Motor/Controller PROBLEMS

[wdenny]

Hey guys,

This is my first post, so easy on me if I missed something about etiquet. I am a senior in electrical engineering and my senior design project is due on Tuesday of next week. My project was to build a CNC mill. Soup to nuts... (start to finish). I have systems working independantly, but I am having major issues when I start hooking things up together. Here are my problems and data.

I am using a NEMA23 Motor from Anahiem Automation listed below.
http://www.anaheimautomation.com/man...ec%20Sheet.pdf

It isn't listed on the sheet, but the tech guy said 3.3V at 2A in Unipolar mode which is what I am running.

In order to get them to pull even close to 2A I am running them at 9V and everything seems to be working. EXCEPT!!! They will only spin at about 37.5rmp. This is WAY WAY to slow when using a polling controller like the one I am using, and I am having trouble with the controller missing pulses because of how long it takes to execute a step. I tried speeding it up, and the motor just jerks around like it isn't switching right, which means one of two things:

1. The motor can't spin that fast
2. My switching Mosfets aren't opening and closing fully

But, my FETs say they have rise and fall times in the nS range, so I doubt that is the case. But I just can't believe that this motor will only spin at 40rpm, am I missing something here?

As for the controller, we originally used BS2 (basicstamp 2) for each axis independantly, but, later found out that they would only run at about 4000 instructions per second, which means, after you slow that down for motor stepping, you are back to polling at the speed of smell again. So we switched to the Motorola 68HC12 and haven't gotten it done enough to test it yet, but still don't know how we are getting ahead if we can only step at 40rpm...

Somebody please give me some heads up here on what I could possible be doing wrong?

I have schematics and code and diagrams for everything if it would help...

[pminmo]

Post your schematics. What are you doing for current limiting? Your problem maybe compound, maybe to low of supply voltage. A 3-1 ratio isn't much for performance. maybe some info on my website might help.

[kreutz]

Nothing is possibly wrong with your motor or Mosfets, what you are experimenting is the result of the L/R time constant of the motor coils, in order to overcome that problem you will need to use a higher voltage power supply and a method of regulating current on the coils. the 68HC12 is perfectly capable of producing the pulses to drive any Stepper motor in the thousands of rpms, the rest is just mosfet drive-mosfet combination.

I always recommend this reading as basic: http://www.cs.uiowa.edu/~jones/step/

Then, if you want, take at look at this thread for a unipolar motor micro-stepper design : unipolar motor microstepping

There are a lot of simpler unipolar stepper drive designs (full and half step) at Pminmo's site: http://pminmo.com/

and, you can see what kind of speeds you can get, in this video: http://youtube.com/watch?v=y6BS4B3yT...elated&search=

[pminmo]

Kreutz,
Did you ever video the setup using G-Code that would be starting and stopping, changing directions, etc?

[SORCHEROR]

i would start first with a 24 volts supply and get some resistors to limit the current,and than it will perform alot better

[kreutz]

Kreutz,
Did you ever video the setup using G-Code that would be starting and stopping, changing directions, etc?

Yes I did, there are 3 videos, the first one is about high speed tests, the second includes low speed tests with reverse movement at different speeds increasing through the midrange and high speeds, the third video is about micro-step angular resolution. All of them are posted at youtube: http://youtube.com/watch?v=y6BS4B3yT...elated&search=

Look under the "related" column in the same page for the other tests.

[wdenny]

Well, the panic is over for now. BUT...

I think you are right about the performance issues. After about 12 cups of coffee with some Ten High kickers, the 68HC12 is finally working the motors. What was the main problem? The 14 year old answer the phone in tech support at the co. that sold me the motors.... gave me the wrong step sequence. Second, we were just trying to go TOO fast for our power supply. We set it up to draw 2A at locked rotor as that was the MAX current rating for the motors. However, this only required 9V. I do however see your point about stepping the voltage up for better L/R performance. And I plan on doing that, AFTER a demo and finals. Working slow still gets me an A.

Right now, our MAX speed is 90 rpm at incredible torque. With a 30 inch X axis however, is still sub par for something I was hoping to use a lot. But since it is due on Tues, there is now way to order new regulators and test by that date.

As for my driver, it looks almost exactly like the one on the site you suggested... http://pminmo.com/3axis/3axis.htm

My biggest concern now is a software/design issue. I have Autocad2002, Ace Converter, Mastercam9, and then the CNCsimulator. I can make stuff to cut out, but I have no idea how to get it to cut anything but the actual lines on the objects? For example, if i were to attempt a cone, when the tool went to cut the bottom circle, it would cut the the entire hight of the cone on that pass. I am fluent in all of the programs above except for mastercam and I am guessing that is what I need to use.

Anyway, any help on the software issue is appreciated, and the suggestions on the driver/controller will be implemented asap after finals and stuff in the next couple weeks.

[pminmo]

Yes I did, there are 3 videos, the first one is about high speed tests, the second includes low speed tests with reverse movement at different speeds increasing through the midrange and high speeds, the third video is about micro-step angular resolution. All of them are posted at youtube: http://youtube.com/watch?v=y6BS4B3yT...elated&search=

Look under the "related" column in the same page for the other tests.

Have you instrumented in some way to know your step count relative to motor position are correct, specifically in the second video?

[kreutz]

Have you instrumented in some way to know your step count relative to motor position are correct, specifically in the second video?

I know that a scientific proof that I did not lose any step would include an encoder and counter attached to the motor axis and making sure the final count is zero (or +/- 1 count due to counter uncertainty) after returning to the starting position. I did not do that, but just looking at the starting and end positions I assumed I did not lose any. Anyways, I will attach an encoder and counter as soon as I have some time left just to be 100% sure, and document the test.

Thanks,

Kreutz.

[pminmo]

No need, I was just wondering as in one sequence, it sounded like on a quick slow down a grinding sound. Looking good !!

[kreutz]

No need, I was just wondering as in one sequence, it sounded like on a quick slow down a grinding sound. Looking good !!

It did a direction reversal in that moment. I noticed the sound too. For some time I have been collecting information in order to build an small dynamometer for the tests. Since I don't have anybody around with a CNC machine it will have to wait a little longer until I have the time to build my own CNC mill. I would like to see the test results from a dyno.