Bench power supplies...

I'm starting to dabble in solid state a bit (mostly a tube guy) and want an inexpensive bench power supply to help. I have a Lambda high voltage supply that was given to me for working with tubes but that won't help with ss. My first ss project is a subwoofer amp so I need something that can handle 150W output or more, ideally under (or close to) $100. I'd like it adjustable to at least 30v, but 60v would be ideal. I've been looking around and new adjustable switching supplies are everywhere and can be had for under $100. I've also found a few linear supplies that might fit the bill, but they tend to be much lower current for a given voltage and price. I'm not looking for a high-end, highly accurate supply here (though maybe I should...).

So I have a few questions. First, I know linear supplies are in general quieter (in terms of output) than switching, but how important is that for building and simple testing? And if it is important, could I add a small cap across the outputs of a switching supply (small for higher frequency noise) to help reduce that noise? I've read that many switching supplies are also not very good at fast transients so I assume I could in theory also add a large reservoir cap. Basically, will I be happier with a linear supply even given the lower current/power? Any recommendations for something that fits the bill?

Or, should I DIY a linear supply to meet these general goals? I did some research on this many years ago but gave it up as too complex. I think I could do it now. If so, can anyone suggest some designs? I could probably design a very simple supply, but am looking for something a bit more carefully designed than I could do.

OK, I'm not a highly skilled builder (though I've built quite a few amps (tube and solid state), preamps, phono-stages, speakers, etc.) and I understand that this may be a lot more complex question than I realize, but some general replies would be helpful. At least for now. Thanks so much!
 
One more thing. I know I could get something used like an HP-6024A for around $150, which might be perfect except I don't know how to test and calibrate it. Would that be hard to do? Also, the 6024A is, I think, unregulated. Not sure if I need regulation but it sure seems like a good idea.
 
Administrator
Joined 2004
Paid Member
Okay, so linear supplies are quieter by far. Cheal anything isn't a good plan.

You may want to build your linear supply as it would be for your amp. THen use a variable (Variac) transformer to adjust the voltage for testing.

For general linear bench use, a bipolar 25 VDC supply is the best, not high current. Find an HP 6236 (A/B/C) or 6237 (A/B/C), they are about the best. The difference is one as an additional 6 VDC supply, the other is 18 VDC. All supplies are adjustable.

After becoming skilled you can build a supply, but even layout is important. If you want to depend on that supply (and you do), buy the best you can find, and you can't beat these HP supplies for performance. I've used mine since the 1970's. I did just get some new Keysight supplies, the EDU36311A is extremely good, very quiet but probably overkill. Never mind about 10X what you want to spend. The current limit feature is great.

-Chris
 
  • Like
Reactions: 1 user
SMPS can be made quieter with the help of additional comon mode filtering. Just slapping extra capacitance on the output is a no-no. Most switchers have optimized feedback loops which can go unstable with big caps on the output.

Linear supplies in my experience are
impossible to build yourself for the price of a used unit. You need big coolers, transformer, case and all odds and ends.

I can highly recommend Delta Elektronika supplies.

SMPS is noisier than linear is often repeated dogma, most SMPS are far noisier. But there are exceptions where a SMPS will defeat a linear psu in noise spec
 
  • Like
Reactions: 1 user
Administrator
Joined 2004
Paid Member
Hi v4lve lover,
You're right to a point. With a wide dynamic current range it is difficult to optimize a switching supply for noise.

With the enclosure and filtering requirements, a switcher is also more expensive to build. If you include experimentation to get it right, it's even more expensive. Then there is the skill and experience plus test equipment needed to get it quiet and stable. A linear supply can be easier than a switching supply to be honest. Less HF noise to worry about and shield from. They both need to have the error amp compensated properly. No easy way out there!

Buying a used supply is by far and away the least expensive way to buy a good power supply. When you get into special requirements, you're in custom design territory. There are some specialized switching supplies for low level signal applications (simple switcher for one), but the performance envelop is much more narrow than a linear supply with the same level of noise.

-Chris
 
Thanks anatech. Certainly an interesting idea to build a supply and use a variac to vary the output. But I want a bench supply that would let me test and try things out before building the final supply. I may end up not liking specific projects and I don't want to build a fixed supply that I'm not going to use. I have multiple amp projects in mind from 16v on up to 60-70v (though I know bench supplies that go that high can be expensive). My problem is that my budget is minimal (kids in college taking half our incomes for 5 years) so cheap (inexpensive) is the only option right now. Likely to change once they are finished with school but I don't want to wait 5 years on this. But...I'll rethink things a bit. I probably have enough parts on hand to build a basic LM317-based supply to reach 30v/5A (with a 2N3055 current amp) and maybe that would get me through the lean years. I could also build a 60-70v fixed supply and use my variac to get some voltages in the gap between a DIY supply and the fixed one. Hmmmm.
 
You are correct, ive been playing with flybacks for a while now, and im starting to get the hang of it, and can guestimate the formulas in my head.

The best, most versatile and price competitive linear regulator is the 723 in my book. Everybody here acts like its just a 723, but with the reference bypassed and a 78XX to feed the control circuitry of the 723 it can get down to under 100uV of noise.

Hail the 723! May your reign last trough the ages!
 
  • Like
Reactions: 1 users
Administrator
Joined 2004
Paid Member
Well, the 6236B is a very nice, small bench supply that will give you bipolar 23 or so volts. You can go for a 6255A, that will give you higher current and 40 VDC per supply. It can be strapped to track and give you bipolar 40 VDC (higher normally). That might be the best solution for you.

It sounds like the 6255A might be the way to go.

For testing larger power amps, I took the power supply from a high power amplifier and stuck it in a case with a variac. Add meters and away you go on the cheap. Now, the HP supplies are regulated (quite well) and low noise. A supply like I described with a variac is a variable supply (not regulated) that will exactly mimic a supply you build, complete with ripple and possibly a little worse stability vs load current.

Look up an HP 6255A and grab one if you can afford it. I have two, one for parts.
 
Administrator
Joined 2004
Paid Member
Simple.
To test power amps, take a power supply and add a variac.
To run any kind of a signal amplifier, digital circuit or anything like that you need a low voltage, low noise supply that is regulated. Old HP supplies are extremely reliable and overall excellent in performance.

I have high voltage supplies, high current variable supplies (for power amps), high-ish voltage regulated supplies at a couple amperes, and several very low noise regulated supplies for general use. Plus a +13.8 VDC high current supply for car amps. Each is used and has a specific type of device it is useful for.
The average bench needs at least a good, low noise variable regulated supply (6236B for example), and a variable higher voltage supply, not low noise or regulated (power supply with variac option). Some folks need high voltage like you and me. I have several from 500 VDC tube units to solid state 600VDC @ 1 ampere!
 
Administrator
Joined 2004
Paid Member
Well, if one current limits before the other, it may reverse polarity. You can also run into issues where one may only withstand so many volts above (below) ground. So you can run into real problems doing that depending on the supplies you are dealing with.

Read some power supply manufacturer's application notes, and manuals for the supply in question. HP has some excellent app notes.