cheap dc power supplies

[Lotus 7]

Looks like a bargain. It would be difficult to purchase the caps, diodes, regulator, heat sink and the PC board for the price. The filter caps are rated for only 35 volts, so heed the recommended transformer rating and don't go over 22 volts AC in with that board. That will keep the cap voltage to a nominal 30-31 volts. The main filter cap is only 1000uF which should be enough for possibly a 300-400 mA load. If you want to draw higher currents, the capacitor value is a little on the low side. There is also no turn on surge limiter between the diode bridge and the first filter cap, so you'd want to use a transformer that can't put out excessive current. I'd keep to a transformer rated for 1 Amp or less, and don't even think about trying to "get away with" 24 volt transformers because they're easy to find. With a 24 volt transformer, a light load and a slightly high line voltage , the input filter cap's voltage rating will be exceeded.

Also, it's nice that they have a terminal point connected to the unregulated DC input of the LM317 (the input filter cap line labeled "DC in"). If the supply has too much ripple for your application, or you want to use it at a higher output current and need more filtering, that's an easy place to hang another filter cap to clean things up.

The circuit is fully floating so there is no problem switching the output polarity by using the terminal labeled "GND" as the negative supply line.

Just be sure to use separate transformers, or one transformer with separate secondary windings for each polarity supply.

[enginefire]

Authoritative response!
For the price I think I am going to experiment a little and try some tweaks you have suggested. I'll work out the draw, but I can't imagine a couple of older opamp or discreet circuits pulling all that much juice.
Thanks lotus7!

sent from my m-3500 using the gearslutz app

[dualflip]

Ive bought that PSU in the past, great PSU i can vouch for it. Caps are Rubycon which is a great brand, the circuit is a standard LM317 PSU taken from the datasheet.

[Lotus 7]

enginefire,

Because you want 24 V DC outputs, you truly need to supply the full 22 VAC to the diode bridge. That's great if you can find a 22 VAC transformer. Unfortunately, 22 VAC is not a common transformer secondary voltage. There are literally hundreds of 24 VAC transformers available but few 22 VAC.

I did find a very inexpensive "wall-wart" 22VAC unit that would be a decent match for the regulator board, if you don't mind the physical arrangement.

The alternative, and better approach would be to use a readily available 24 VAC transformer (like this one) and to replace the input filter capacitor (C-1) with one rated at 50 VDC. The LM317T is rated for an input to output differential of 40 volts so that's not an issue.

[Matt Syson]

Hi
Relatively little to fault especially at that price. However although the LM317 is rated at 1.5 Amps in an 'ideal' world, the heatsink is not large and asking it to deliver high current with many volts 'across' the regulator (which is NOT the output voltage) will make it overheat and shut down. As a guide the heatsink should not be allowed to run at greater than 'hand comfortable' when it is enclosed in it's final home. If it does, you need a bigger heatsink!
Matt S

[dualflip]

Originally Posted by Matt Syson

Hi
Relatively little to fault especially at that price. However although the LM317 is rated at 1.5 Amps in an 'ideal' world, the heatsink is not large and asking it to deliver high current with many volts 'across' the regulator (which is NOT the output voltage) will make it overheat and shut down. As a guide the heatsink should not be allowed to run at greater than 'hand comfortable' when it is enclosed in it's final home. If it does, you need a bigger heatsink!
Matt S

Ohh yes, that heatsink wont be enough for higher current demands, using that PSU at 24V ive noticed that anything above say 250mA will make it VERY hot, the lm317 will shut down when the temperature reaches 125ºC at the junction, but remember that its always hotter at the junction than on the outside, so 70ºC on the case means a lot more in the junction.

It depends on how much current do you need, the API 500 series lunchbox PSU uses LM317 and LM337 and they have a MASSIVE heatsink which is practically the entire enclosure, to be able to dissipate the amount of heat being released at currents such as 1.5A

[vince @ speck]

For +/- dual outputs you'll need two PCB kits + a transformer + a larger heatsink (No way that kit will handle 1.5A)

Why not just get this

Power One HBB24 1 2 A 24VDC 1 2Amp Power Supply New in Box | eBay

[dualflip]

Originally Posted by vince @ speck

For +/- dual outputs you'll need two PCB kits + a transformer + a larger heatsink (No way that kit will handle 1.5A)

Why not just get this

Power One HBB24 1 2 A 24VDC 1 2Amp Power Supply New in Box | eBay

+1 on the power one, however they must be mounted outside of the equipment in question, the transformer on those PSUs radiates just too much EMI, ohh yeah and they tend to mechanically buzz a bit more than i would like when mounted on metal.

[Richard Crowley]

If you need bipolar, then you would need TWO of those PowerOne supplies. Better to just get a bipolar version, IMHO.

It it not clear how a traditional "heavy-iron" linear power transformer transmits "EMI"? This is NOT a switchmode power supply.

The typical application of this type of open-frame supply module is to mount inside the equipment. I have never had any problems with these kinds of PS units, especially PowerOne who make good stuff.

OTOH, that form-factor (the old-school big, open-frame linear supplies) are likely too big for internal 500 rack style applications.

[vince @ speck]

Originally Posted by rcrowley

If you need bipolar, then you would need TWO of those PowerOne supplies. Better to just get a bipolar version, IMHO.

It it not clear how a traditional "heavy-iron" linear power transformer transmits "EMI"? This is NOT a switchmode power supply.

The typical application of this type of open-frame supply module is to mount inside the equipment. I have never had any problems with these kinds of PS units, especially PowerOne who make good stuff.

OTOH, that form-factor (the old-school big, open-frame linear supplies) are likely too big for internal 500 rack style applications.

The HBB24 is a dual output ... bipolar.

Who said this was for a 500 rack? If it is, then it would be short-sighted not to include a +48 as well.

[Richard Crowley]

Yes, it IS bipolar. My error. The manufacturer's data sheet is pretty generic and not as straightforward as it could have been.

I just got the impression that it was for a 500 rack (or some sort of modular assemblage.) The details of the application were not revealed by the OP. And indeed it seems like something for P48 would be necessary if this includes mic preamps.

[dualflip]

Originally Posted by vince @ speck

The HBB24 is a dual output ... bipolar.

Who said this was for a 500 rack? If it is, then it would be short-sighted not to include a +48 as well.

Never said it was for a 500 rack, I was just trying to give an example of a well know LM317 based PSU with a massive Heatsink

Originally Posted by rcrowley

If you need bipolar, then you would need TWO of those PowerOne supplies. Better to just get a bipolar version, IMHO.

It it not clear how a traditional "heavy-iron" linear power transformer transmits "EMI"? This is NOT a switchmode power supply.

The typical application of this type of open-frame supply module is to mount inside the equipment. I have never had any problems with these kinds of PS units, especially PowerOne who make good stuff.

OTOH, that form-factor (the old-school big, open-frame linear supplies) are likely too big for internal 500 rack style applications.

Im sorry, EMI was a bad choice of words, induced magnetic field would be a better one? and yes they are great units. Believe it or not, from own experience ive put these power one PSUs inside equipment and the induced hum is considerably loud, they should be taken out of the box.

[enginefire]

I will probably put the PSUs in an old computer PSU box and run power to the units from there, hopefully being able to power more than one thing with a pair.

I got two to power a few things that will run different + voltages, and I might try to experiment with getting +/- 24 as well for some other equipment. Don't worry, I wont try this at the same time! Although,, maybe it would be possible to run them +/-24v and take some extra lines from the +24 and reduce the voltage with a resistor to +20 and +18. Would this actually work or would it negatively affect the performance of the +/- 24?

Phantom power... Yes that does sound useful, I don't have a closet full of c1000s is it possible to make a voltage doubler circuit from the -24v and then invert it to +48? Maybe this would be asking too much from these humble PSUs.

The other option would be to construct a separate box to go between the pres and mics.

sent from my m-3500 using the gearslutz app

[doulos30]

I'd make a voltage trippler and regulate it down to 48v. Though you could proably regulate it up to 48v as well with a doubler

[Lotus 7]

Originally Posted by enginefire

I got two to power a few things that will run different + voltages, and I might try to experiment with getting +/- 24 as well for some other equipment. Don't worry, I wont try this at the same time! Although,, maybe it would be possible to run them +/-24v and take some extra lines from the +24 and reduce the voltage with a resistor to +20 and +18. Would this actually work or would it negatively affect the performance of the +/- 24?

If you need a stable voltage, dropping it simply with a resistor is usually not a good idea. The voltage drop will, of course, change as the current changes so you have no voltage regulation. Dropping with a resistor also creates a high source impedance for the power feed which is never a good thing to have for any audio circuits.

To drop the +/- 24 to a regulated +/- 18, just add a couple of three-lead fixed reguator ICs (on small heatsinks or mounted on a metal chassis). A "7818" and one (1uF) output cap will give you a stable +18 volts and a "7918" plus one (1uF) output cap will give a stable -18 volts. If you need the +/- 20 volts a pair of resistors connected from the respective output terminals as a 9:1 divider to the common leg of the 7818/7918 regulators will shift them up to 20 volts.

Remember all the current you are using at both +18 and +24 is going through the first regulator so if you're drawing 250 mA at +24 and an additional 250 mA at +18 then the first power supply must be able to provide 500 mA.

I've attached a schematic of a circuit that will provide both +20 and +18 regulated outputs from your +24 volt supply. If you also want -20 & -18. the same can be used with "7918" regulator ICs on the -24 volt supply.

If you never need the lower voltages at the same time as +24 v, then the simplest thing is to wire up an external pot, or a switch with some fixed resistors to the LM 317 control pin to substitute for the one on the PC board to allow easily changing the output voltage.

[Matt Syson]

Hi
The 'Power One' units as mentioned above use 'industrial grade' transformers and are not optimised for minimum magnetic hum radiation so you do need to be careful where you put them in relation to sensitive circuitry.
They can also emit 'RF' energy due to 'switching' of the rectifier diodes. I have not personally had a problem with this but should be borne in mind.
Filters on the incoming mains are a good idea to reduce noise coming in, and any noises from the rectifiers getting back out.
They should be mounted in a metallic enclosure (properly grounded of course!).
They may also need the metallic case for heatsinking duties if any significant current is being drawn, although up to 1 amp should be fine.
Matt S

[Lotus 7]

Jeeez!!
This thread is soooo 19th & 20th Century.

AC power.... at least 180 year old technology
Transformers....130 year old technology
silicon diode rectifiers....109 year old technology

To really provide an induced hum and RFI free supply the real answer is a stack of 3.7 volt Lithium Polymer batteries recharged by a large bank of monolithic silicon solar cells. To reduce the possibility of any noise pick-up, the battery pack should be encased in an air-tight container made up of at least three layers of (0.25 mm) MuMetal to shield against induced electromagnetic currents, plus (2) layers of (0.50 mm) oxygen-free copper as an electrostatic shield (Faraday cage)(grounded by a copper rod driven at least 20 feet into the water table), plus at least a single layer of 2.5 cm thick Mallory-Metal (tungsten, copper, nickel alloy) shield to stop ionizing radiation from inducing micro-currents in the battery dielectric layers.

The above supply can be positioned right next to any low level circuits with no risk of inducing any noise.

It's so simple it's a wonder that some "boutique" mic pre company is not offering a front-end powered by such a system.*

*The patent application will go out in Monday's mail.

Any Venture Capitalists out there who want to invest....PM me.

[Matt Syson]

Hi
Darn, I missed one of the shielding layers on my unit, but the good news is that the water table here is about 6 inches below the ground surface, or above it when it rains!
Matt S

[Lotus 7]

Originally Posted by Matt Syson

Hi
Darn, I missed one of the shielding layers on my unit, but the good news is that the water table here is about 6 inches below the ground surface, or above it when it rains!
Matt S

Of course the long ground rod is not necessary in most of the Eastern US because the water table is now unfortunately above ground. :-(

[dualflip]

Originally Posted by Lotus 7

Of course the long ground rod is not necessary in most of the Eastern US because the water table is now unfortunately above ground. :-(

black humor but it made me laugh

[enginefire]

Lotus 7, I already have the boys pounding a grounding cage into the foundation of the new facility under construction, the whole building is going to be a big grounded shield. This way I can take off my tinfoil hat when I'm inside! I'll upload a picture of the progress tomorrow.

So the rough plan right now is to install the boards into a salvaged metal computer PSU case, after removing the regulator and heatsink from the board. I will remount the regulator along with a few others on a larger salvaged heatsink and/or the case. I will use yet another salvaged item (I took an old amount my father in law was throwing out home), a switching pot (the amps input selector) and wire up the regulators in all the combinations I think I will use.

I will use one board to give +24 and +20 and the other to give -24 and +18 to balance out the load.

Once all this is here and put together, I'll start on the +48! Thanks for all the great help so far!

sent from my m-3500 using the gearslutz app

[Lotus 7]

Originally Posted by enginefire

I will use one board to give +24 and +20 and the other to give -24 and +18 to balance out the load.

You can easily get +24and +20 from one of the supplies by adding the "bootstrapped" 7818 as in the schematic on my post, but you can't get -24 and +18 from the other (at the same time). The -24 supply has its positive output grounded to provide the "upside-down" negative voltage (on the output terminal marked "GND"). No positive voltage is available at the same time.

[enginefire]

Hard at work! Just not on Sunday.

Mega grounding super faraday change headquarters in progress.

No cell signal inside (at least that's what the specs say)

sent from my m-3500 using the gearslutz app

[Lotus 7]

Originally Posted by enginefire

Hard at work! Just not on Sunday.
Attachment 316081
Mega grounding super faraday change headquarters in progress.

No cell signal inside (at least that's what the specs say)

sent from my m-3500 using the gearslutz app

Hope you finish it soon. Then you can take off the aluminum foil hat.

[dualflip]

seriously a Faraday cage? are you next to a radio station or something?

[enginefire]

I need it to stop the voices!

naw seriously im just messing around thats a condo foundation retaining wall going in to hold up the rest of the city around the big pit they're gonna dig.

I rent a 240sq live room right next door.

[enginefire]

how do these look for a phantom power box:
hammond 229E120
2 secondary outs of 60v 40ma each.

i guess this could be used to make a nice 48v with enough juice to power maybe 8 mics per sec winding?

i will redraw the schematic for the psu with the additional parts to get the extra voltages later and post it for review!

[enginefire]

so a little update and some good news,
I am still waiting on ordering some caps to upgrade those cheap DC supplies, but in the meantime I scored an original power supply from a Scully 280 reel to reel... along with two preamps with all the cards and iron Dont ask me how I did it, but I did!

My plans are now changing to use the scully psu as a main supply to power just about everything I could need:
24v at 3 amps
24v at 5 amps
multiple AC taps at 3 amps (I think its 70v, 50v, and 30v).

I would like to get +/-24v from the large trafo (its 48v ct from what I understand) and add parts to get phantom from the smaller AC trafo. both the transformers are toroidal.

I have a question, should I replace all the diodes and capacitors inside the PSU? If so, does anyone feel like helping an EE noob like me with the design? I am pretty much a read and replace kind of technician.