Looking for PWM controller

Thread Starter

R!f@@

Joined Apr 2, 2009
9,918
Remember the motors I have posted.
Well I have tested those except the steppers.
All the 2 wire ones have a good torque and speed. Once I get the tacho I can measure the RPM.
So for now I am planning to build a small winding machine to complete my woofers, doing this hand is really tiring.

All the motor can handle up to 24V , the big one takes (like a star body) more current around an amps or so and I cannot even hold the shaft while even at low RPM. as for my engraver I'll get the nema steppers u guys suggested, and I can buy them once the cable job is finished,

For now like to use one of the DC motors at the winding machine. The shaft and bobbin locking machs and reduction gear and the frame work is finalized.

All I need now is PWM controller that can rev up the motor from 0 to max RPM, Vcc will be 24V and current will be less than 3A.
I did a lot of search and nothing good came up except one at silicon chips 2008, issue 234. It has a very good PIC16F88 controller with enough current capability plus a counter, but for the detail they are asking for my credit card. And I cannot do that.
I like to ask you members that does any of you have a controller that I can use, if you don't mind, a PWM one is preferable.
I have the PICF88 if needed, but I guess logic will also do.
I thing the machine builders here will have something atleast I can modify.
I want to use 2 Wire DC motor as it is easier for me to make the machine and all the rest of assembling is finalized.
Just for your info I am attaching the power drivers that I have.
L297, 1 pc
L298, 1 pc
L6203, 4 pc
MPM3003, 1 Pc
MPM3004, 2 Pc

These H bridges and DMOS full bridges.
I can use these as the out put power drives.

So please, if any one have any thing, can you share it.
If all fails I guess I will pay silicon, but this is the last resort, and after this I guess I have to renew my Credit card.

PS . I can't seem to attach more than 2 PDF's

Thanks
 

Attachments

rjenkins

Joined Nov 6, 2005
1,013
You do not need PWM for speed control of a DC motor.

The simplest method is just a variable DC supply - as long as the torque on the motor is not excessive, the speed will be proportional to voltage.

Another very simple dc speed control uses a small sensitive-gate thyristor.

You basically run the motor on rectified but unsmoothed DC, using a thyristor wired as if it were a transistor in 'emitter follower' configuration feeding the positive side of the motor.

You connect the gate via a diode to a variable reference voltage, something like a zener with a pot across it & feed from the pot slider. If the thyristor is not sensitive enough you may have to add a small transistor as an emitter follower to give the gate enough current.

(You can get the reference supply using an extra diode off the output of the main bridge & feed a small smoothing cap, then to a resistor & the zener etc.)

This works as the thyristor only fires if the back EMF from the motor is low enough for it to draw current through the gate from the pot.

The lower the pot reference voltage, the lower the speed.
Torque is not reduced.

I used to build a version of these for friends who were model railway enthusiasts, many years ago (for trains working on 12V DC). They could control the train from full speed down to something like one inch per minute.
 

Thread Starter

R!f@@

Joined Apr 2, 2009
9,918
So you are saying that the motor won't have a jerking effect, in low rpm's the motor tend to have a rpm jumping effect. That's why I wanted a Speed controller, but if you are sure that this won't be there then I can give it a try.

But still why there are PWM controllers for DC motors, is it not that RPM will not change with load. My winding machine will have a quite a load to the motor as the wire will have tension, This will tend to stall the motor at the starting stage as I will be at a very low rpm when the winding begins and plus I will be applying fiber glue between the kapton and the coil layer. So any problem here will create a lot of difficulty . The fiber will harden quite fast so I will have zero tolerance for errors.
I have tried with a hand operated winder, that was a bad idea as most of the wire have to be discarded as even a slight speed reduction results in loosening the winding. I cannot apply glue with one hand driving the bloody handle..
I wll post the machine details once I get my itunes working. No camera just my new i phone. I have started the work on it, but for now only the coil reel support is half way through. I cannot continue as I ran out of 6mm acrylic plastic, so now I am planning to make the frame with 10mm sheets.
I don't have a machine shop, so no metal work as most of you, all I have is laser cutter, so plastic will do fine, by this I can gaurantee that it would work on my engraver :)
 

rjenkins

Joined Nov 6, 2005
1,013
The thyristor based controller should give very even speed control with virtually no load dependence.

You should still gear the motor down as far as possible so your maximum winding speed is toward the limit of the motor speed.
 

Thread Starter

R!f@@

Joined Apr 2, 2009
9,918
Yea about gear reduction. I have a few ideas and some gears for that, Final gear is the third one pinned to the bobbin shaft, supported by two bearings. I do not need such high RPM, as I don't have any plans to wind thousands of turns, I only repair speakers and woofers, for this a 100 RPM is even more than sufficient, just need a steady low rate. The motor in mind is capable of 1000 RPM. If this motor creates a problem I have other powerful ones that can be placed with the same gear assembly.

OK then I am off to find a good SCR. I'll post back ones the circuit is tested.
 

retched

Joined Dec 5, 2009
5,207
Of you have a 10,12,18 or whatever speed bicycle, they make easy transmissions for electric motors. I used one for my VAWT, but it was too noisy. (I live in the city)

For you, the gears are there, and they work with easily sizable, readily available chain.

This also gives you the ability to change speeds per application to get to one your comfortable with, without having to do anything but move the gearshift lever.
 

Thread Starter

R!f@@

Joined Apr 2, 2009
9,918
Guys ... I am using 60V full bridged without smoothing fed to the anode of a 3A SCR, cathode is connected to the motor.
Gate is grounded by 150K and driven by an emitter follower like rjenkins said, gate supply is tapped off via diode and 10uf 100V cap giving gate max of 80V.
I get full speed plenty of torque but the low speed is just as I suspected, it jumps.
the lowest speed is not applicable as motor is like a grasshopper.
Told ya. that jerking will be there, so what's the remedy guys?
 

kkazem

Joined Jul 23, 2009
160
Hi all,
I design PWM power circuits for a living. For motor control, yes, at low speeds especially for a winding machine (and I have designed and wound many power transformers and inductors myself using a variety of winding machines), where the torque needs are likely to be the highest, you need a PWM controller for the DC motor and it isn't all that complicated or expensive. In fact, all you really need is an SG3526 control IC, a single power MOSFET, a single catch diode, and a low-ohm current-sense resistor to protect the power switch from over-current, and a bias supply for the control IC. I'm going to assume that your 24 VDC is unregulated and won't go over 35 VDC for the circuit I designed and uploaded. The parts are relatively inexpensive. For the SG3526N PWM Controller IC, you can use the TI-Unitrode UC3526 instead of the Microsemi part, they are completely interchangeable. I've provided pdf datasheets for all the important parts and the schematic.

Please reply if you have any questions about the circuit.
Regards,
Kamran Kazem
 

Attachments

Thread Starter

R!f@@

Joined Apr 2, 2009
9,918
Thank you so much. I am checking 'em now

Later.
I got 3842/43 ..........UC3524 UC3525 too, but no UC3526.
Not even in my junk yard....come to think of it, I never came across a 3526 before.
I checked, cannot be interchanged as 26 has current sense inputs :(
 
Last edited:
Top