that much. You'd notice a variation in speed, maybe, but more likely than not tension won't be affected very much. Maybe use a small extension spring, but that's about it.

[Robot360]

Originally Posted by EricH

Running that sort of system... I don't think 330 used anything more than a turnbuckle while running wire rope through their lifts. (Up to 10' long lifts, for anybody keeping track at home.) Generally, though, we were able to keep the wraps flat, just by having enough space. I don't think we did anything else special. (Turnbuckle, if used, was a between-match tensioning if things were getting loose.)

Remember that the cable is often only about 1/4" thick, or maybe less, counting the coating, so if you have two wraps, the effective diameter only increased by 1/2" or so; most likely, the other direction decreased by that much. You'd notice a variation in speed, maybe, but more likely than not tension won't be affected very much. Maybe use a small extension spring, but that's about it.

Thanks for the quick reply, EricH. Did 330 do anything specific to ensure the wraps remained flat? Regardless, do you have a few pictures of the setup to share?

Regarding your comment about the diameter changing with wrapping cable on itself: Even a wrap diameter increase of 0.5" is approximately an additional 1.5" of circumference, meaning with one drive pulley revolution, one cable will have fed in or out 1.5" further (assumes the up-feed and down-feed drive pulleys are driven off of the same shaft), which seems significant. Thoughts?

[EricH]

Originally Posted by Robot360

Thanks for the quick reply, EricH. Did 330 do anything specific to ensure the wraps remained flat?

Nope, not that I can recall. I've seen some industrial solutions that involved grooved drums. I think we may have done the wrapping before stringing cable, but that's it.

Regardless, do you have a few pictures of the setup to share?

Not specifically. However...
http://www.chiefdelphi.com/media/photos/17610; Zoom in on the 330 machine, down low in front, and you'll see the winding drum.
http://www.chiefdelphi.com/media/photos/17599 is another of the same machine; similar location.
http://www.chiefdelphi.com/media/photos/12345; You might be able to make out the winch behind the green light.

There's another picture, from 2002, but it has a similar problem: The drum is obscured behind the robot frame, which happened to include the guide rails for what we were lifting. I can't recall us ever having an issue with slack cables, though, without some other major mechanical problems. (You'll also notice that 2004 is the last time 330 used a lift of that type in competition--2010 used a leadscrew, IIRC--but it had more to do with us finding single-joint arms to be at least as effective as single- or multi-stage lifts for easier design than with any problems with the lifts.)

Regarding your comment about the diameter changing with wrapping cable on itself: Even a wrap diameter increase of 0.5" is approximately an additional 1.5" of circumference, meaning with one drive pulley revolution, one cable will have fed in or out 1.5" further (assumes the up-feed and down-feed drive pulleys are driven off of the same shaft), which seems significant. Thoughts?

Consider this: Should your cables happen to be attached to the same moving object in the lift, which I consider highly likely, you'll have a retract full and an extend empty at lowest position. The extend cable speeds up as it wraps up; the retract cable slows down as it gets pulled out. Going down, the opposite is true. That's a setup that will generally keep tension in the system, no?

[jspatz1]

In both our cascade-rigged telescopic lift in 2011, and our continuous-rigged telescopic lift this year (2013), we maintained tension and accommodated small variations in overall length by terminating the cable with a strong extension spring. The spring takes up variations in the rigging length as the apparatus extends and retracts, and is placed on the non-loaded side (or "down" side) of the rigging. A small loop of slack cable is also placed around and parallel to the spring as a shock cord, to create a maximum limit to how much the spring can stretch. This protects the spring in cases where the mechanism gets caught or snags, and the down direction sees a high load.

[Al Skierkiewicz]

We would put a wire tie around the lift string and hold it tensioned (to the side) with surgical tubing. That gave a shock relief to the assembly when it ran into a mechanical stop as well.

[sanddrag]

968 used a bungee cord between the carriage and the down wire rope.

[Robot360]

Originally Posted by EricH

Nope, not that I can recall. I've seen some industrial solutions that involved grooved drums. I think we may have done the wrapping before stringing cable, but that's it.

Not specifically. However...
http://www.chiefdelphi.com/media/photos/17610; Zoom in on the 330 machine, down low in front, and you'll see the winding drum.
http://www.chiefdelphi.com/media/photos/17599 is another of the same machine; similar location.
http://www.chiefdelphi.com/media/photos/12345; You might be able to make out the winch behind the green light.

There's another picture, from 2002, but it has a similar problem: The drum is obscured behind the robot frame, which happened to include the guide rails for what we were lifting. I can't recall us ever having an issue with slack cables, though, without some other major mechanical problems. (You'll also notice that 2004 is the last time 330 used a lift of that type in competition--2010 used a leadscrew, IIRC--but it had more to do with us finding single-joint arms to be at least as effective as single- or multi-stage lifts for easier design than with any problems with the lifts.)

Thanks for the pictures!

Originally Posted by EricH

Consider this: Should your cables happen to be attached to the same moving object in the lift, which I consider highly likely, you'll have a retract full and an extend empty at lowest position. The extend cable speeds up as it wraps up; the retract cable slows down as it gets pulled out. Going down, the opposite is true. That's a setup that will generally keep tension in the system, no?

Yes, the feed rates change as the lift extends and retracts, but anytime the cables are not at the same feed rate, there will be cable slack (or binding) in the system.

For example, when lowering the lift, the up-feed cable's feed rate will start out being relatively fast since its drive pulley is fully wrapped, and then slow down as the cable unwraps. The down-feed cable's feed rate will start out being relatively slow, and then speed up as the cable wraps. The result will be initial cable slack. When the lift passes through its mid-portion of its stroke, the wraps on both drive pulleys will be relatively equivalent, so the feed rates will be relatively equivalent (thus, the slack will still be present, but not increasing or decreasing). As the lift approaches being fully lowered, the down-feed cable's feed rate will now be faster than the up-feed cable's feed rate since the down-feed cable's drive pulley has more wraps. The result is that the system will take the slack back up. However, during the stroke, the system will have slack (for the reasons mentioned above). Thus, unless you have a system that keeps the cable wraps flat, it seems a tensioner would be needed to keep the cable tight, right?

[Robot360]

Originally Posted by jspatz1

In both our cascade-rigged telescopic lift in 2011, and our continuous-rigged telescopic lift this year (2013), we maintained tension and accommodated small variations in overall length by terminating the cable with a strong extension spring. The spring takes up variations in the rigging length as the apparatus extends and retracts, and is placed on the non-loaded side (or "down" side) of the rigging. A small loop of slack cable is also placed around and parallel to the spring as a shock cord, to create a maximum limit to how much the spring can stretch. This protects the spring in cases where the mechanism gets caught or snags, and the down direction sees a high load.

Jeff, thanks for the response and details. For your setup, could you ever get in a situation where your lift snagged, maximizing the spring stroke (that is, to its limiter), and then if you continued driving the lift, cable slack would end up in the system?

Also, how many layers of cable wraps are typically on your full drive pulley, and with what diameter cable? Even a few layers of 1/4" cable could result in several inches of cable slack.

[MichaelBick]

Originally Posted by sanddrag

968 used a bungee cord between the carriage and the down wire rope.

It looks like 254 did the same using a spring on slipstream.

http://media.team254.com/2012/08/slipstream.jpg

With cascade there is usually not a problem. From what I've sent eh more popular solution with cascaded elevators is chain, belts, or rack/pinion, which either isn't prone to losing tension or can easily be tensioned using a method other than a spring.

[jspatz1]

Originally Posted by Robot360

Jeff, thanks for the response and details. For your setup, could you ever get in a situation where your lift snagged, maximizing the spring stroke (that is, to its limiter), and then if you continued driving the lift, cable slack would end up in the system?

Also, how many layers of cable wraps are typically on your full drive pulley, and with what diameter cable? Even a few layers of 1/4" cable could result in several inches of cable slack.

When the carriage was obstructed there would be a small amount of slack develop on the up side but not enough to de-track the cable or cause problems. I'm talking about a small amount of spring range, just enough to take up small variations in the rigging. If stretched tight and driven too hard we would stall or break something.

We size the diameter and width of the winch drum to require only one layer of cable wrap. With the proper alignment the cable can be made to wind uniformly in one layer. The spring tension helps with this. Note on the photo of 254 the drum is wide enough to easily allow one layer. This keeps the rigging length constant with only small variations, such as for parellelism. We have used 1/16 to 1/8" wire cable. It is surprisingly strong, more supple than larger cable, and adequate if the apparatus is properly counterbalanced. Few FRC mechanisms really need 1/4 cable IMHO. That's dang strong.

[Robot360]

Originally Posted by jspatz1

When the carriage was obstructed there would be a small amount of slack develop on the up side but not enough to de-track the cable or cause problems. I'm talking about a small amount of spring range, just enough to take up small variations in the rigging. If stretched tight and driven too hard we would stall or break something.

We size the diameter and width of the winch drum to require only one layer of cable wrap. With the proper alignment the cable can be made to wind uniformly in one layer. The spring tension helps with this. Note on the photo of 254 the drum is wide enough to easily allow one layer. This keeps the rigging length constant with only small variations, such as for parellelism. We have used 1/16 to 1/8" wire cable. It is surprisingly strong, more supple than larger cable, and adequate if the apparatus is properly counterbalanced. Few FRC mechanisms really need 1/4 cable IMHO. That's dang strong.

Jeff, thanks for the clarification, and I agree with your comments regarding cable sizing.