My Design for a home made Pendant

[Steve-Tee]

Hi folks,

I have designed and built a simple Pendant for use with the CNC machine I am building.
It uses discrete parts, so as to make for easy testing during development.
A stepper motor is used as a carrier for the jog wheel and the optical encoder disc. The stepper motor has another use too, it generates a very pleasant degree of tactile feedback when turning the jog wheel.
I thought I would include my circuit diagram here, so folks could take a look and suggest possible improvements / weaknesses.

There are pictures of the completed device in the photo callery.
It's quite a small unit, so it fits nicely in my hand.
I need to use both hands to operate it. This helps to keep my digits clear of the moving machine parts .

The 4 axes can be driven at the same time, but will all go in the same direction - since there is only one direction control coming from the Pendant.
Single axes can be selected by pressing the appropriate switch while delivering pulses from the 2 possible onboard sources.
1) A variable oscillator provides 'auto pulse' mode (200Hz to 6000Hz),
2) An optical MPG provide 'manual jog pulses' when the jog wheel is turned.

The MPG has a velocity control switch to select between full jog speed and 1/64th of full jog speed.
This comes in handy when making final adjustments to the positions of the various axes.

Direction control comes from the jog wheel. Its last movement dictates the direction in which the axes move when driven in 'auto pulse' mode.

When the Pendant is plugged into the mill, the computer detects its presence and halts the program, if it was running.
This function is afforded by the 'Computer lockout' pin on the connector. Its purpose is to prevent the two sources of motor pulses from interfering with each other.

The Pendant has to be unplugged before the computer can resume control of the machine

This design is not perfect, as it is possible for the 'auto pulses' to be delivered while the 'manual pulses' are being generated.
A bit of extra circuitry could have fixed this, but I figured it wouldn't be too bad .

If I were to do this again, I am sure I would opt for a PIC Micro.
This would save on parts and make for a much more versatile device.

Best wishes,
Steve-tee.

[neilw20]

Ever tried using the output from a small stepper motor as an MPG.

Shouldn't take much to clean up the signal and create suitable pulses.
Picking the right size resistors as a load on each phase can adjust the 'feel'.

[Steve-Tee]

Howdy Neilw20,
Yes, stepper motors do work as MPGs. Not only this, but they also generate a convenient quadrature signal.
This can be obtained from any one of the other coil phases in the motor.
At first, I thought I had found the winning combination, as the signals being generated seemed to be perfect.
Sadly, it all fell over when I tried to turn the stepper motor at very low speeds.
When I did this, the signal amplitude became too low to be readily detected.
Perhaps I should have more carefully amplified those weak signals for digital detection.
Hmmm, better yet, I could have used a differential amplifier to operate on opposing coil phases.
Ok, that experiment was done 20 years ago, well before I needed such a device for a CNC application.
Crikey, I didn't know what CNC meant until a year ago .
Now, I am a CNC addict

Considering that Pendants have velocity selectors in them, perhaps it's not such a problem that one can't use a stepper motor at very low speeds. After all, one can turn the dial at moderate speeds and use the velocity selector to slow the stream of stepping pulses.
Well, it looks like you have given me the needed impetus to reinvestigate this possibility .

It ought to be a lot easier to build than this Pendant was, so it's worth another look.

Best wishes,
Steve-tee.

[neilw20]

Try some of the low voltage CMOS families.
15K Resistors on the inputs, and a supply of about 1v, then a level translator.
We must stay digital (LOL).
Or, maybe just drive (through resistors) straight into the base of a transistor
that is almost biased on. A current mirror comes to mind. That will be good over a wide temperature range if you use a dual transistor in a package.
Make sure they get a couple of emitter resistors to improve balance.
A stepper motor designed for higher voltage should generate a bit more at low speeds too.
All is not lost.

[Steve-Tee]

Hehehhe, ahhh another convert to the great ways of the 'Digital world' .

Yes, I am sure I could get some great results by redoing those experiments. Hmmm, now where's that little pile of broken floppy drives?

Originally Posted by neilw20 Try some of the low voltage CMOS families. 15K Resistors on the inputs, and a supply of about 1v, then a level translator. We must stay digital (LOL).

[flyboy1015]

Steve,

I'm with Nielw20 on the stepper MPG. I think it's a great idea, however, I don't have any experience with it, so I probably have an idealistic idea of how it should work. I'm a mechanical engineer struggling to work in an electrical environment. I understand the concepts very well, but the implementation is something else.

Anyway, here is my idea: Use the stepper to generate pulses that are received by an op-amp. The op-amp is set up as a schmitt trigger. The op-amp is biased such that it will only trigger once the input has reached a certain level, which you can adjust. This may require an amplifier prior to the op-amp as you previously mentioned. The purpose of this whole exercise is to produce a nice, clean, consistent output from the device. There are probably 100 other ways to do this and a lot of them are probably better than this, but at my current level of experience and understanding, I think this is how I would proceed.

I would really like to be able to do this with a picmicro. Unfortunately, I have a minimal understanding of programming in assembly and even less in other programming languages. That being said, the next major hurdle would be communication between the pendant and the computer. I would like the cnc software to interface with the pendant.

I am also currently running turbocnc. I really with it had velocity contouring. I would probably continue using it if it did.

Thanks for your post. This helps me understand more about how to interface and how the machines work.

Chad

[Steve-Tee]

Hi Chad,
Yer pretty much on the mark when it comes to the use of op amps for pulse shaping and amplification.
A lot of op amp packages on the market have 4 devices in them.
Hmmm, how convenient. 2 devices per chosen stepper phase .
The first op amp would be set up as an amplifier, with diode clipped inputs for protection from over speeding of the stepper spindle. This would be most likely to be DC coupled to the motor for best response at very low speeds.
The output from this would then feed another of the 4 opamps which would be set up with positive feedback to make it into a schmitt trigger. this arrangement ought to be quite elegant in its simplicity.
By duplicating this arrangement with the other 2 opamps in the package, perhaps a TL074, we could extract the required quadrature pulse structure from the motor coils.
Once done, those pulse groups are then sent to a 'D type clocked latch'. This device combines the signals to produce a 'Forward' / 'Reverse' direction control signal.
So, with just 2 chips and an old stepper, you have made a pendant .
Now, at this point, a PIC Micro starts to look interesting. As these have analogue to digital convertors in them, it is possible to produce a 'virtual schmitt trigger' in software.
The resulting virtual states are now in the form of data bits in a register.
By deciding which occurs first, and which occurs after, one can determine the direction of rotation and hence produce a direction control signal to accompany the motor clocking pulses.
The same PIC Micro can be used as the autopulse generator, controlled by a potentiometer driving an ADC input.
But why stop there?
An LCD would fit in quite nicely, as PICs love to play with these .
Oh, the ideas are really starting to flow now. Looks like the morning coffee is kicking in .
At the moment, my thinking cap is in the wash. It's really quite hard to get all the little bits of brain off it .
Hmmm, come to think of it, that could be grease from my hair...
Once it's ready to wear again, I'll give this some more considered thought.

Best wishes,
Steve-tee.

[neilw20]

Carefuly done, with a capacitor and a few diodes, you can power an 8 pin PIC direct off the power generated by the stepper.
A some PICs will run for minutes, on a really small capacitor. 10uF 1206 SMD
Stuff a big cap in there, and the number of wires required keeps diminishing. Blue tooth!(LOL)

Put an 3 axis accelerometer chip (5mm x 2mm under usd$5) in the pendant and point it where you want to go!
The invention improves. Hit the pendant on the correct side for rapid. Who needs a knob. Now we have a matchbox size pendant, with no wires.
Battery (now required) will last over 1 year.
Drop it on the floor for emergency stop.

[Steve-Tee]

Howdy Nielw20,

Oh boy, not only are you inventive but jolly funny too .
I must remember not to be drinking coffee while reading your posts, it makes such a mess when one suddenly bursts into laughter.


Best wishes,
Steve-tee.

[neilw20]

Originally Posted by Steve-Tee Howdy Nielw20, Oh boy, not only are you inventive but jolly funny too . I must remember not to be drinking coffee while reading your posts, it makes such a mess when one suddenly bursts into laughter. Best wishes, Steve-tee.

You should read the comments in my C code!
I think I'm coming down with a Lisp. (to generate GCode)

[Steve-Tee]

Hehehehe, thanks for the warning.
I will be sure to keep coffee, and other sticky beverages well clear of the 'dot com machine' while reading your posts

Oh, for a waterproof keyboard .



Hmmm, now I'm off to design some more 'lektronic bits.
I have an idea for an SMPS that will supply the current that the motors will need.
It's an odd design, so it should work oddly? - or perhaps evenly?.
We'll see .

Originally Posted by neilw20 You should read the comments in my C code! I think I'm coming down with a Lisp. (to generate GCode)

[neilw20]

Originally Posted by Steve-Tee Hehehehe, thanks for the warning. I will be sure to keep coffee, and other sticky beverages well clear of the 'dot com machine' while reading your posts Oh, for a waterproof keyboard .

Water Proof Flexible Keyboards about $25 at ROCKBY ELECTRONICS (I think)

Originally Posted by Steve-Tee Hmmm, now I'm off to design some more 'lektronic bits. I have an idea for an SMPS that will supply the current that the motors will need. It's an odd design, so it should work oddly? - or perhaps evenly?. We'll see .