DC Motor drive

[duke37]

I have a very old AVO coil winder which I have fitted with a 24V DC motor. I need a variable speed and have achieved this with a Variac and 12V battery charger. This works OK but is bulky and I need my Variac for other duties. The motor takes 400mA unloaded at any speed. The current will be somewhat more when driving the winder and could be considerably more when winding thick wire.

What would be the best way of driving the motor?
1. Variable stabilised DC voltage. (LM2596)
2. PWM with no feedback. (555)
3. Linear voltage supply.

I have bought a buck convertor board based on a LM2596 (2Amp?) and will modify this to fit an external variable control.
I repaired a CB radio power supply modified for variable output but this has excess reservoir voltage and gets too hot even when the motor is unloaded.

On a separate topic, I need a couple of half nuts for the lead screws. Half inch diameter, 40tpi, buttress thread. Any idea of where to get these. I have a ordinary thread tap and will get a tame model engineer to make me two half nuts and see whether these will do.

 

[Harald Kapp]

Variable stabilised DC voltage. (LM2596)

O.K. as long as the max. current is not exceeded. Being switch mode this regulator will not dissipate much power as heat. Note that you have no regulation of the motor speed. Constant voltage does not ensure constant speed. Speed will vary with the load.

PWM with no feedback. (555)

For this application this is comparable to the LM2596, having no fedback for speed regulation either.

An advantage of the PWM over the switch mode DC is that the motor can develop more torque at low speeds due to the high voltage and current during the on time of the PWM. This may come in handy especially when winding thick wire.

If you want/need speed regulation, have a microcontroller generate the PWM and add feedback e.g. by using a tachometer generator or by evaluating the flyback voltage of the motor during the PWM's off time.

 

[(*steve*)]

I would tend to go for PWM because it's going to give you the best torque at low speeds.

 

[hevans1944]

On a separate topic, I need a couple of half nuts for the lead screws. Half inch diameter, 40tpi, buttress thread. Any idea of where to get these.

You could check with this company.

 

[duke37]

I did try a feedback circuit. This used a rectified but unsmoothed supply. A SCR was activated when the motor voltage was low, giving a pulse if necessary 100 times per second.This was used on a 12V Ford fan motor with strong cogging. It managed to run the motor well at very low speed but I did not pursue as there was a possibility of DC through the transformer causing saturation.

 

[duke37]

You could check with this company.

I contacted Barnes but they did not have tools to make this thread. Thank you for the link.

 

[hevans1944]

I contacted Barnes but they did not have tools to make this thread. Thank you for the link.

Well, maybe this company is on the other side of the pond from you has a hand-tap you could use. Lots of useful information on their website too. You could drill and tap a hunk of brass or bronze and then saw it in half to make a split-nut for your lead screw. Or get your tame model engineer to do it. Some additional machining required.

For a buttress thread, it might be cheaper to have a local machine shop make a thread-cutting tool with the proper profile, and then cut the thread on an engine lathe instead of trying to find a hand tap. Sounds expensive, either way. Can you replace the lead screw with something else? How accurate does it have to be for winding coils?

 

[duke37]

Thank you for the link to North American Tools, I see they have a distributor in Derby, I will give them a ring on Monday.I have purchased (cheap) a 0.5 inch diameter, 40 tpi tap with standard thread and will get a friend to make a nut to cut in half. Perhaps the standard thread form will give sufficient bite. The carriage will run to the right but is loathe to run to the left on the second lead screw. On the last coil I did of 30000 turns for a slave clock solenoid, I had to help it.

The accuracy will depend on the lead screw and the speed that the screw turns relative to the coil. The speed ratio is set with a rubber pulley set to a radius on a rotating flat plate. I set it to traverse at slightly more than the wire diameter (4 thou).

 

[hevans1944]

... The accuracy will depend on the lead screw and the speed that the screw turns relative to the coil. The speed ratio is set with a rubber pulley set to a radius on a rotating flat plate. I set it to traverse at slightly more than the wire diameter (4 thou).

Wow! That's some coil winder! If you can't get the carriage carrying the wire to run in both directions, you could still wind some boss Tesla coil secondaries.

I saw on the North American Tools blurb page that they are supposed to sell buttress taps and dies, but I couldn't find any on their catalog pages. Probably a special-order item, especially in 1/2-40 thread.

I like the variable-radius rubber pulley-on-disk arrangement for pitch control. Haven't seen anything like that in maybe fifty years. I think my dad's Air Force sextant had a similar arrangement for integrating the altitude angle of whatever star he was shooting, a variation of the old ball-and-plate mechanical integrator. You had to wind his sextant up and let it run for about two minutes while using a knob on the side to move an internal prism to keep the star centered on the bubble level. Stars "move" a lot in two minutes! Dad wouldn't allow me to take his sextant apart to "see how it worked" but he did bring home a tech order manual that gave a simplified explanation of the mechanical integrating mechanism. Pretty neat gadget.

Does your winding machine do low-capacitance "pancake" coils like you sometimes see on RF chokes? Why do you even have this ancient machine?

 

[duke37]

The 1/2-40 thread is designated ME or model engineering but of course is a normal thread form. I managed to get a tap for about £5; one company will make a buttress tap for £100 which is over the top.

The half nuts need to be fitted the right way round but my eyes are not good enough to be sure if they are right. There is a dint on the lead screw but I think I have rubbed this out.

Pancake coils as I understand it are wound on a flat former with an odd number of slots to make a coil not much more thickness than the wire diameter. These were used on early radios and were moved to alter the coupling for regenerative detectors. The machine will not do anyrhing like that.

The capacitance reduction you mention is done by wave winding, The machine will not do this but I have seen stepper motors suggested to do this. Cotton covered wire was used to keep the wire in place, it is not easy to find.

I made some remarks when I saw the machine on ebay and surprise, surprise it was delivered as a present. The intention was to rewind transformers for valve radios but the first one is still here years later as other jobs and a broken back have taken precedent. I made some missing parts out of wood but a neighbour has converted some to aluminium. He is too busy to make me some half nuts.
I have made some transformers for 12V to 400V for radios using TV line output transformer cores (1V/turn at 10kHz) and some coils for an audio pickup. I friend looks after the clocks on a preservation railway and needs some solenoids rewound for the slave clocks. I have borrowed his valve tester. I have also wound an inductance for the low tension ignition for a 1902 petrol engine. Another job which has not been completed is an invertor for an old electric fencer, it will need to be efficient to run off batteries of 6V to 12V.

The next project is to get a variable speed motor drive in a small neat box, then add some counters to count the turns taken off a transformer as well as those put on. The machine was solid when I had it and I had to thump the balls in a bearing to free from the rust.

 

[eKretz]

I was a machinist for almost 20 years before I blew my back out and the docs were unable to fix it well enough to get me back to work. I might be able to find a tap for you, give me some time to look around. No promises, but there are a LOT of old oddball taps where I last worked (I still have a key to the shop so I can work on little things if I feel up to it).

 

[duke37]

I was a machinist for almost 20 years before I blew my back out and the docs were unable to fix it well enough to get me back to work. I might be able to find a tap for you, give me some time to look around. No promises, but there are a LOT of old oddball taps where I last worked (I still have a key to the shop so I can work on little things if I feel up to it).

I would appreciate any help here. I have a 1/2-40 tap and and will see a friend on tuesday, he may be able to get a couple of half nuts made with standard thread form. Perhaps there will be enough bite to work, obviously the correct waveform will be much better.

 

[eKretz]

I will take a look as soon as I can. I personally wouldn't try making a half-nut with a standard thread form though - it wouldn't close, and would probably damage the lead screw if it was used for very long.

It's actually not very hard to make a tap that will hold up in brass by using a piece of drill rod or even any old tool steel (I assume you will be making it from brass, right?) Just machine the rod to the same dimensions as the lead screw and then cut some grooves like on a tap. Heat it up with a torch until a magnet no longer sticks to the steel and quench it in oil while swishing it around. Polish it up and slowly heat it until the polished area turns straw color (straw like as in hay). Ready for use!

On another note - you're positive that the thread form is buttress? Most lead screws are Acme or square thread. Since you mentioned that your machine has two lead screws though, the maker may have used buttress in opposite hands for opposite directions - that was more common in the very old days.

 

[skenn_ie]

eBay has loads of speed controlers. I suggest though, get a couple of Arduinos and learn a bot of programming while you use them.

 

[daberbaber]

I have a very old AVO coil winder which I have fitted with a 24V DC motor. I need a variable speed and have achieved this with a Variac and 12V battery charger. This works OK but is bulky and I need my Variac for other duties. The motor takes 400mA unloaded at any speed. The current will be somewhat more when driving the winder and could be considerably more when winding thick wire.

I know this goes against convention wisdom but I would start with a light dimmer, maybe a commercial version, to replace the Variac and then replace the battery charger with a step down transformer, possibly one used to run furnace controls. The VA rating would have to be observed for the secondary. The inductance may fry lesser dimmers but may not be a problem with commercial ones since they are designed for longer wire lengths and might have some tolerance. You can then employ a simple bridge rectifier with some filtering as necessary. You won't have much torque tolerance like PWM would provide though.

 

[duke37]

eBay has loads of speed controlers. I suggest though, get a couple of Arduinos and learn a bot of programming while you use them.

I am not going to learn a lot of programming. I last programmed a 3502 in an Apple ][ to drive a 250kW DC motor about 50 years ago, not starting again now. I did learn that correct spelling was mandatory.

A dimmer needs to be symmetrical if driving a transformer. My previous trial of a SCR (with feedback) worked well but I did not go ahead because of possible saturation of the transformer core if the output is not symmetrical.
I have bought a PWM module and a switched mode buck convertor (no feedback) to compare them. One of these has a twiddle knob so easier to use. Accurate speed control is not required.

Buttress thread.
There are two lead screws,both right hand thread, rotating in opposite directions with a half nut for each. I think one half nut was fitted the wrong way round so did not engage properly and there were particles of brass where one lead screw had a dint.

 

[daberbaber]

A dimmer needs to be symmetrical if driving a transformer. My previous trial of a SCR (with feedback) worked well but I did not go ahead because of possible saturation of the transformer core if the output is not symmetrical.

Can you assume that the triac used in an AC dimmer, not SCR, puts out symmetrical pulses on both sides of the AC? If so, then maybe a simple RC smoothing filter can shape them into somewhat of a sine wave?

 

[duke37]

An AC dimmer is triggered with a diac. Some are more symmetrical than others. You do not need to smooth any output to drive a motor.

 

[daberbaber]

I have never heard of a diac, just triac and SCR. I will investigate though. My concern for smoothing was more for the step down transformer from 110 or 220VAC mains to the 24VAC or lower feeding the bridge rectifier for the DC requirement.

 

[hevans1944]

I remember diacs. Not the most ideal component for symmetrical triggering, plus it depends on an RC network to determine where on the phase it triggers.

You could really complicate this and (probably) have a lot of fun doing so: Phase-lock a high-frequency (20 kHz) oscillator to the line frequency; use a zero-crossing detector to reset an 8-bit counter that counts up clock pulses from the oscillator; trigger back-to-back SCRs or a TRIAC when the count reaches the desired number on each half-cycle. That would be pretty damned symmetrical and a good exercise in logic.

However, as interesting as this might sound for a retro digital design, I think a modern PWM chip driving a MOSFET H-bridge operating off rectified line voltage might be more cost effective. Most of these chips have a current feedback input that allows you to set the motor current with an analog or, perhaps, digital command. And the ones I have seen also boast a charge-pump circuit to drive the high-side MOSFET gate. And, being an H-bridge, the motor direction is reversible. These circuits are especially adaptable to driving bi-polar stepper motors.