PDA

View Full Version : Switching power supply fo small DC motor PWM drive



Forrest Addy
05-30-2013, 02:49 AM
I'm working on a design for a motor driven feed unit for a lathe compound. The object is a design of low cost new components a home shop guy can accumulate for about $60. This may lead to an article for HSM but of course I have to build it and prove it first. The speed is varied (I found a slick cheap little 12V PM motor/worm reducer) with a cheap $9 PWM drive. This combination works great on the breadboard. The motor current is about 1 amp under simulated load. The motor output has plenty of torque to run the compound lead screw via timing belt reduction. It's smooth and quiet with good speed regulation. I'm happy with this part.

Next I need to select a caged power supply. I was looking at $10 to $20 24 VDC 2 to 3 Amps switching power supplies on eBay like these:

http://www.ebay.com/sch/i.html?_sop=15&_from=R40&_sacat=0&_nkw=24v%203a%20power%20supply%2072w%20-adapter&_mPrRngCbx=1&_udhi=20

I understand from half-remembered posts that switching power supplies can be quirky when they feed a DC motor. Something about load sensing leading to nuisance surges and shut-offs.

Anyone have experience with this? How do I recognise power supplies that can run small DC motors within their current capacity and how do I rule out those that won't.

Quirky power supply behavior has me a little spooked. I hate expensive failed experiments.

darryl
05-30-2013, 03:43 AM
I don't know if there's a way to decipher this by looking at the power supply specs- at least I haven't found any good info on this. Recently I used a switching supply rated at 1.5 amps to drive some led modules. When these things are operating, they draw about 250 ma- the power supply would not stay on, it just continued to pulse on and off. It would charge a capacitor ok, but still you could not hook up the led module and have it work. For its rated voltage and current capacity, the power supply should have been able to drive 5 of these led modules. I would normally suggest that using a well over-rated power supply would assure that it would be able to drive a small load, but obviously to me, and this is your concern, it won't always be the case. It's almost a matter of trial and error.

I will suggest that a computer power supply is cheap enough to try. It might be physically larger than what you want, but I have run dc motors directly from them without problems. I have not tried using them with motor speed controls in the circuit, but I think there's a good chance that they will work fine. For 24 volts output, you would tap both the + and - 12v outputs, being aware that both outputs have to have the high current rating. It won't work if the +12 is rated at 10 amps, but the -12 is rated at 1/2 amp, for instance.

In many cases, I end up opting for the 'brute force' transformer/rectifier/filter power supply that was the standard before switching supplies became popular. The reliability is unsurpassed- of course we do have these things coming from China, so maybe that isn't true anymore-

Paul Alciatore
05-30-2013, 04:11 AM
Some switching supplies need a certain minimum amount of load current to operate properly. I know because I am doing a design now. Actually, my supply would work for you, but it is no where near ready yet. Anyway, this, the minimum load current, is one spec you need to watch out for.

One way around this is to add a dummy load resistor to draw that minimum current. Wasteful, but it works.

Personally, I would look for a hefty wall wart if all it is doing is running a motor. Or several motors. For a motor it does not need to be regulated, just filtered.

J Tiers
05-30-2013, 08:30 AM
The minimum load has been an issue, but most do not have that problem if designed for general use. If for a specific use (computer, etc) where a known minimum load exists, it may be required.

Another issue is response to surge current. Motors draw a surge, which can be considerably higher than running current, at starting, and some supplies may respond to that by going into a shutdown/restart mode, which they never exit until load is removed.

You want a supply which will at the least just go into a constant current mode, and preferably one with a short-term overload capability matching or exceeding the motor surge.

+1 on not getting too fancy.... why use an SMPS when a simpler supply works better?

For those building their own SMPS, look into the Power Integrations chips, available through Digikey. Up to a couple hundred watts, off-line universal power supplies are possible, and Digikey has a line of transformers which can work with them. There are other chips besides the PI parts, and the transformers usually can work with ones they may not have been strictly designed for.

ed_h
05-30-2013, 09:27 AM
Forrest--

I can relate an experience in trying to use a switcher to drive a DC motor:

I was doing some mods to my MIG welder that involved providing a separate supply to drive the wire feed motor. I bought a small switching supply with plenty of capacity, but when I connected the motor to it, the motor would just jog at about a 1-second rate. I finally surmised that the inrush current of the motor triggered a foldback circuit in the supply that took about a second to reset. I finally had to build a dumb linear supply to make it work. The switcher was too smart for that application.

Here's the application if you're interested:

http://bullfire.net/Welder/WP_Welder.html

MaxHeadRoom
05-30-2013, 10:16 AM
I wouldn't bother with switching supply either for this application, it would not take much to make up a simple linear supply, more rugged and easily fixed, if need be.
Max.

DICKEYBIRD
05-30-2013, 11:45 AM
ebay usually has a variety of new & used 24vdc linear power supplies. I got lucky & found a great ol' 2.4A Tamura for $15 shipped. That was maybe a bit unusual but not by far. Thing's built like a tank from man-sized components that a guy can replace when/if nec.

ironmonger
05-30-2013, 12:19 PM
Here is a cheap transformer for a linear supply
http://www.antekinc.com/details.php?p=20

here is a rectifier
http://www.radioshack.com/product/index.jsp?productId=2062583

and a filter cap
http://www.radioshack.com/product/index.jsp?productId=12460895&filterName=Price&filterValue=%244.00+-+%245.99

nice solid linear supply for under $20 or so
less if you scrounge...

paul

Jaakko Fagerlund
05-30-2013, 01:25 PM
In here at least I can buy off the shelf transformers or power supplies in 5, 8, 12, 24 and 48 V in DC or AC from just about any store that sells electrical stuff. Comes in a nice casing as a DIN module, just needs a length of DIN rail to mount to, hook up your mains to it and get what you need. Costs from 10 EUR and up.

darryl
05-30-2013, 08:34 PM
From the link that ironmonger posted, you would likely want the AS-1218. Once rectified and filtered, the output voltage would be about 25 volts dc with no load, and about 23 or so with a 5 amp load. That's more current capability than you're asking for, but at least the voltage will not fluctuate widely between loaded and unloaded. You'd have a bit more capacity available for future use, and it won't use much power when it's not loaded, even though it has a 100va capacity. As suggested, use a bridge rectifier and a good sized capacitor, and you have a linear power supply good for a long time with little to go wrong with it.

Switching supplies have their strong points- light and small- and their weak points- complexity and potentially troublesome. For any fixed station where weight and size aren't really important, I avoid switching power supplies. The brute force ones (linear, or analog, iron transformer) rule.

By the way, that series of toroidal transformers is designed for audio use, but that doesn't mean they're not good for simple to complex power sources for other uses. They are probably better, since they will be quieter both sonically and electronically- nothing wrong with that, and potentially good if your circuitry being driven is sensitive to incoming noise or radiated magnetic fields.

I'm also going to suggest more capacity than 4700- maybe three times that much. You could use a 25 volt rated capacitor, as that is about what the open circuit voltage would come to, and that is a working voltage rating so it's not like you'd be cutting it too close. Use a 35 volt rated part if it feels safer to you, but that's not really required. More important will be the ripple current rating, and the ripple voltage, which is why I suggest a much higher capacity value. Nothing wrong with using 20,000 or more mfd, and a bridge rectifier rated at 25 amps costs little more than a 10 amp device. That's just more added reliability- to me if you're going to go to the effort to make it up, you might add $5 to the cost and make it strong like bull.

MaxHeadRoom
05-30-2013, 09:57 PM
One advantage of Toroidal is they are easily modified as to voltage, take turns off or add.
One thing to watch with going too large on capacitance it increases the required VA when used at or near rated current.
Max.