Switching vs. linear power supply??

[bomscho]

So I have an Astron VS-35M, and it's been great for me. Now I need something bigger.

I thought I HAD to use a linear power supply for ano. The shop that sold me the Astron linear told me they're being phased out by the EPA b/c they're very inefficient (20%-50% efficiency), whereas switchings are 90% and accomplish the same thing.

Astron is the only co. that makes linear anymore, for now. It was a radio gear shop that advised me, but they sell to a lot of platers and they said switching is fine for plating. I'm only after ano.

He said switching technology has improved so much that they're actually better than linear now, and not just in efficiency. It used to be you couldn't regulate voltage and crossbar (?) -- amperage -- on switching supplies, which he said was why folks used to favor linear. He told me "DC is DC"...true?

I saw in another thread that Acidrain bought an Agilent power supply that was probably switching...how was that for you?

Switching supplies weigh less, you have more choices on manufacturers, and they're cheaper, but it should be established decisively that they're ok for our applications -- especially if linears are dinosaurs soon to bcm extinct.

Does anyone else have a switching supply that works fine? What voltage and amperage? I'm on the east coast and can't wait forever for a 50 lb supply to drop ship from California. Thanks!!

[woodjames]

Linear versus Switching Power Supply for Plating

So I have an Astron VS-35M, and it's been great for me. Now I need something bigger.

I thought I HAD to use a linear power supply for ano. The shop that sold me the Astron linear told me they're being phased out by the EPA b/c they're very inefficient (20%-50% efficiency), whereas switchings are 90% and accomplish the same thing.

Astron is the only co. that makes linear anymore, for now. It was a radio gear shop that advised me, but they sell to a lot of platers and they said switching is fine for plating. I'm only after ano.

He said switching technology has improved so much that they're actually better than linear now, and not just in efficiency. It used to be you couldn't regulate voltage and crossbar (?) -- amperage -- on switching supplies, which he said was why folks used to favor linear. He told me "DC is DC"...true?

I saw in another thread that Acidrain bought an Agilent power supply that was probably switching...how was that for you?

Switching supplies weigh less, you have more choices on manufacturers, and they're cheaper, but it should be established decisively that they're ok for our applications -- especially if linears are dinosaurs soon to bcm extinct.

Does anyone else have a switching supply that works fine? What voltage and amperage? I'm on the east coast and can't wait forever for a 50 lb supply to drop ship from California. Thanks!!

[BobFrapples]

I use a switching power supply. Works fine. There is less ripple and ~V so its more controllable and tolerable to any changes in the solution.

Just make sure you use a diode on the positive end of it, to ensure nothing kicks back at it.

[bomscho]

@woodjames, i did find that thread. problem is the info is all very contradictory, even on here. and this guy, from a shop that didn't steer me wrong before, told me he's been selling lots of switchings to platers, and they're happy. hmm.

@Bob, what do you mean a diode at the end?

i was hoping for something around 50a/30v, dismayed that nothing is less than $950 or so. and i want a solid product that will last. after reading that acidrain cooked 3 Mastechs i don't think i'm heading towards that brand.

[dmiom]

So I have an Astron VS-35M, and it's been great for me. Now I need something bigger.

I thought I HAD to use a linear power supply for ano. .....

For radio use, there was a preference for linear supplies because the high-frequency switching in an SMPS can be radiated/conducted as interference - but that's of no concern for anodising.

Regarding "It used to be you couldn't regulate voltage & current on switching supplies" - efficient voltage/current control has always been one of SMPS strong points ! The ONLY aspect, again in a radio-specific environment, is that older SMPS can have less reserve and be slower to respond to a step change in load - think of keying a 500w or 1kw transmitter - an old SMPS might suffer momentary output voltage droop. Again, since since you ramp the current/voltage up to avoid contact point burn; and there are no step changes thereafter; this won't affect you when anodising either.

Otherwise, I think there may be a certain amount of folk mythology in what your radio man has told you.

Don't understand reference in that context to "crossbar" - two un-related things do come to mind (1) a technique for tripping-out a faulty power supply due to over-voltage on the input is crowbar - effectively short it and blow a fuse - but that has no relevance here; or (2) maybe he was alluding to trying to common-up two power supplies on a bus-bar for extra current capacity? - that will be a problem with SMPS, but almost as much of a problem for a linear supply.

Have a look for a lot more info & discussion in this thread here on linear vs. switching supplies

Dave

[BobFrapples]

If you view Dave's link he gave you, he fully explains switching and linear power supplies, its great information and should anwser all questions.

By a diode in the positive end I mean on the positive wire heading to your bus bar, input a diode (make sure its the right way or it won't work! and make sure its a high enough rating, double the amps and volts I usually do). Input the diode, and continue to your bus bar. It will look like:

P.S. ----- Diode ------Bus Bar (Positive)
P.S. -------Cathode (negative)

That's how your two wire should be set up. Diodes are cheap, I have a 400a 600v blocking diode for $12

[woodjames]

Not sure what to tell you, I don't have the knowledge of PS's to help-I just use them; I just remembered that post and linked to it. When you said "but they sell to a lot of platers and they said switching is fine for plating". "A lot" is subjective (and how would they know they were happy?)- there are two large rectifier dealers that come to mind, KingSupply in Franklin,Il and David Riddle in Van Nuys, Ca. They are both large operations dealing with many many brands and sizes of rectifiers, I'd get an answer from them first. At first I thought that the linears were the only way to go for large size, but then I checked out King's site and they advertised a switched 1000 amp as "They take up so much less space than a traditional unit, this is the Future of plating!".

$950 for only 30amps? Keep looking. My power supply puts out up to 250 and wasn't nearly as much as that, albeit slightly used. Never have a problem with it, same one for 2 years now. I mounted it into the wall, which rears to an adjacent room. I've got some fuming acids that I work with that reek havoc on everything, I'm concerned about those ruining the unit, not the unit burning itself out. They sell conformal coating in a spray, have yet to go that route yet, but I feel pretty good about the installation protecting it from the fumes.

@woodjames, i did find that thread. problem is the info is all very contradictory, even on here. and this guy, from a shop that didn't steer me wrong before, told me he's been selling lots of switchings to platers, and they're happy. hmm.

@Bob, what do you mean a diode at the end?

i was hoping for something around 50a/30v, dismayed that nothing is less than $950 or so. and i want a solid product that will last. after reading that acidrain cooked 3 Mastechs i don't think i'm heading towards that brand.

[bomscho]

Well...thanks all! And that link (why can I never find these threads?) was very helpful, dmiom. Bookmarked so I can link it for someone when the question comes up here again

So that's helpful, now if I can just find a switching PS that can ship to me before week's end. Found a lot of sites but there are so many brands and so many features as well, most probably irrelevant to my uses...at least I know it doesn't have to cost me $950, which I definitely do not have right now.

Are there any PS features that could substitute for putting a diode in my positive line? That would be a pain because my PS connects directly to my tank by battery cables.

[dmiom]

.....Are there any PS features that could substitute for putting a diode in my positive line? That would be a pain because my PS connects directly to my tank by battery cables.

A few points on Diodes which may help :

1. Firstly, if you are going to use one, just think of it as a non-return valve. it can be anywhere in the electrical circuit, it needn't be immediately adjacent to the busbars. It may well be the best place is at/near the PSU end of the leads, as that is probably the best environmental position. Also, so long as you connect it with the appropriate polarity, it makes no difference whether you connect it in the +ve or -ve lead - all you're doing is fiting the non-return valve in the flow or return line.

2. Do you need a diode - well to be honest, I've yet to be convinced (we don't use them here on either SMPS or linear supplies). With the exception of the point (4) below, there isn't really any disadvantage (provided the diode is of sufficient rating). Have never seen the electrochemical cell EMF from the tank cause any problems - the only thing I can imagine is that the minute standing voltage across the cell might be detected by some SMPS at startup and inhibit its operation.

3. Does a diode affect constant current operation? - no, it shouldn't. A diode acts as a non-return valve, but in common with most real-life fittings, it does provide a small disturbance to the flow - in the diode's case there is a voltage drop, which is typically of the order of, or just less than, one volt. This does generate some heat, and you may need heatsinking to keep it reliable. The constant current power supply will simply increase the voltage it supplies in order to push the commanded current through the load circuit.

4. The possible, but important downside - the effect of that voltage drop on an under-specified power supply. Using the Caswell LCD process, and the 720 rule (especially the on-line calculator), you get not only the current required, but also an indication of the likely peak voltage. That is the voltage across the anodising tank at PAR. As a rule of thumb, you need 'a handful' of volts in hand to make sure your supply has sufficient headroom to always push your commanded current. LCD peak voltages can be of the order of mid-teens - so to be safe for LCD operation you might want to look for a PSU which can run at least 17 - 20 volts. There are ongoing reports of those who find their supplies (maybe using a lab supply which can't go high enough) max-out and revert to CV operation because their voltage rating isn't high enough. If you then compound this by adding a diode in the circuit which just takes another volt or so and just throws it away as heat, then IF your supply is of limited voltage ability, you will find your useable current density reduced even further. So, if your PSU has plentiful voltage headroom, then there's no downside - but if its of limited voltage capacity then beware.

Dave

[bomscho]

Thanks, Dan. I've been back and forthing with several online used equipment vendors or brokers, and the results are not encouraging. Without rewiring my studio, I can't do more than ~115VAC supply on a 15-20a breaker. Doesn't seem you can get the kind of v/a i need from that input.

Moreover, I still haven't seen anything less than $900. A couple Sorensens somewhere I think, but that was it. Not a lot of choice so I might have to grab one of them.