Sprockets as Chain Tensioners?

[NickE]

Keep in mind that anything you add into the chain run will add friction and decrease the efficiency of the system.
The only way to avoid this is to either size the center-to-center distances perfectly so that no tensioner is needed or to have a system where the axles slide to tension the chain.

[Ether]

Originally Posted by NickE

Keep in mind that anything you add into the chain run will add friction and decrease the efficiency of the system.
The only way to avoid this is to either size the center-to-center distances perfectly so that no tensioner is needed or to have a system where the axles slide to tension the chain.

Does anyone have any data quantifying this effect ? At least some ballpark numbers. It would be useful to know just how "floating sprocket" tensioners, rotating tensioners, sliding tensioners, and adjustable axles compare.

[Chris is me]

Originally Posted by Ether

Does anyone have any data quantifying this effect ? At least some ballpark numbers. It would be useful to know just how "floating sprocket" tensioners, rotating tensioners, sliding tensioners, and adjustable axles compare.

I've been interested in making a simple test bed for something like this but I have no idea what I should be measuring - I assume current draw?

Personally I'm going to wager a guess that friction in tensioners is essentially negligible. Shaker just uses blocks of Delrin or nylon pushing against the chain (not spinning) and we have no problems at all. We even had a multi-purpose Delrin rod we used as both a tensioner and as a structural member (standoff)

[Ether]

Originally Posted by Chris is me

I've been interested in making a simple test bed for something like this but I have no idea what I should be measuring - I assume current draw?

You need current and speed, or current and voltage, or voltage and speed.

Ideally one of the two parameters is held fixed to make it easier to compare.

If you can measure all three then the extra data can be used as a cross-check of the test's integrity.

[Cyberphil]

On our first failed ball control system, we used a free-floating sprocket as a chain tensioner, just as you are describing. It worked great, but did fall off once or twice with a hard hit. I think it was a 24 tooth 25-chain sprocket that was in the middle of two 12 tooth sprockets spaced about 6 inches apart. So it was a small system, but it did work very well.

[Cory]

Originally Posted by Chris is me

I've been interested in making a simple test bed for something like this but I have no idea what I should be measuring - I assume current draw?

Personally I'm going to wager a guess that friction in tensioners is essentially negligible. Shaker just uses blocks of Delrin or nylon pushing against the chain (not spinning) and we have no problems at all. We even had a multi-purpose Delrin rod we used as both a tensioner and as a structural member (standoff)

I doubt that a static tensioner has a negligible effect on efficiency.

If you've got a two stage spur gear reduction and a single chain run from the gearbox you're already down to 83% efficiency or so. If you then shove a block of plastic into the chain to tension it, it would seem you would be liable to lower efficiency quite a bit.

We have no data, but empirical evidence has suggested to us that our overall drivetrain design is noticeably more efficient than the average drive similar to ours in terms of same number and type of motors, similar gear reductions, top speeds, and wheel type/friction.

[Ether]

Originally Posted by Cory

We have no data, but empirical evidence has suggested to us that our overall drivetrain design is noticeably more efficient than the average drive similar to ours in terms of same number and type of motors, similar gear reductions, top speeds, and wheel type/friction.

I'm not saying you're wrong, but without any data how do you draw this conclusion? What "noticeable" observations did you make? Not criticizing, just very curious.

[Chris Fultz]

I think the difference is evident in a few ways.

First is that engineering judgement would tell you anything and everything you add into the chain circuit is going to increase friction and reduce efficiency.

Secondy, the robot performance on the field confirms the statement. 254's robots always move faster and are more durable, even compared to other robots that are using a very similar drive.

Anyone who has seen their robots up close, or competed with them on the field, would agree.

[Ether]

Originally Posted by Cory

I doubt that a static tensioner has a negligible effect on efficiency.

If you've got a two stage spur gear reduction and a single chain run from the gearbox you're already down to 83% efficiency or so. If you then shove a block of plastic into the chain to tension it, it would seem you would be liable to lower efficiency quite a bit.

We have no data, but empirical evidence has suggested to us that our overall drivetrain design is noticeably more efficient than the average drive similar to ours in terms of same number and type of motors, similar gear reductions, top speeds, and wheel type/friction.

If I may, I have a couple of questions, Cory.

I infer that you do not use static tensioners, is that correct? If so, do you use rolling tensioners, floating-sprocket tensioners, or no tensioners at all?

When you are comparing the performance of your drivetrain design to other drivetrains similar to yours, is "similar" limited to designs which use static tensioners (but are similar in all the other ways you mentioned), or does it also include designs that use the same tensioner design as yours?

[EricH]

254 has a "mobile" sprocket system, where the end bearing blocks get pulled outwards/inwards to tension the chain. The only things on the chain are the gearbox sprockets and the wheel sprockets (it's a live-axle system). I don't know how they compare, but there aren't a lot of people who use the bearing-block tensioning system that I know about. So it seems reasonable that they look at the basic drive system without regard to the tensioners.

[Ether]

Originally Posted by EricH

254 has a "mobile" sprocket system, where the end bearing blocks get pulled outwards/inwards to tension the chain. The only things on the chain are the gearbox sprockets and the wheel sprockets (it's a live-axle system). I don't know how they compare, but there aren't a lot of people who use the bearing-block tensioning system that I know about. So it seems reasonable that they look at the basic drive system without regard to the tensioners.

My questions for Cory were an attempt to discern from the anecdotal evidence given whether 254's superior drivetrain performance is due exclusively to tensioner design, or whether other design issues or craftsmanship play a larger role. This is an important distinction which has practical ramifications for other teams trying to emulate their success.

[AdamHeard]

Originally Posted by Ether

My questions for Cory were an attempt to discern from the anecdotal evidence given whether 254's superior drivetrain performance is due exclusively to tensioner design, or whether other design issues or craftsmanship play a larger role. This is an important distinction which has practical ramifications for other teams trying to emulate their success.

The design has an emphasis on efficiency throughout. All ball bearings, a minimization of parts & gear reductions whenever possible, and no tensioners touch the chain (sliding bearing blocks as already said).

Each item is a small percentage, but I imagine they certainly add up. We run a very similar drive, and teams often comment on how quiet it is.

[Jared Russell]

I will buy that static tensioners are inefficient (seems obvious). But what about idler sprockets? Worst case you are adding in the friction of an extra ball bearing, plus the angular momentum of the sprocket itself. The losses from such a system should certainly be dwarfed by factors like alignment, lubrication, and gearing.

[Aren_Hill]

Well considering every link in a chain is essentially a steel pin in a steel bushing, the more you make those move or twist (bends in the chain path), the less efficient it will be.

[Ether]

Originally Posted by Jared341

I will buy that static tensioners are inefficient (seems obvious). But what about idler sprockets? Worst case you are adding in the friction of an extra ball bearing...

My gut instinct is that the friction at the interface of the idler sprocket teeth with the chain would be a bigger factor (albeit still small) than the idler's ball bearing friction.

The losses from such a system should certainly be dwarfed by factors like alignment, lubrication, and gearing.

That's my guess too. Which is why it would be most instructive to see some real apples-to-apples data comparing a well-designed system to itself: with and without a tensioner touching the chain.