Elevator Design Help

[GeeTwo]

Originally Posted by Ether

You might want to compare the output power of a window motor to the power required to lift the load at the desired rate.

Agreed. Max output power of a (denso) window motor is about 85W, or 63 ft-lb/s. Even if you managed to get its output at its best speed to the totes with those weightless, frictionless components you get to use in intro physics class, that would raise six totes (47 lb) at about 16 inches per second. Using real gears, chains/belts, and a lift plate, the answer is quite probably zero. Further, we have had window motor gearboxes nearly explode when we've tried to get too much out of them.

A CIM drawing 40A has about three times the power of a window motor at its maximum. If you're hoping to lift six totes briskly, one of these is probably not enough.

[Ether]

Originally Posted by GeeTwo

Agreed. Max output power of a (denso) window motor is about 85W

23 watts.

[Nathan Streeter]

The CIM's a very powerful motor... A 20:1 reduction with a CIM should be capable of lifting an elevator with 4-6 totes on it IF (and maybe only if?) the drum/winch is fairly small (think 2" diameter or so... smaller means more torque, less speed), your elevator is moderately efficient, and your gear reduction's not binding badly (due to a poorly manufactured gearbox, big side-loading from the spool belt/chain/wire tension, no greasing, or a variety of other things).

If you have a relatively large drum/spool - which it looks like you may have - then you'll need extra gear reduction... this may be a big part of the problem. I've never made a custom worm gearbox, but they introduce a variety of axial (thrust) and bending loads that you don't have with ordinary spur gears. These need to be handled well (bearings, stiff enough housing, well toleranced design & fab, etc.) or you'll get some bad binding/wear. I'm guessing the combination of large drum/spool and an inefficient gearing are combining to cause your issues... to check if your elevator itself is inefficient, does your carriage fall down by itself? Does it still fall by itself when you have tote(s) on it? Can you grab the cable (with the robot off!) and pull the carriage up and down by itself? Perhaps with a tote on it? It should at least fall by itself most of the time (although it's best to drive it down anyway!), otherwise you probably have serious binding in the elevator itself.

For reference, we're using a CIM in a WormBox (16:1 AndyMark Worm Drive gearbox with a 2-start worm and a 32-tooth helical gear) which we have modified to put a needle thrust washer at both ends of the worm... it now runs butter-smooth but had worked well without this improvement (we had kind of abused the gearbox last year so we wanted to be darn sure it lasted in this application). We then have a 30t HTD Timing Pulley (1.88" PD) cantilevered .25" away from the gearbox. This goes over an idler pulley at the top, and drives our top carriage, which has ball bearings riding on 80-20. Webbing then connects this to our lower carriages. This elevator system can pickup and hold 4 totes very easily... it's all efficient enough that 3 totes (and sometimes 2 totes) can backdrive it, so we actively hold position. The CIM has gotten warm - but not hot - to the touch after a day of practicing. So far we're very happy with it.

TL;DR, you shouldn't need more reduction... you probably have some considerable inefficiencies in your gearing (very likely) or your elevator (likely, but probably not as detrimental), which combined with a large spool/drum diameter are giving you problems. I recommend switching to a different gearbox (go COTS, probably!) and a smaller spool.

[GeeTwo]

Originally Posted by Ether

23 watts.

Oh, right - I was looking at the left side of the table (speed). Make that 4" per second in theoretical intro physics, and quite possibly zero for one tote. I knew that sounded like too much.

[grem110]

Thanks for all the advice! We will definitely reduce the diameter of the spool. I assume that keeping the worm gear for its anti-backdrive ability is a must? I ran some numbers and discovered that using an RS-775-18 with a 35:1 reduction would work pretty well--I may have forgotten to mention we have a block and tackle pulley system that acts as a 2:1 reduction, so the system would have an overall 70:1 reduction.

Also, how does amperage fit into all of this? Not sure how much amperage to give this motor when I still want to have the robot drive and grab the totes (grabber is motor driven).

Thanks so much for everyone's help.

[GeeTwo]

Originally Posted by grem110

Thanks for all the advice! We will definitely reduce the diameter of the spool. I assume that keeping the worm gear for its anti-backdrive ability is a must? I ran some numbers and discovered that using an RS-775-18 with a 35:1 reduction would work pretty well--I may have forgotten to mention we have a block and tackle pulley system that acts as a 2:1 reduction, so the system would have an overall 70:1 reduction.

Also, how does amperage fit into all of this? Not sure how much amperage to give this motor when I still want to have the robot drive and grab the totes (grabber is motor driven).

Thanks so much for everyone's help.

Not necessarily on the worm gear. We have a spur gear system (AndyMark Toughbox Mini) driving through a reducing chain, powering two lift chains, actuating a sliding (not rolling) lift plate. Among all this, we have enough drag that four totes will hold their position without extra braking. Other options include active brakes (such as a bicycle, drum, or disc brake -- usually powered by pneumatics, but you can use a lead screw) and motor brakes (applying a bit of current, but you do need to be careful that your motor can handle it for a long period of time), and anti-backdrive clutch (e.g. an adaptation of Dewalt a drill clutch).


Current (amperage) is the real limiting factor when you get to the larger motors. FRC rules essentially mandate that no single motor can draw more than 40A continuous because otherwise you will trip a breaker. A CIM motor delivers its peak 12V power at about 70A and 2500 RPM. You can get away with this for a second or so, but if you sustain it, you lose all use of the motor until the breaker resets. If you do this with two or more motors at the same time, you'll trip the 120A main breaker, and your match is over.
Here's a link to some motor performance curves at 12V.

And of course, on behalf of everyone, you're welcome. Always glad to help out.

[asid61]

Originally Posted by grem110

Thanks for all the advice! We will definitely reduce the diameter of the spool. I assume that keeping the worm gear for its anti-backdrive ability is a must? I ran some numbers and discovered that using an RS-775-18 with a 35:1 reduction would work pretty well--I may have forgotten to mention we have a block and tackle pulley system that acts as a 2:1 reduction, so the system would have an overall 70:1 reduction.

Also, how does amperage fit into all of this? Not sure how much amperage to give this motor when I still want to have the robot drive and grab the totes (grabber is motor driven).

Thanks so much for everyone's help.

If you use the jvn design calculator it makes all this calculation much easier. According to that, you are running 38 amps when you are lifting 70lbs- a lowish estimate for claw + totes + bin. However, when you factor in efficiency and random junk getting in the way, you should be gearing it down even further, maybe with a second 2:1 reduction to make your overall reduction 140:1.
That nets you a speed of 1.7ft/sec and has a big margin of safety.
Now, you still have to use a worm gear to prevent backdriving, as braking with an RS-775 alone may cause a fire. We added a bike brake to our lift to stop it, and that works wonderfully. There are specs online for where to mount a disc brake if you choose to go that route.
You can use a 70:1 reduction if you run a cim, as that will only draw 21 amps at max load and give you a max loaded speed of 1.6ft/sec. I do not know how slow you would have to run if you choose to brake with the motor alone.

[philso]

Originally Posted by shewejff

We made our gearbox out of a 2" X 3" X 1/8" rectangular aluminum tube. That works really well for the size of our gears, but the holes really need to be drilled perfectly. With our rudimentary workshop tools, it took us a couple tries to get it perfect.

Your first tournament is later this week and another the week after. You need to evaluate your risk and determine what you can do in the time available. There is no "playing defense because you can't score" this year. Can you make a new worm gear box accurately enough? Can you get an AndyMark RAW box in time? Can you do something that emulates what oRyon Labs did? The last two options may be your best.

Originally Posted by grem110

I assume that keeping the worm gear for its anti-backdrive ability is a must? I ran some numbers and discovered that using an RS-775-18 with a 35:1

oRyon Labs and others (including our team) have found that anti-back drive is not really necessary. Even if you choose gearing such that the winch backdrives slowly, your drivers can probably learn to live with it. The alternative is to not be able to do anything significant on the field.

The RS-775 and RS-550 motors rely on the integral fan for cooling. The can overheat and fail if you are going to run them at low speeds (including stall condition = 0 rpm). The CIM, Mini-CIM and BAG motors rely on the thermal mass and heat dissipation through the motor casing, without air being forced over them, for cooling and will be much more durable in this kind of application.

[lark95]

We are using a worm box this year as well. At first we had some real problems with the worm gear getting worn very quickly and the jamming the whole assembly. Our problem was the the center to center distance was to large. we made a new box with tighter spacing and that fixed all our problems. You also might want to make sure that the output shaft on the gearbox is not bent at all, one of ours was and as it rotated it varied the center to center distance to the point that sometimes it would be to tight and sometimes to loose.

[Ether]

Originally Posted by asid61

If you use the jvn design calculator it makes all this calculation much easier. According to that, you are running 38 amps when you are lifting 70lbs- a lowish estimate for claw + totes + bin. However, when you factor in efficiency and random junk getting in the way, you should be gearing it down even further, maybe with a second 2:1 reduction to make your overall reduction 140:1.
That nets you a speed of 1.7ft/sec and has a big margin of safety.

The free speed of the RS775-18 at 12 volts is 13000 rpm. With a 140:1 gear reduction and the 6" diameter winch drum and 2:1 pulley reduction in the OP's design, that's only 1.2 feet/sec load speed. And that's with the motor at free speed... so with a load you'll be nowhere near that speed.

Start with required power and work from there. Take the 70 pound load and add a 50% margin (or more!) to account for friction and other factors. That's 105 pounds.

The mechanical power required to lift a 105 pound load at 1.7 ft/sec is 105*1.7*1.356=242 watts. From the motor curve at 12 volts, the RS775-18 will generate 242 shaft watts at 8632 rpm while drawing 30 amps. At that operating point it will be generating 122 watts of heat, which is about 4 times the heat it generates at max efficiency (and at much lower speed). So this is far from optimal for that motor.

Now calculate the required speed reductions. To get 1.7 ft/sec with 8632 motor rpm, you'll need a gear ratio of about 66:1 (with 6" diameter winch drum and 2:1 pulley).

You can use a 70:1 reduction if you run a cim, as that will only draw 21 amps at max load and give you a max loaded speed of 1.6ft/sec. I do not know how slow you would have to run if you choose to brake with the motor alone.

The CIM's free speed is 5310 rpm. With a 70:1 gear reduction and the 6" diameter winch drum and 2:1 pulley reduction in the OP's design, that's only 1 foot/sec load speed. And that's with the motor at free speed... so with a load you'll be nowhere near that speed.

Again, start with required power and work from there. You need 242 watts (see above). From the motor curve at 12 volts, the CIM will generate 242 shaft watts at 4065 rpm while drawing 33 amps. At that operating point it will be generating 157 watts of heat, which is about twice the heat it generates at max efficiency.

Now calculate the required speed reductions. To get 1.7 ft/sec with 4065 motor rpm, you'll need a gear ratio of about 31:1 (with 6" diameter winch drum and 2:1 pulley).

[AdamHeard]

Originally Posted by philso

The RS-775 and RS-550 motors rely on the integral fan for cooling. The can overheat and fail if you are going to run them at low speeds (including stall condition = 0 rpm). The CIM, Mini-CIM and BAG motors rely on the thermal mass and heat dissipation through the motor casing, without air being forced over them, for cooling and will be much more durable in this kind of application.

You mean they can overheat if you run sustained high currents. Stalling them doesn't imply high current draw (when stalled at lower voltages).

[pmangels17]

There has been a lot of good info posted about gearing and motor selection, so I will leave that topic alone. However, if you decide you want to do a worm, make sure that the worm gear is bronze and not cast iron. Our team has had nothing but trouble with cast iron worm gears, and far better success with bronze worm gears. Also, make sure you keep the gearbox greased and somehow shield it from debris that might enter the box, either via proximity to the field floor or from maintenance work done above. Often a major contributor to inefficiency is something as simple as one small chip getting stuck in the grease in the gearbox.

[philso]

Originally Posted by AdamHeard

You mean they can overheat if you run sustained high currents. Stalling them doesn't imply high current draw (when stalled at lower voltages).

Yes, you are correct. I have not seen anything in their data sheet indicating what current they will tolerate with no cooling. The stall current in an application like a winch would be dependent on a number of factors (like the load, the gearing, the efficiency). I doubt that the OP has time before his first tournament to test for this so it is probably best for him to try a less risky solution than using the RS-775.

[lark95]

Originally Posted by philso

The RS-775 and RS-550 motors rely on the integral fan for cooling. The can overheat and fail if you are going to run them at low speeds (including stall condition = 0 rpm). The CIM, Mini-CIM and BAG motors rely on the thermal mass and heat dissipation through the motor casing, without air being forced over them, for cooling and will be much more durable in this kind of application.

We had this problem in 2013. we had an elevator that was con counterbalanced which was run by a BB550 on a planetary gearbox. In order for the elevator to stay in one spot the motor had to run stalled, it got very hot very quickly.

[Certified Nerd]

Originally Posted by dradel

Why go for 6 totes. From what I have seen from most teams is it is going to take to long to make 6 stacks. And really a 6 stack is pointless without a can on top.

My thought is it is better to have a team make a 2 or three stack quickly and place on the scoring platform and perhaps have another team make another stack with a can at same time and combine them. Or perhaps a team that can top a stack with a can with a noodle.

There are many ways to go about it, efficiency is the key in my opinion.

I had the same thought. that's why we built a general purpose grabber.