DC motor constant speed regardless of load

[ricko_uk]

Hi,
I don't have much experience in DC motors other than basic speed control using PWM but I would like to build a very simple small coil-winding machine consisting using this simple lathe from Amazon: https://www.amazon.co.uk/gp/product/B07HKZPDDD/ref=ox_sc_act_title_1?smid=AJPBV39SXHGIV&psc=1

and I am going to guide the wire by hand.

It uses a 12V DC motor and ideally I would like it to:
- run at very low RPM, maybe even as low as 30 to 60 RPM
- maintain constant speed regardless of load (a very small range in load). The load is just my finger which presses against the coil while it is being wound

From past experience, especially at extremely low RPM, the smallest change in load made the DC motor behave a little erraticaly.

Are there simple low cost solutions/controllers that can achieve both of the two points mentioned above?

Perhaps some back EMF controller or some other technology/system/solution?

I could not find any on Amazon (but perhaps I was searching for the wrong keywords)?

Thank you

[IanB]

Have you looked this other thread on a similar topic? There were various ideas posted there:

https://www.eevblog.com/forum/projects/brushed-dc-motor-closed-loop-speed-control/

[Benta]

Not worth the trouble, and you'll never get it down to that speed... well, you might, but it'll be so jerky that it's useless. 60 rpm? That's one rev per second with a normal 3-winding, 2-magnet DC motor. No way.
Make a stand for your battery screw machine instead.

[IanB]

Well, yes, that machine is designed to run at 5000 rpm. To get low speeds, you are going to want a gearbox or other reduction gear. With a reduction gear the torque will be so high there will be no significant speed variation with load.

As Benta suggested, look at using a cordless drill or screwdriver with speed control on the hand trigger.

[ricko_uk]

Thank you all,
yes it looks like the cordless screwdriver is a much better solution. 

Any suggestion on a practical way to connect a chuck or other coil-supporting means to the hex screwdriver head hole/slot?

I thought of a chuck with an hex shaft but have only found relatively small chucks (max 6.5 mm drill bit).

Thank you

[IanB]

Well, cordless drills come with a chuck. Cordless screwdrivers not so much.

Just get a cheap cordless drill from Lidl or Aldi or some such. Or get a more expensive one if you also want to use it for drilling.

[ricko_uk]

   Sorry... in the rush I wrote drill but meant cordless screwdriver (I now corrected it).    And when I read the previous post in my mind I "read" screwdriver instead of drill.   

I think it might be time for me to have a break and relax a bit...

Yes a low speed drill would work perfectly. Disregard my previous silly post...!!!

Thank you all again!

[Siwastaja]

Position sensing e.g. by optical encoder, fast motor current control loop, and very finely tuned speed control feedback loop. And still getting it run stable at such low RPM is very difficult. Hence +1 to gearing the motor down. Making it run smoothly at 300rpm is so much easier than at 30rpm.

[David Hess]

Why not use a small stepper motor?  Then the RPM can be controlled by a function generator.

[ricko_uk]

Yes, all good alternatives.
Thank you both

[Siwastaja]

Stepper is not bad - open-loop of a motor which is built as very high pole count (thus low fundamental RPM) side-steps all control difficulties. The downsides are much worse torque density than similar geared DC motor - a small stepper might not cut it - and poor efficiency. Neither is probably a problem here: use a tad larger stepper and let it heat up. But do get a micro-stepping controller, otherwise the vibration from full-stepping drives you mad.

[SteveThackery]

But do get a micro-stepping controller, otherwise the vibration from full-stepping drives you mad.

Seconded!

Also, the torsional vibration might stress the wire. The last coil I wound had wire not much thicker than a hair. I had to build a very soft mechanical "buffer" (a pulley wheel on a very long, very low-rate spring) just to avoid breakages arising from the wire tugging on the reel.

I think I'd go for a geared motor myself, with a nice, smooth speed control.

[David Hess]

But do get a micro-stepping controller, otherwise the vibration from full-stepping drives you mad.

20 years ago we were using stepper controllers which worked great, and had one external resistor for current programming.  But the current crop of Chinese steppers which are available do not seem to work nearly as well.  My experience is that microstepping on a cheap Chinese controller was *worse* than full or half steps.

Also, the torsional vibration might stress the wire. The last coil I wound had wire not much thicker than a hair. I had to build a very soft mechanical "buffer" (a pulley wheel on a very long, very low-rate spring) just to avoid breakages arising from the wire tugging on the reel.

Maybe add a pulley reduction with a thinner rubber belt?

[ricko_uk]

Thank you all for the additional suggestions, all valid and interesting

[SteveThackery]

Maybe add a pulley reduction with a thinner rubber belt?

Yeah, maybe. Myself, I just don't think a stepper motor is right for this application. A geared motor with a speed controller is just easier and better. Especially if it could be lifted whole from a cheap cordless screwdriver.

[BennoG]

20 years ago we were using stepper controllers which worked great, and had one external resistor for current programming.  But the current crop of Chinese steppers which are available do not seem to work nearly as well.

You can still get these gecko drivers.

Benno

[David Hess]

20 years ago we were using stepper controllers which worked great, and had one external resistor for current programming.  But the current crop of Chinese steppers which are available do not seem to work nearly as well.

You can still get these gecko drivers.

I will check them out.

I tried to find the company that I used 20+ years ago, but could not remember the name, something like "SMC", and searches turned up nothing to jog my memory.  Maybe they no longer exist.

The Chinese stuff all seems to rely on the TB6600.  Maybe the DM556 is better.

[Benta]

Just buy something like this. Can't get much cheaper:
https://www.amazon.de/JGY-370-4632-370-reducer-motor-self-locking/dp/B0CQPHG8GC

[SteveThackery]

Just buy something like this. Can't get much cheaper:
https://www.amazon.de/JGY-370-4632-370-reducer-motor-self-locking/dp/B0CQPHG8GC

Although a speed controller would be required as well.

[EPAIII]

The trick to slow speed, DC motors is pulse width modulation. I have had good results with using a 555 timer circuit, followed with a current booster. It provides full rated torque (well to a degree) at very slow speeds. That high torque will help provide a constant speed over a reasonable range of loads. If that is not good enough, then a closed loop servo can be employed.

A stepper motor is another way. It also employs the rated torque (also with limits) on each step so there is, in theory, no lower limit on RPM. But it will be in steps. A high belt or gear ratio can help make the steps too small to notice. And it will also provide more torque.

I would choose the DC motor with pulse width modulation if the requirements are not too severe. The project would be easier to finish.

[Benta]

DC or PWM or whatever. It will never solve the problem of commutator cogging.

[Circlotron]

Automatic sliding door motors come in PMSM or DC with x pulses per rev tachometer. And typically 15:1 worm gear reduction. Dunkermotoren  DC ones in particular are very smooth. Motor size in the pictures is typically about the size of a drink can.

https://www.google.com.au/search?q=automatic+sliding+door+motor&sca_esv=560d0aa0f230caa6&sca_upv=1&udm=2&biw=1309&bih=957&ei=ckG_Zq-gBtCdseMPldm4iQI&ved=0ahUKEwivof2ivfmHAxXQTmwGHZUsLiEQ4dUDCA8&uact=5&oq=automatic+sliding+door+motor&gs_lp=Egxnd3Mtd2l6LXNlcnAiHGF1dG9tYXRpYyBzbGlkaW5nIGRvb3IgbW90b3IyBRAAGIAEMgUQABiABDIGEAAYCBgeMgcQABiABBgYSJguUKwRWJ8ocAF4AJABAJgBuAGgAZIMqgEEMC4xMLgBA8gBAPgBAZgCC6AC1QzCAgoQABiABBhDGIoFwgIGEAAYBxgemAMAiAYBkgcEMS4xMKAHthM&sclient=gws-wiz-serp

[David Hess]

A stepper motor is another way. It also employs the rated torque (also with limits) on each step so there is, in theory, no lower limit on RPM. But it will be in steps. A high belt or gear ratio can help make the steps too small to notice. And it will also provide more torque.

Microstepping reduces the step size so much that it might be ignored, however my own experience is that some microstep drivers are much better than others.  They should all be the same because it is easy to implement, so I have no idea why results vary so much.

Analog quadrature signals can also drive stepper motors without cogging, however I think the sine drive has to be corrected (arctan?) because the rotor position is a non-linear function of the magnetic flux.  Digital microstepping should be much easier because the correction can be built in, but see above.

DC or PWM or whatever. It will never solve the problem of commutator cogging.

Low duty cycle PWM really does help though, and my own observations is that the commutator cogging becomes insignificant.  Model trains take great advantage of this to operate at very low speeds without stalling.

[Benta]

Model trains take great advantage of this to operate at very low speeds without stalling.

Have you seen the gearbox on model trains? Not exactly simple. Also, the motors have a 5-pole armature, not just 3-pole.

[Doctorandus_P]

As a very simple alternative, you can use a brushed DC motor, combined with an adjustable constant current power supply. (A lab power supply in CC mode works) Current determines torque for a DC motor. The goal is to set the current high enough to have a bit of torque, and then you can just feed in the wire with whatever speed you like. You can even hold the wire to stop it or pull on the wire to unwind it a bit if you overlapped windings.

I have done this a few times and it worked quite well. Of course it depends on the motor size, wire thickness and such.