Controllable torque electric motor questions

Perhaps a more relevant point is that if nothing is moving, nothing useful gets done in the context of the OP's inquiry. He cited 0 to 10 RPM at the load as the range of interest. For zero speed he could skip the motor and gears, just hang a weight on a lever. If he wanted to apply and release that torque, then the weight might hang on a rope and be lifted by a pneumatic or hydraulic cylinder.

Might you be a consultant?

Mr Patmore may well have been a genius, but this is a straightforward (classic even) closed loop control with torque feedback. That said, it is elegant in its simplicity and economy.

Yes, as Xeno taught us.

I use motors a lot in the shop. I just took a quick look at the link you posted. Thanks a bunch. Eric R Snow

No. Not unless you're looking at the torque developed by the armature of the motor, which is a singularly useless place to be looking when what you care about is the output of a 100:1 gear box.

At the output of the gearbox, the torque will be

torque = (motor torque) * (gear ratio) - (gearbox friction)

No. Not unless you're looking at the force developed at the face of the piston, which is a singularly useless place to be looking when what you care about is the force developed by the cylinder assembly.

At the output of your hypothetical air cylinder the force will be

force = area x pressure - friction.

If friction is much bigger than area x pressure then you have to regard it a great deal, or resign yourself to poorly working machinery.

To the OP:

Sorry for the confusion. This is one of those special USENET moments when you have to read the posts, think a bit, and figure out who's right. Have fun.

And the friction vanishes as the system slows. The ultimate torque as you approach a stall is not affected by friction.

Friction is a monotonically increasing function of velocity, and zero at zero velocity. So a moving air cylinder pushing against a monotonically increasing force will develop an ultimate force = area x pressure when it comes to a stall from equilibrated forces, regardless of friction. The friction only affects the *rate of approach* to the final force, not the magnitude of final force. Similarly for torque in the rotational case.

I would say friction bigger than area x pressure is going to amount to locked brakes.

To the extent you have stiction, that matters.

It may help to think of it intuitively in terms of a more perfectly frictional system, such as a magnetic induction brake, rather than gears that have a combination of friction and stiction combining through gear ratios.

You can also think of it as a restriction in an air line. It slows the delivery of downstream air volume, and it lowers the downstream pressure under flow, but in the no-flow case you still have full pressure at the terminal.

That was one of his claims to fame. Anyone could design a textbook circuit. He could design something that worked using a bent paper clip and a marble. I had known him for years before I knew he held a patent. I had had a chassis bent up in the sheet metal shop and had set it up in the lab when he walked by. In passing, he said "I have a patent on that". I said "you can't have a patent on a three sided box". He crooked his finger. I followed him back to his cubicle. He opened his file drawer and pulled out a stack of patents about an inch and a half thick. He shuffled through them and pulled out a patent on a three sided box. I picked up the stack and started going through them. The first electronic digital voltmeter (Non-Linear System's was electromechanical with a stepping relay, Jim's used triode flip-flops). The complementary symmetry output stage ("using a n-type transistor and a p-type transistor" says the patent). The shift register keyboard debounce circuit. Jim could do anything. He was a pilot. Scuba diver. Dirt bike racer. Sang in a barbershop quartet. Kept race horses. Had a lathe and made model airplane engines. I was a kid in the lab and he took care of me. I could learn more working for him than I could in any school.

Kevin Gallimore

Thanks to everyone who responded. An interesting and educational discussion.

I think that I understand the issues of friction and stiction. ;-) I will side with those who think that friction and stiction will reduce the delivered torque, even at zero speed. As long as the friction and stiction was fairly consistent, it should not be a big concern. However, a prudent designer might plan ahead about how a torque sensor could be installed if more precision was required.

I like the idea of a DC permanent magnet motor rated for stall use. Apparently these types of motors are commonly used as servomotors. Regulating the current to control the torque makes sense, and sounds fairly straightforward, with an appropriate programmable power supply. Looking at the curve on the Baldor website, it appears that torque is proportional to current, while RPM is a function of current (torque) and voltage. I noticed that someone said that temperature was also a factor, although I did not see it on the Baldor site, but perhaps that effect could be minimized by an oversized motor or other conservative design.

I can understand that twisting square stock by machine is usually done cold, but this application is a little different, and does not involve square stock. The process already exists, albeit in a manual form, the goal is just to automate it.

The more I think about this project, I realize that a good controls designer is key. I could probably handle the mechanical design and overall system design, working with the guy who developed the process, but obviously I am not up on servomotors, programmable power supplies, PLCs, etc. The project gets bigger the more I look at it. ;-) I may just offer a bit of free advice and bow out.

Thanks again, I learned some stuff.

Richard

D>

Friction (Newtonian) is independent of velocity and is non-zero at zero velocity.

Viscous drag is a monotonically increasing function of velocity and is zero at zero velocity.

If it were otherwise, nuts and bolts would not be of much use :-)

Mark Rand RTFM

Love that HTML. :-) ...lew...