Shop built servo

B

Beone

Active User
Registered
Joined
Aug 4, 2013
Messages
50
I am operating way out of my league here but: there is a thread about using a cordless drill for a power feed, not exactly revolutionary. But part of the discussion was that you can get pwm controllers for less than ten bucks on ebay.
Here is the question/challenge: can we then affix a chopper wheel and a optical switch or two and create a servo? Would it be usable with a gecko type driver?
dave
 
Beone said:
I am operating way out of my league here but: there is a thread about using a cordless drill for a power feed, not exactly revolutionary. But part of the discussion was that you can get pwm controllers for less than ten bucks on ebay.
Here is the question/challenge: can we then affix a chopper wheel and a optical switch or two and create a servo?
Click to expand...

Sure. You'll have to design the control loop and the servo won't be very high quality (low speed cogging, backlash in the gears) but it will work. I have plans to do that very thing to power the knee on my mill (except that I'm building the entire controller and using a "corded" drill).

Would it be usable with a gecko type driver?
dave
Click to expand...

What's that?
 
Well, like a G320x servo drive by geckodrive products
dave
 
Depends on your motivation. You can build one for fun and "hey lookie at what I made." It will cost more than off the shelf and most likely not work as well. The biggest problem will be the encoder. You would want to probably start at about 200 pulse per revolution. Not too bad and you can get an encoder for about $30. Then you will want a dual H-bridge that you can control the individual pulses on. Either something like a Darlington Array or build it from transistors. Then you need some diodes. The biggest problem is that you are not going to be buying stuff in enough bulk to get the prices of components down to reasonable levels.


[video=youtube_share;webaVx_mpC8]http://youtu.be/webaVx_mpC8?list=UUTClOLIPPO4IqRdNt5DJGzQ[/video]
 
David Kirtley said:
Depends on your motivation. You can build one for fun and "hey lookie at what I made." It will cost more than off the shelf and most likely not work as well. The biggest problem will be the encoder. You would want to probably start at about 200 pulse per revolution. Not too bad and you can get an encoder for about $30. Then you will want a dual H-bridge that you can control the individual pulses on. Either something like a Darlington Array or build it from transistors. Then you need some diodes. The biggest problem is that you are not going to be buying stuff in enough bulk to get the prices of components down to reasonable levels.
Click to expand...

Actually I have 90% of what I need on hand, mostly salvaged from junk. That includes several encoders (though I know how to build those). Mechanical components are the hard part, not having a useable lathe.
 
Assuming the motor voltage was in the range the the 320X, you wouldn't need anything other than an encoder...
 
I have an objection to the "your biggest problem will be..." clause stated previously. I posit your biggest problem will be that a drill motor is series wound, and speed is load-dependent. Sure, speed is load-dependent for any motor to a greater or lesser degree, but in the types of motors traditionally used for servos, it is the very lesser degree. Permanent magnet DC motors put out a pretty predictable speed for a given applied voltage and all that's needed is a little trim here or there if there is a heavy load. AC polyphase induction motors only drop a little speed (slip) as load is increased; again, only a little trim needed. AC polyphase permanent magnet motors (BLDC AKA PMAC) have very predictable speed output for given frequency. A series wound motor on the other hand has wildly variable speed output for a range of load conditions. I have given thought to making a down-&-dirty series wound servo before but haven't gotten around to trying it yet so I can't say for sure whether it's possible or not, but my suspicion is that it is possible (it seems theoretically possibke to make ANY motor into a "servo") but would be very challenging to tune and probably wouldn't be precise enough for a CNC application unless there was a great deal of down-gearing from the motor to the load. I predict you will aim for SERVO and achieve "servo."
 
strantor said:
I have an objection to the "your biggest problem will be..." clause stated previously. I posit your biggest problem will be that a drill motor is series wound, and speed is load-dependent. Sure, speed is load-dependent for any motor to a greater or lesser degree, but in the types of motors traditionally used for servos, it is the very lesser degree. Permanent magnet DC motors put out a pretty predictable speed for a given applied voltage and all that's needed is a little trim here or there if there is a heavy load. AC polyphase induction motors only drop a little speed (slip) as load is increased; again, only a little trim needed. AC polyphase permanent magnet motors (BLDC AKA PMAC) have very predictable speed output for given frequency. A series wound motor on the other hand has wildly variable speed output for a range of load conditions. I have given thought to making a down-&-dirty series wound servo before but haven't gotten around to trying it yet so I can't say for sure whether it's possible or not, but my suspicion is that it is possible (it seems theoretically possibke to make ANY motor into a "servo") but would be very challenging to tune and probably wouldn't be precise enough for a CNC application unless there was a great deal of down-gearing from the motor to the load. I predict you will aim for SERVO and achieve "servo."
Click to expand...

The people talking about using drill motors for CNC are using battery drills which use PM DC or BLDC motors. My plan (which I may never complete if I can't scrounge the needed mechanical parts) is to use an AC electric drill to power the knee on my mill. I intend to "servo" it because, as you note, the open loop speed control of series motors is very poor.

There will, of course, be a great deal of "gearing down": the armature of a typical AC drill maxs out at upwards of 20000 RPM.

In a previous life I designed motor speed controls for medical equipment.
 
I think that the "evil experiment" factor is really cool. It sure does make replacement of failed components problematic though.

I would personally be more interested in a home brew a closed loop stepper.
 
You must log in or register to reply here.
Back
Top