Laptop power supplies for amps

Laptop power supplies for amps??

I was at a DIYer's house recently and noticed what he explained to me was a Nelson Pass Zen variation 5 and right there in the amp was 4 laptop power supplies (the bricks you throw in your bag for when the battery dies...). 2 per channel powering each of the push-pull.

Now this amp design is a class A, so I am guessing its always pulling a lot of power.

So it got me curious and I did some research.
1.) on eBay you can buy 90 watt 19 volt power supplies all day, and cheap. They can deliver almost 5 amps each! Seems like enough if I'm intending to use it for a low mids and up on a tri-amp'd set up (horns, so no real need for lots of power).
2.) Nelson Pass himself says the PS in this amp has to be really clean.

So has anyone else done this? Just how "clean" are laptop supplies? What/how much would you need to do to clean up the output?

Thoughts.

P.S. - I didn't have a chance to hear it, but he said he and another audiophile thought very highly of its sound.
 
Last edited:

paulb

Member
2001-06-01 4:53 pm
Calgary
Nelson Pass says (Amp Camp #1 article):
I addressed this by choosing a commercially available switching supply of the type you routinely see powering up your portable computer. There are literally tons of surplus supplies available for this purpose, providing regulated 19 volts DC at more than 2 amps or so on their output and having their input going safely to the wall AC power outlet through a safety approved power cord. These supplies are isolated for shock safety and are also short-protected. At 19 volts output they do not represent much of a hazard to humans.

What? An audiophile component with a switching power supply? Get over it – it works fine.
 

Jay

Banned
2003-02-11 9:02 am
Jakarta
For expensive amps I will not use laptop's supplies, or switching supplies in general. I know that I don't understand this power supply technology. I may miss one or two tricks on how to use it.

Grounding, I still don't understand it. The relation between the switching supply ground and other components' grounds, I don't know how to treat them well.

Last night I built a cheap chip amp (but using expensive and exotic components). I wanted to use IBM laptop supply so that the amp is as small as possible. Prepared to use RC or LC on board. No problem. So I removed the R, still no problem. I guess the on-board bridge rectifier acted as R (I connected the regulator output to bridge diode).

I think it is cleaner with the supply (because IC amp has terrible PSSR so I think the IC benefits from the regulator). Couldn't hear fatigue either.

But I guess transformer supply with simple pre-regulator will be better for my application.

As you said, the used laptop supply is very cheap and capable of very high current. If it is as good as conventional supply, everyone would have used them everywhere. I have tried and modified many, and I always found that such supply is a headache (talking about high-end here).
 

Jay

Banned
2003-02-11 9:02 am
Jakarta
Last night I built a cheap chip amp (but using expensive and exotic components). I wanted to use IBM laptop supply so that the amp is as small as possible. Prepared to use RC or LC on board. No problem. So I removed the R, still no problem. I guess the on-board bridge rectifier acted as R (I connected the regulator output to bridge diode).

I think it is cleaner with the supply (because IC amp has terrible PSSR so I think the IC benefits from the regulator). Couldn't hear fatigue either.

But I guess transformer supply with simple pre-regulator will be better for my application.

I just had a chance to listen to my chip amp with bigger speaker (10cm wide :D ) and bigger volume. The sound is very awful with IBM laptop power supply. With around an ampere of 10V transformer the sound was much much better.
 
In one sense, it's easier to get cleaner power from an SMPS, because none of the stuff you need to get rid of is in the audio band. With a CLC filter, you can reduce the spikey SMPS artifacts to a few microvolts.

For the CLC, Mouser.com, and many others, have those nice JW Miller / Bourns toroidal inductors that have extremely low DCR and can carry very high currents, such as their 2300-series, a little over an inch in diameter and about 1/2-inch thick, which are available from 10 uH to 1000 uH, with corresponding DC current capacities from 20A to 2.4A, and DCRs from 0.005 Ohms to 0.3 Ohms.
 
So has anyone else done this? Just how "clean" are laptop supplies? What/how much would you need to do to clean up the output?

I'm fairly sure that SMPSUs are the future for amp supplies, can't beat the bang for the buck as they've got huge economy of scale now.

There are a few things to watch - common-mode noise being the biggest. Also check the PSRR of your amp at high frequency, you probably should be using chokes and/or ferrites in filtering to kill all the HF hash you can. I'm using my own custom-wound CM choke on the input to suppress common-mode noise - happy to provide details on request or have a look at my blog (link on the left of this post). Oh and I'm also using ceramic decouplers on the power supply to my amps. Listening result is very impressive - I'm getting sound with a cheap chipamp now that I previously needed a whole (discrete) integrated amp to achieve.
 
Mixed experiences

I am using a laptop power unit to feed my Asus eeePc with mpdpup, itself feeding WaveIO USB Spdif converter to DAC.
- first try with a generic powersupply - awful sound very very noisy - I had to listen on batteries
- replaced it with a Sony Vaio power supply with a tweak for connectors, all the noise gone compared to battery supply, and it sounds good. Probably better designed SMPS.

Can someone draft a filter to get read practically of the noise, including how to calculate the components, as when you buy a SMPS you do not know before if IT has been adequately designed...

Also I am now using a USB cable with ferrite beads. I did not notice sound improvement compared to no beads cable. Is it nevertheless generally speaking a good approach to use bead equipped cables ?

BR
Jean-Louis
 
You won't find a 'one size fits all' approach to getting rid of the noise, you just have to work to improve your knowledge and experience of noise reduction. I found this article by analog design guru Jim Williams very useful in relation to getting to grips with SMPSU noise : http://cds.linear.com/docs/en/application-note/an101f.pdf

I'm a big fan of ferrites (as is Jim Williams in that article, ferrite beads are the most effective noise reduction with SMPSUs as you'll note if you study it) so yes, use as many ferrites on your USB cable as you can fit on. In my opinion, better to spend your money on ferrite clamps than on so called 'audiophile' cables.
 
Yes, that what the "theory" says. When your hands-on experiences are really against it you will start to think that the theory is not as simple as it looks, or...?

I rather suspect that gootee has yet to discover the hidden perils of common-mode noise (because he's talking about normal mode noise in that post) and that you have, by experience Jay? ;)