CIM Voltage and Motor Suggestions

[jakep]

Hey,

I was looking at the CIM Motor specs, and I was wondering if anyone has tried running them at 24 volts? All the spec sheets I saw say 12 Volts, but I was wondering if that was just because FIRST bots are always on a 12 volt system, so performance at higher voltages just isn't reported.

For this one robotics project, we are looking at using 2 CIMs to power it. However, according to our calculations, we will need to double the Horsepower on these motors if they will ever work for our needs. They are rated at ~.34 HP, and we need something that performs at ~.7 HP. My theory was if we doubled the voltage, this would work. However, I'm not sure if this will burn out the motors.

If this is not possible, and our small team is getting a bit desperate with this summer project of ours, can anyone recommend some relatively cheap motor/gearbox combination that would have around 250 inch-pounds of torque at 40amps? I saw: http://www.npcrobotics.com/products/viewprod.asp?prod=42&cat=20&mode=gfx
but those just seem way overpriced.

Any help would be greatly appreciated.

[Kevin Sevcik]

Hey,

I was looking at the CIM Motor specs, and I was wondering if anyone has tried running them at 24 volts? All the spec sheets I saw say 12 Volts, but I was wondering if that was just because FIRST bots are always on a 12 volt system, so performance at higher voltages just isn't reported.

For this one robotics project, we are looking at using 2 CIMs to power it. However, according to our calculations, we will need to double the Horsepower on these motors if they will ever work for our needs. They are rated at ~.34 HP, and we need something that performs at ~.7 HP. My theory was if we doubled the voltage, this would work. However, I'm not sure if this will burn out the motors.

If this is not possible, and our small team is getting a bit desperate with this summer project of ours, can anyone recommend some relatively cheap motor/gearbox combination that would have around 250 inch-pounds of torque at 40amps? I saw: http://www.npcrobotics.com/products/viewprod.asp?prod=42&cat=20&mode=gfx
but those just seem way overpriced.

Any help would be greatly appreciated.

Is your estimated horsepower need at the 250 in-lb point on the torque-speed curve? Basically, do you need roughly 180 RPM at this torque or can you just gear something really low?

Generally speaking, the maximum power output of a motor goes up as the square of the voltage. If you up the voltage you up the current by the same ratio, so twice the voltage is four times the power. If you wanted twice the peak power, up the voltage by root 2. So figure about 18V, not 24V.

I think you could operate the CIMs at peak power for a reasonable amount of time at 12V, but I'm not sure sure they'd like having to dissipate twice the heat.

[jakep]

Yes, we would need roughly 180RPM at this torque, if not closer to 200 or so.

I see what you mean about the 18 volts, but can the CIMs handle it?

[Cuog]

DC motors can run at as much as double their intended voltage "safely" however 150% is usually the maximum recomended so 18V should be ok, just make sure to stop using it if the motors start getting too hot to avoid damaging the motors.

[sanddrag]

You could try one and if it doesn't work you're out $28. However, how about just using more motors?

[Kevin Sevcik]

Well at max power, the CIM is gonna be drawing around 100A at 18V, so it's going to have to sink about 1200W of heat or so. Which is a good bit. I'm not saying it's going to catch fire, but you're probably going to want some sort of cooling on it. I figure a ballpark estimate would be a rise of atleast 1 deg C every second the motor is running at max power at 18V, probably more.

And don't forget you'll want a pair of Victor 885s to power those.

[Richard Wallace]

Check out the CIM data (http://www2.usfirst.org/2005comp/Specs/CIM.pdf) given by its manufacturer. In particular, read these notes:

1) DUTY CYCLE:INTERMITTENT, CCW 3 MINUTES ON 2 SECOND
CW 3 MINUTES ON 30 MINUTES @64.0 OZ-IN,
2) LIFE:1000 CYCLES MINIMUM

shown near the lower right corner under the heading "Special Features." Admittedly it's a little cryptic, so let me try to parse it. In plainer English it says that, if you use the motor to drive a 64 oz-in (4 lbf-in) load that requires a cycle of motion in which it first turns counterclockwise as fast as it can from a 12V supply for 3 minutes, then rests for two seconds, then turns clockwise as fast as it can from a 12V supply for 3 minutes, and finally rests for 30 minutes before starting the same cycle again, then the motor will last for at least 1000 of these cycles. The load cycle thus described is typical of a winch, which is the application for which the CIM was originally designed. So the total operating time for this load rating (12V, 27A, 4 lbf-in, +/-4320 RPM intermittent reversing duty) is 6000 minutes or 100 hours.

Back to your question: can you get more power out of the CIM by increasing the supply voltage. Kevin correctly pointed out that you can approximately double the peak power if you increase the voltage by a factor of sqrt(2). Now your question is: can the CIM handle that?

The question amounts to asking how to de-rate the CIM for operation at higher voltage and higher loading. Cuog recalls correctly that dc motors can typically be operated at up to 150% of rated loading; however, this increase does not come for free -- the penalty for overloading a motor is that it will wear out faster. You may not care about this if your robot only needs to last for a few operating hours.

Two overload stress factors will tend to shorten the CIM's expected life. One is running at higher speed, which will cause both the bearings and the brushes to wear faster. The other is running at higher currents, which will cause the armature coils to run hotter, thereby degrading the life of the insulation, the commutator, and the bearings, since these are all parts of the overheated rotating assembly.

I wish I could give you a rule of thumb for de-rating the CIM -- something like "subtract half the operating life for every 10 Celsius degrees beyond the rated insulation temperature." But I don't have one, and I don't know who does.

[Kevin Sevcik]

I'll also point out that if you power just these two motors from even a 20Ah 18V battery, you're likely to get about 3-4 minutes out of a fully charged battery. Assuming the battery doesn't end up in a puddle on the floor. A constant 200A draw is a huge load.

[Kevin Sevcik]

Okay, I've got a sort of lead for you, but it'll probably require some leg work. I ran across this (http://www.longrange.net/BattleBots/motor_chart8.xls) list of battlebots motor specs. two motors in there look like a good idea for you. The Dewalt 24V hammerdrill motor with gearbox, and the NPC 64038, which you'd have to gear up to twice the speed.

It says the dewalt costs $85, but I have no idea what the part number for all the stuff is.

EDIT: So I was bored and did the legwork for you. The drill model is DW006K-2. Get an account on www.dewaltservicenet.com (http://www.dewaltservicenet.com) and you can gaze confusedly upon the exploded view of the thing. It looks to me like the parts you'll need will run you from $100-$150 per motor. Depending on how much case you want around it.

[John Gutmann]

Isn't there something you can do to the brushes on the CIMs to give it higher RPMs? Also I recall seeing a DeWalt impact wrench that had 300 ft-lbs of torque on it. That should work.

[jakep]

Hey Kevin, WOW thanks for that awesome spreadsheet! And it's got those numbers and specifications we all love!

For those Dewalt motors/gearboxes:

What do you mean by "the parts you'll need will run you from $100-$150 per motor. Depending on how much case you want around it." ?

I was looking at the ServiceNet link you gave me, the "motor pack" part number is 389010-00 and the transmission part number is 388974-00. Buying one of each costs around $90. For our project, we have a protective aluminum shield around the underside of our robot. (Where the motors, electronics, batteries, etc. are mounted) So, I don't think we need much case around the motors/transmission. Would we be set if we just bought the motor pack and transmission?

[KenWittlief]

One of the limiting factors of electric motor rpm is the centrifugal force on the wires, armature, commutator...

With no load you can keep increasing the voltage, and the motor will spin faster, until you reach the point where the motor rips itself apart.

I think that is the real issue you will face.

Heat dissapation is the second. You can push the HP rating of the motor as long as you back off on the duty cycle, to give it time to cool between application of full power. This would more of less be a matter of trail and error. The failure mode for a motor that overheats is usually the wires melting/unsoldering themselves from the commutator, which is not easy to do since most motors are soldered with silver solder, and the wire is hooked or twisted to give it extra mechanical strength.

right in the same range of abuse, the varnish on the wires will start to cook off from the heat, making a wonderfull pungent scent that all FIRST teams have experienced at some point. If the varnish burns through and wire shorts to wire, then the motor turns into a space heater, and you get those wonderfull clouds of white smoke spewing like Mt Saint Helens.

[Kevin Sevcik]

Hey Kevin, WOW thanks for that awesome spreadsheet! And it's got those numbers and specifications we all love!

For those Dewalt motors/gearboxes:

What do you mean by "the parts you'll need will run you from $100-$150 per motor. Depending on how much case you want around it." ?

I was looking at the ServiceNet link you gave me, the "motor pack" part number is 389010-00 and the transmission part number is 388974-00. Buying one of each costs around $90. For our project, we have a protective aluminum shield around the underside of our robot. (Where the motors, electronics, batteries, etc. are mounted) So, I don't think we need much case around the motors/transmission. Would we be set if we just bought the motor pack and transmission?

Have you looked at the Nothing But Dewalts (http://www.chiefdelphi.com/media/papers/1592) white paper? You can see a bit of how complicated these Dewalt trannies can be. I suspect you need more than just the motor and tranny to make that work. Specifically, I think you need some of the plastic case parts to screw things into and other parts to hook up a shaft to.

So, I think you also need parts 17, 18, 19, 38, and maybe 24. 18, the gear case, would hold everything together. 38 is a nice round-ish spindle to use for a shaft. 17 and 19 almost certainly engage the spindle with what's probably an odd looking asymmetrical output on the tranny itself. 24 is a spring that probably helps the clutch work, but you can probably simply weld/glue and 19 together instead.

EDIT: oh. and 29. A nice left-hand screw to go into the spindle. And 22,to hold the spindle in. So all that works out to $125. I'd personally want the case as well to give it more support and some of the screws for it. That works out to $154 or so.

[jakep]

Thanks for the quick reply! ,

I looked through that whitepaper briefly, but it looks as if it's meant more for CIMs and FIRST motors hooked up to dewalt gearboxes. If we are using everything dewalt, won't it be easier? Also, would there be much machining to do? We don't exactly have easy access to a machine shop :(

"I'd personally want the case as well to give it more support and some of the screws for it. That works out to $154 or so." - The case is part number 1 right, but what about the other screws?

EDIT: Nevermind, I think the screws are part 6, correct?

[sanddrag]

Is there a specific reason you just can't use 4 CIMs (at 12V) to power this project?

[Kevin Sevcik]

Thanks for the quick reply! ,

I looked through that whitepaper briefly, but it looks as if it's meant more for CIMs and FIRST motors hooked up to dewalt gearboxes. If we are using everything dewalt, won't it be easier? Also, would there be much machining to do? We don't exactly have easy access to a machine shop :(

"I'd personally want the case as well to give it more support and some of the screws for it. That works out to $154 or so." - The case is part number 1 right, but what about the other screws?

Other screws are numbers 6 and 27. 3 of #6, 4 of #27. You also need 2 of #22.

I know the NBD won't work for you, I was just referring you to that so you could get an idea of the innards of one of these lovely trannies. If you're going with all Dewalt stuff, the only machining necessary would be to attach something to the spindle. I'm unsure of the thread on the spindle, but the easiest option would be threading the center of a sprocket with that thread. Then you just screw the sprocket on and put on a washer and put in the left hand screw. Look at the end of the whitepaper to see how it works on those. It should work similarly for this spindle.

[jakep]

There is no easy way to mount 4 CIMS with our gearboxes. Doing so would require redoing the gearboxes completely, so I am searching for perhaps something better.

Edit: Kevin, sorry for my lack of knowledge in this subject, but the only machining required would be to attach something to the output of the gearbox, right? I know on regular drills you unscrew the cap, place a drill bit or something inside, then screw it tight. I'm guessing that strategy wouldn't work for us though, right?

[Andy A.]

The CIM motors are sealed units, meaning that the motor can has no venting and the armature (assumingly) has no fan on it like you might find on a Fisher Price motor. This means that all the heat generated by the motor has to migrate out through the motor can, either by radiation or through the bearings.

Neither is efficient or fast, and I suspect that the heat generated is going to fry the armature long before cooling the can with heat sinks and fans will do you any good. The CIM motors get pretty toasty as it is at 12 volts. As was stated, at double the power output, the motor will need to dissipate 1200 watts. To put that in perspective, an electric space heater I own has a maximum of 1500 watts.

You might consider removing the end cap and drilling out some holes in it. Then place a muffin fan behind the motor blowing into the now perforated end cap. The added airflow might help keep the motor cool while running. It will certainly help speed up cooling it in between run times. Just how much of an improvement is might make is difficult to say, but some air flow is better then none.

Modifying motor parts, especially removing metal, can be a bit hazardous to the motor's health, and I wouldn't recommend that you try it unless you are confident you can do so with out risking damage to the motor.

[Kevin Sevcik]

With some further info from jakep, this isn't necessarily a continuous use application. Atleast, it won't be operating at max power for more than a few seconds. It would hit peak power and past for a bit and then level off and operate somewhere below that. So it might not destroy itself too quickly. Perforating both endcaps and using a muffin fan might be enough to handle the extra heat. Certainly it's a cheaper easier option than the re-engineering and building that'd be necessary to use those dewalts. Though the dewalts would work exceptionally at this and I'd be highly amused to see a pair of hammerdrills in this application.

[jakep]

Hmmm... See the problem is that if we upgraded the voltage we were using with the CIMs, we would have to re engineer our gearboxes with a different gear ratio so that extra speed would translate into more torque. Since we need to redo the gearboxes either way, those Dewalt ones are looking awfully nice :)

The one problem at this point seems to be attaching the output shaft on those to our wheel, which has a key slot-thing already cut inside.

Any ideas? I can get some dimensions for you guys if that would help.

[Al Skierkiewicz]

Very interesting discussion so far...
I would like to add that the output power of a motor is not only a function of input current but also of the available field in the magnet structure. At some point, the magnetic field developed by the electrical structure may exceed the permanent magnets. At this point, no additional speed or power will be available and all excess electrical input is turned to heat. Without some testing at these extremes, we can not determine a point at which this will occur. Please remember that the resistance of the motor windings and brushes are also affected by heat which will change performance. Unfortunately, as nice as this motor is for a two minute match, it is not designed for anything approaching continuos duty, included are heat dissipation, brush life and bearings.

[jakep]

'Unfortunately, as nice as this motor is for a two minute match, it is not designed for anything approaching continuos duty, included are heat dissipation, brush life and bearings."

Which motor are you talking about, the CIM or the recommended Dewalt?

[Al Skierkiewicz]

'Unfortunately, as nice as this motor is for a two minute match, it is not designed for anything approaching continuos duty, included are heat dissipation, brush life and bearings."

Which motor are you talking about, the CIM or the recommended Dewalt?

Chalupa, of course.

[jakep]

Hey Kevin, or other mechanically inclined person,

Could we use one of these (referenced fromt he NBD whitepaper) as a motor mount with the 24V Hammerdrill motors?

http://www.chiefdelphi.com/media/img/412/412b176a09ff4026a52e915de988043e_l.jpg

[sanddrag]

There is no easy way to mount 4 CIMS with our gearboxes. Doing so would require redoing the gearboxes completely, so I am searching for perhaps something better.

Edit: Kevin, sorry for my lack of knowledge in this subject, but the only machining required would be to attach something to the output of the gearbox, right? I know on regular drills you unscrew the cap, place a drill bit or something inside, then screw it tight. I'm guessing that strategy wouldn't work for us though, right?

What are your gearboxes? I thought you wanted to buy Dewalts but this makes it sound like you have something already. Do you have something already?

Using a stock drill chuck probably wouldn't work because you couldn't get it tight enough. It would slip. To adapt a drill motor to a robot, you could take off the chuck by removing the left hand screw you'll whole chuck will unscrew from the gearbox output shaft. You'll be left with a threaded shaft. On the lower voltage XRPs, it is 1/2-20. I'm not sure if it is the same on the 24V model. So then, you just use a tap (preferably set up in a lathe or a mill) to put internal threads in the bore of your sprocket/gear/flex coupling/etc so you can screw it on the gearbox output shaft. Then you put the left handed screw back in the gearbox output shaft with a nice thick washer and that keeps the sprocket/gear/etc from unscrewing itself. It is all explained in the NBD whitepaper.

[jakep]

sanddrag, I was talking about the 24V hammerdrill motors that Kevin Sevcik was suggesting. We would be buying the Dewalt motor pack, transmission and some other Misc. parts from the DewaltServiceNet. The part we are trying to work out now is a bearing system or something to handle the side loads. Also, we will be doing a shaft drive system, no chains.

[Kevin Sevcik]

Sorry, detained for the weekend. That kind of thing will work, but I'm not 100% sure the size for the NDB trannies will work for those trannies. But you should certainly be able to get some clamps like that that are the proper size once you measure the diameter on the hammerdrill.

As for mounting, etc. I still don't think you'll be able to get a direct drive to work. I think you'll want to get atleast one of those in hand or find one at a Home Depot or tool store and make sure that if you put them back-to-back that they'll still fit on your base.

That said, if you're bound and determined, you could drill and tap the end of a shaft, lock-tite it onto the spindle, drill it and roll pin it, and maybe even weld it to the spindle as well for good measure.

Yes, I'm serious. This thing is going to put out a lot of torque, and I think welding might be your only chance of getting a direct shaft connection that can handle it. The problem is that your direct shaft means you can't use the left-handed screw to efficiently hold something down and keep it from unthreading. Atleast I don't see any easy way.

[sanddrag]

Support your shaft with your wheel or whatever it is on its own. Put half of a flex coupling (lovejoy makes good ones) on the end of it with a keyway and key for transfering torque and perhaps a set screw to prevent axial movement. Put the other half of a flex coupling threaded onto the drill gearbox output shaft. Hold it on with a left hand screw and washer. Put a flex copuling spider (insert) in between. Put the two together. Mount the drill.

[jakep]

Ahh, I didn't think this sort of thing would be easy. The problem is, a chain drive would be too laggy switching directions to ever work for us. Direct drive has to be the way to do it.

So, let me get this straight:
The mounting-clamp-thing from the NBD whitepaper will support the side loads right? I hear that side loads are bad for gearboxes, so this will protect it right? So, once we get our 5/8in shaft somehow attached to the gearbox (perhaps the way you mentioned it) it should work right?

Thanks also for your suggestions, sanddrag.

[Kevin Sevcik]

sanddrag's shaft mounting should work brilliantly. I should've thought of that. at any rate, the mounting collar doesn't particularly support any side-loads. All of the load will still be supported by the shaft. With the flexible coupling, the shaft out of the drill won't support anything at all. You'll need to support the shaft in your wheel with a pair of bearing pillow blocks on the other side of the flexible coupling.

[jakep]

So what exactly is a flexible coupling and coupling spider? What function does it perform?

Bearing pillow blocks are the blocks of aluminum that have a bearing in the middle right? And then you mount them to your frame and the shaft goes through the bearing, right? This supports the load then, I assume.

You say that "With the flexible coupling, the shaft out of the drill won't support anything at all.", but then why do you still need the bearing blocks?

[Kevin Sevcik]

Here's a link to some on McMaster Carr (http://www.mcmaster.com/ctlg/DisplCtlgPage.asp?ReqTyp=CATALOG&CtlgPgNbr=1077&RelatedCtlgPgs=82,89,90,987,1077,1078,1079,1080,10 81,1082,1083&term=Flexible+Couplings&sesnextrep=622778393191956&ScreenWidth=1600&McMMainWidth=1393)

Drill shaft goes in one side, wheel shaft goes in the other. As you can see, the coupling istelf is only held together if it's pushed from both sides. If you try to bend it or anything it'll just fall apart. So. You mount the drill with that clamp on one side, and then you'll need two bearings supporting the wheel's shaft on the other side.

[jakep]

Ahh, so it's sort of a protection mechanism?

[Kevin Sevcik]

That and you can use it to more easily couple your direct drive shaft. You can thread one half the coupling and screw it on to the drill, and then use the left hand screw to hold it on. The spider means you can still easily couple it to the other half of the coupling and your wheel shaft.