DC Linear Power Supply : Pro or Audiophile device ?

[tgb]

hello,

I reviewing the market of linear power supply for my digital source.

I found quite a udge difference in price between audiophile devices like Teddy Pardo PSU types (roughly all are around 300-400$ with only 1 output, some are much more expensive) versus "professionnal" devices like linear power supplies for use in laboratories (these products can offer up to 3 outputs for the same price of only ONE Teddy PSU !!).

 

The trick is that :

- Teddy fo instance gives very little info about the quality of the DC output current. They (and customers) just say that it sound "nicely".

- Pro devices give more info like the voltage ripple at the output. For instance, rather good products give informations like a voltage ripple of <2mV at the output.

 

Questions :

- did anybody tried these "pro" linear PSU ? Difference vs a Teddy type PSU ?

- a voltage ripple of <2mV seems to be very low, according to me. Is it suitable for audio ? What is the voltage ripple of audiophile product (data I could not find anywhere), and is it much lower than these tiny 2mV max of pro devices ?

 

I found a laboratory linear power supply with 3 outputs (perfect for my NAS + interface + DAC) at 350$. I wonder why paying 3 times more for 3 audiophile products with only 1 outputs...

 

Thanks in advance if you have info about it.

Regards

[mav52]

Can you provide some examples of these ""pro" linear PSU's" manufacturer & model number as I would also like to look into a external power supply.

[4est]

There are different grades of lab supplies. 2mV is not very good to be honest, but it does not tell the whole story either.

[ALRAINBOW]

I do not have an answer but I know who will. Steve nugent of imperical. Audio.

fined his website audio circle and post it. I paid about 750 for one PSU from him. Yes it makes a difference but would one like you are describing work just as a well ? My thoughts yes , but there may be other measurements that matter . An example before I purchased the dynamo from Steve I was going to use a battery . Lipo or life . Steve said it works but his is better . I asked why as I am aware of noise in batteries , his answer was not noise it was the battery could respond as fast to current demands as the linear PSU . And he did sell battery PSU , s at one time . So in the end maybe it's splitting hairs , not sure but we must go with someones advice or our ears. Post some links to the PSU your are posting about both the audio grade and the lab grade. Another point might be the output impedance might not be low enough in the lab grade .

Al

[tgb]

Thanks fo your feedbacks.

I agree with you Allrainbow, "pro" devices do not give any idea about sound quality... producers did not test that point simply because they do not care :-)

 

This is an example of what I talk about :

VOLTCRAFT VLP 2403pro, 252W Triple Output Variable DC Power Supply, Linear, Bench from Conrad.com

Quite interesting in theory because you have 3 outputs for the price of 1 Teddy !

Note that I have a look at theses products because you can manage different voltage outputs ; quite useful for audio because a NAS + interface + DAC need quite always specific & different voltages. With audiophile products you need 1 box / voltage ; some audiophile PSU can be done with 2 different voltages but price rises a lot.

Rgds

[bluesman]

This is an example of what I talk about :

VOLTCRAFT VLP 2403pro, 252W Triple Output Variable DC Power Supply, Linear, Bench from Conrad.com

Quite interesting in theory because you have 3 outputs for the price of 1 Teddy !

Note that I have a look at theses products because you can manage different voltage outputs ; quite useful for audio because a NAS + interface + DAC need quite always specific & different voltages. With audiophile products you need 1 box / voltage ; some audiophile PSU can be done with 2 different voltages but price rises a lot.

Rgds

 

Those lab power supplies are generally excellent for stability, but they're made for reference-stable voltage under load rather than for audio silence.

 

One down side of a lab PS like that is that the voltage has to be dialed in, so you need to protect the knobs from accidental "readjustment" or you could damage a circuit. But it's intriguing for sure. I may run over to our biomedical engineering department and see if I can borrow one to play with.

[Superdad]

Indeed there are a TON of inexpensive bench "lab" supplies out there. You can pay $50-$400 for them. They look a little fancy from the outside. They are mostly all made in China as cheaply as possible. If you want a high quality bench supply that truly performs, then look to Keithley or some of the other test brands who have been making precision supplies for decades. But be prepared to pay dearly--and for performance parameters such as output accuracy, temperature stability, and tight line/load regulation--which are not of that great of importance to our music computers and DACs.

 

Ripple, noise, and transient load regulation are what is important to us. Rejection of AC line noise is also important. Most of those cheap bench supplies are offering totally incomplete specs. Stating "<2mV ripple" tells us almost nothing. Are they talking about just the AC ripple (at whatever line frequency--50 or 60Hz--you are plugging into), or are they including all the wideband noise on the output? And over what bandwidth?

 

Really most of those supplies are craptastic. Physically noisy transformers (though some brands are okay in this respect), cheap (and spiky) bridge rectifiers, inadequate or poorly chosen pre and post regulator capacitance. Yes, some of the higher current units (I'm talking 10A and up) are using big heatsink-mounted transistors for discrete regulator circuits (they have no choice at those amperages), but that does not mean they are any cleaner/quieter.

 

I still admire how inexpensive all those supplies can be, plus they include those nice meters, knobs, and the circuitry to support them. Ah, the wonders of mass production economies of scale. Still, I think 2A-3A is a little lightweight for music server/small DAC application. Maybe look in a bit higher range, say 4-5 amps. I'll probably pick up one of the popular 5A Chinese "lab" supplies just to have John Swenson measure it versus our forthcoming dual-rail, choke-filtered unit. Comparative scope shots of noise and transient load capability should go a long way in helping folks understand the differences between a generic and a purpose built design.

[JakovH]

How is your LPSU coming along Superdad?

[iago]

This is an example of what I talk about :

VOLTCRAFT VLP 2403pro, 252W Triple Output Variable DC Power Supply, Linear, Bench from Conrad.com

 

I'm using exactly the one you linked to as a voltage source. It usually supplies my audio computer, the USB-to-S/PDIF converter and a RB clock. 3 A are sufficient for smaller computers without SSD components; my desktop computer overloaded this PSU (even in bridged mode). Lab power supplies are not capable of over-current, because they immediately throttle the supply voltage. For larger computers, especially SSD components, you need a large current overhead capability.

 

The ripple and drift specifications from the manual are met. Stability over time is sufficient in normal mode, in bridged mode drift is perceptible long term.

 

This component replaced one plug-in PSU plus USB bus power and made a noticable difference (sound stage is more realistic). My impression is that it's not good enough to supply clock circuits with.

 

The build quality of this component not as good as found in vintage gear (also something you might consider). Internal wiring is not up to audiophile standards and the transformer is very sensitive to DC on the mains and tends to hum audibly; fortunately it's not noticable from 6 ft. away.

 

Conclusion: the Voltcraft PSU does its job, the performance-to-price ratio is very good, but next time I would look for something used from Agilent, Gossen or Philips.

[bluesman]

Conclusion: the Voltcraft PSU does its job, the performance-to-price ratio is very good, but next time I would look for something used from Agilent, Gossen or Philips.

We used Bruel & Kjaer in the labs where I was a resident, but it appears that they're no longer making a full line of lab electronics. The original B&K equipment was absolutely tops (and priced accordingly). If you can find a used piece of B&K, it's well worth grabbing.

 

I see a "B & K Precision" that makes linear lab power supplies, and they may be a modern offshoot of the original B&K (although the pictures of their products don't look anything like the original B&K). The specs do appear to be better than the VoltCraft at a slightly higher price. I have no experience with this line, but it looks interesting enough to make me want to buy one from Amazon and see if it improves the sound from my projects.

[Superdad]

I see a "B & K Precision" that makes linear lab power supplies, and they may be a modern offshoot of the original B&K (although the pictures of their products don't look anything like the original B&K).

 

B & K Precision is an American maker of moderately priced but decent (at least they used to be) test equipment. Power supplies were not their specialty. They made a wonderful VTVM (with a beautiful, easy to read needle meter) that was extremely popular (I'd like to get one again).

 

Of course Brüel & Kjær is the old-line Danish company famous for their measurement microphones. The quality--and look--of most everything they made was first-rate.

 

My favorite brand of vintage power supplies is Power Designs in New York. Amazing pieces inside and out.

 

If you want to ooogle vintage bench supplies, there is a long and fun thread over at EEVblog:

Show your favorite and most used benchtop PSU - Page 1

[iago]

We used Bruel & Kjaer in the labs where I was a resident, but it appears that they're no longer making a full line of lab electronics.

 

Brüel & Kjaer were very well known for acoustic measurement devices; the first gear I used for speaker measurement was made by B&K. B&K are still very renowned in acoustical and vibration measurement. I was not aware that they also made power supplies.

 

B&K Precision seems to be a US only company, which makes buying their equipment in Europe rather difficult /expensive. From what I have seen, their 3 rail PSUs have similiar prices to that of the other prime vendors (around 1200 EUR). The equipment sold on Ebay looks different from what you find on their home page, though.

 

Apart from the comfort of using a bench power supply (voltage at your finger tips), I'm currently thinking of building my own PSU for a change. Application note 3657 from Maxim Integrated looks quite interesting.

[iago]

My favorite brand of vintage power supplies is Power Designs in New York. Amazing pieces inside and out.

 

This one looks really cool. It would make a nice addition to an Audio Rack.

[Superdad]

How is your LPSU coming along Superdad?

 

Extremely well, thanks. Sent $4.3K to Japan last week for the first 25 unit run of customized chassis; the dry fit went well and production PCBs get going next week. Chassis arrive at the very end of May/first week of June; all other parts, transformers, copper heat sinks, cartons, etc. should be ready and waiting before then. I'm excited!

 

[Superdad]

This one looks really cool. It would make a nice addition to an Audio Rack.

 

Yeah well, they are pretty on the front, but not so much the rest of the casework. Plus, they are huge and very heavy.

(The cute one I posted a pic of before is only good to 0.5A)

 

[bluesman]

Brüel & Kjaer were very well known for acoustic measurement devices; the first gear I used for speaker measurement was made by B&K. B&K are still very renowned in acoustical and vibration measurement. I was not aware that they also made power supplies.

The Audiology research labs at Penn had a ton of B&K and other top shelf stuff in the mid-'70s, including signal generators, analyzers, amps, displays, power supplies, etc. I don't remember which was what, so the power supplies could have been part of other components. I suppose they could even have been from a different source - all I remember is the magnificent quality of every single piece in that lab. I think we even had open reel tape decks labeled B&K, but it's been 35 years since I was a resident, and I spent little time in that lab because I had little interest in Otology - I only went there to ogle and play with the equipment.

 

BTW, all amplification in the clinical audiology suite was McIntosh, from mono 30s to MC275s.

[Animaniac]

Astron Power Supplies RS-5C - Free Shipping on Orders Over $99 at DX Engineering

Would this be any good for a celeron 1037U based computer? It could be 18W or even 10W depending on the board. 3 pin XLR output and heatsink, the voltage can be internally dialed in to 12V. I've built a DIY balanced power conditioner and gainclone amp from kits, but I can't even read a schematic so I shouldn't be messing with this stuff! Still, I'm curious.

[Superdad]

Astron Power Supplies RS-5C - Free Shipping on Orders Over $99 at DX Engineering

Would this be any good for a celeron 1037U based computer? It could be 18W or even 10W depending on the board. 3 pin XLR output and heatsink, the voltage can be internally dialed in to 12V. I've built a DIY balanced power conditioner and gainclone amp from kits, but I can't even read a schematic so I shouldn't be messing with this stuff! Still, I'm curious.

 

I have and used for a while the Astron RS-12A. It is 12 amp bigger brother to the little 4A unit you are looking at. About $30 more, though I got mine used for $50.

 

I don't think you will like it as the Astrons hum a LOT (mechanically) due to use of cheap EI transformers. The hum of mine modulated up and down with the load from the Mac mini.

Also, since it is 13.8V, you will have to open it up and try to get at the small potentiometer that adjusts the voltage, but be careful not to shock yourself! It may take a few turns of the pot to get it down to 12V. On mine, the pot was very hard to reach.

[InfernoSTi]

I have tried a couple of industrial LPS and an audiophile LPS and would rank them as audiophile first, nicer industrial second, cheap industrial third, no LPS (i.e. cheap switching PS) fourth. Just one person's experience…YMMV.

 

John

[Animaniac]

Wow, HN12-5.1-AG Power-One | Mouser

Thanks guys. The comparable model by Power One is $100.

Here's a funny thread about ham radio folks trying to get rid of the Astron hum. One guy managed it with varnish: [HAM] - Astron hum/vibration

I got a case of hum when I plugged in a Blue Circle power line pillow into my balanced power conditioner, the nice US rep said that it may be due to the capacitor frequency resonating with the secondary winding. It's silent now, thankfully.