Variable DC power supply question

[convoluted]

Hey all,

I'm looking to pick up a variable DC lab power supply in the near future and I'm stuck between two options. Would it be better to go for a 0-36V @ 3A supply, or a 0-20V @ 5A supply? I will be using the supply to power servos, brushed DC motors, LED matrices, microcontrollers, and basic electronic components. I really only see myself using up to 20V @ 3A (which both supplies can provide), but which option do you think would be better to go with (ie, will offer better flexibility in the future)?

The supply I was looking at is a Tenma switching supply: http://canada.newark.com/tenma/72-8345/power-supply-bench-36v/dp/01N4887. Anyone with any experience with these supplies? Any other suggestions for supplies? I was also looking at this one: http://cgi.ebay.ca/0-30V-0-5Ampere-DC-Adjustable-accurate-Power-Supply-/160456228081?pt=LH_DefaultDomain_2 but I couldn't find much info on it, which raises some flags.

Thanks in advance

[waltr]

Tough one but since you may be powering motors etc I'd go with the high current one. Unless you may use higher voltage motors.

[convoluted]

Thanks for the reply waltr. I'm now even contemplating buying a cheaper 1.5-15V @ 2A supply for simple IC testing, then converting a PC PSU for a high power applications. I particular I was looking at maybe this supply http://cgi.ebay.ca/MASTECH-HY152A-DC-POWER-SUPPLY-VAR-15-VOLTS-2-AMPS-/230504940342?pt=LH_DefaultDomain_0, which is fairly cheap. I could even buy 2 and run them in parallel to get 4A - I believe the supply is linearly regulated, though I couldn't find and documentation for it.

[voyager2]

I've found the best thing in terms of testing motors is am 18v 5A variable model train controller.
Go to any half decent hobby shop that has Hornby model trains (or any other brand of model train).
Be warned though that the voltage does fluctuate slightly (in the mV) so you'll probably need some capacitors for running ICs.
It is mostly negligible with DC motors though...
Hope this helps

[Soeren]

Hi,

[...] I was also looking at this one: http://cgi.ebay.ca/0-30V-0-5Ampere-DC-Adjustable-accurate-Power-Supply-/160456228081?pt=LH_DefaultDomain_2 but I couldn't find much info on it, which raises some flags.

The page has got the most important info and if the specs hold water (it's probably a Chinese made supply), it's really-really worth the price.
I cannot say if it does - the data sounds a bit too good compared with the bad looking plastic box wrapping it, but I would go for that one (if not something else entirely).

The  "1.5-15V @ 2A" supply is rubbish for experiments, as it doesn't have an adjustable current limiter like the other one.


I'd recommend a 30V supply if there's a chance that you will be testing 24V motors, or at least a 24V supply.

An adjustable current limiter is the single most important thing on a lab supply (without it, it doesn't deserve the title "lab supply").

Going down to 0V is nice if you wanna test really low voltage circuits, but 1.5V will do for most stuff.

If you wanna test motors, you cannot get too much current, as the start current can be huge compared to normal running currents and 5A is the lowest I'd go as most <=12V motors of some size will have a higher stall current.

For anything but motor testing, 20V and 2A to 3A should do nicely - still with an adjustable current limiter.

[convoluted]

Yep, I agree. The cheap $30 1.5-20V supply is probably not the best option for circuit experimentation, I only initially considered it for the cheap price. I'm having doubts with the $100 Chinese supply too, really don't trust the quality. I'm sure it would be fine, but I prefer something that will last and works without issues. Really prefer a supply with a brand and a warranty. Good call about the motors. I would probably use a converted ATX PSU for driving motors as I could easily achieve 20+ A.

So I think I'm down to two different supply options. One is a BK Precision switching supply (1-36V, 0-3A), the other is a Tenma supply at the same rating and form factor. The BK is $140, the Tenma is $200. The BK has a strange clicky control where as the Tenma has rotary dials. I like the look the Tenma better, but does that warrant an extra $60? Which would you folks go for? Here's the links:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=BK1550-ND
http://canada.newark.com/tenma/72-8345/power-supply-bench-36v/dp/01N4887

Thanks again

[madsci1016]

I bought the $60 0-18V 0-3A   Power supply HERE And have run everything off your list with it, including a hefty 12V 3Amp Marine washdown pump (moving saltwater between aquariums) and have had no issues. Best $60 every spent. Even charged Li-Ion cells with it when I was in a pinch. Just need to set the CC/CV dials right.

Looks like they are out of stock right now though. 

[Soeren]

Hi,

So I think I'm down to two different supply options. One is a BK Precision switching supply (1-36V, 0-3A), the other is a Tenma supply at the same rating and form factor. The BK is $140, the Tenma is $200. The BK has a strange clicky control where as the Tenma has rotary dials. I like the look the Tenma better, but does that warrant an extra $60? Which would you folks go for? Here's the links:

Of those two, I would go for the BK and not only for the price tag. I didn't study their specs though.

But far cheaper supplies are out there and you could build one yourself as well, then you could tailor it to whatever specs you need, want and can afford.

[convoluted]

Hmmm, yep good call. No point spending a tonne of money on something that can be done for much cheaper. I'll take you advice and build a cheap supply using an LM317 and then a more hefty, higher current supply using a converted ATX PSU. Will be good learning experience nevertheless. Then I can distribute the money I would have spent and pick up some additional servos and whatnot. Thanks for the help everyone!

[Soeren]

Hi,

I'll take you advice and build a cheap supply using an LM317

You'd need to boost it with a power transistor (an MJ4502 is a good choice) to make it work properly.
If you build eg. a 30V @1A supply and run something at say 5V/1A, the regulator will have to drop 25V/1A=25W and it would have shut down way before that.

With an LM317, an MJ4502, an op-amp and some glue, you can make a proper lab supply like 0..30V @ 0..5A.
Remember that you need a transformer (around 25V @ 7A), a large capacitor (10,000µF+ and 60V+) and a 7A+ bridge to get the unregulated power feeding the circuit.
And a reasonable heat sink for the active parts is needed too.

Such a homemade lab supply served me well for a couple of decades, untill the main cap blew (waking me up, as I had dozed off on a chair right next to it). Then I rebuilt it into a variable constant speed supply for a mini drill and it has been driving that for around 15 years, making a rather tiny drill more versatile than a Dremel and almost as powerfull (actually appearing more powerfull than a Dremel, as it doesn't slow down when eg. you load a large cutting disk by cutting steel).

[convoluted]

Hmmm, well I suppose I could get by with the LM317 if I strap on a nice heat sink, maybe. Let's say I drop 20V at 1.5A, which is 30W. Without a heat sink the T-220 package has a thermal resistance of 50°C/W (yuck), so the heat sink dissipating 30W will have a temperature difference of 50×30 = 1500°C, so the temperature will rise to 25+1500 = 1525°C (clearly above the 125°C thermal cutoff). However if I use a 3°C/W heat sink ($6 @ digikey), the temperature difference will be 3×30 = 90°C, so the temperature will rise to 25+90 = 115°C, which is close to (but below) the cutoff temperature of 125°C. So it is possible to get by with the LM317, but I'll need a really nice heat sink.

I'll also look into the transistor option as well. Do you have a circuit diagram? What about this circuit: http://electronics-lab.com/projects/power/004/index.html

Edit: Oh wow, check out this heat sink: http://canada.newark.com/h-s-marston/890sp-01500-a-100/heat-sink-extrusion/dp/15J9645 0.08C/W thermal resistance! I can pump through 1500W through this beast!

[Soeren]

Hi,

Hmmm, well I suppose I could get by with the LM317 if I strap on a nice heat sink, maybe. Let's say I drop 20V at 1.5A, which is 30W. Without a heat sink the T-220 package has a thermal resistance of 50°C/W (yuck), so the heat sink dissipating 30W will have a temperature difference of 50×30 = 1500°C, so the temperature will rise to 25+1500 = 1525°C (clearly above the 125°C thermal cutoff). However if I use a 3°C/W heat sink ($6 @ digikey), the temperature difference will be 3×30 = 90°C, so the temperature will rise to 25+90 = 115°C, which is close to (but below) the cutoff temperature of 125°C. So it is possible to get by with the LM317, but I'll need a really nice heat sink.

I wouldn't bet on it, unless you used liquid nitro.

You need to use the entire chain of thermal losses. Here is a good read from Altera on Thermal Management Using Heat Sinks

I'll also look into the transistor option as well. Do you have a circuit diagram? What about this circuit: http://electronics-lab.com/projects/power/004/index.html

The LM723, however nice a circuit, is older than the LM317 and may be hard to come by (and it really should be given a rest after a hard life of regular abuse  )

Try this 0..30V, 0..5A supply, it has served me well for more than 3 decades.
(Some part of it reconstructed from memory (and late at night), but if you choose to build it, just drop me a note if it acts up in any way).

Edit: Oh wow, check out this heat sink: http://canada.newark.com/h-s-marston/890sp-01500-a-100/heat-sink-extrusion/dp/15J9645 0.08C/W thermal resistance! I can pump through 1500W through this beast!

Not through an LM317 
I wouldn't try it (or even a measly 30W) through an LM317 if it was mounted on one of my heat sink tunnels with forced convection!

Besides, it would be quite expensive, compared to a booster transistor and more moderate heatsinks.
You'll still need a heatsink of course - How large would you make it?
(Best way is to use one for the LM317 and another for the MJ4502).

[convoluted]

Wow, thanks very much Soeren for the circuit! Hope you didn't go into too much trouble preparing the diagram. I'll take a closer look at in the following days and will let you know if I need any assistance. Looks fairly straight forward however. Thanks again