[Bob Katz]

Prove me wrong 🙂

Two days ago I purchased a bargain headphone amplifier on special for $68 that is all-discrete, direct coupled and uses a 36 volt dc external switching power supply. It’s the Liquid Spark, on special at Monoprice. Monoprice is a scary place, but in this case they did very well

I shot the Liquid Spark out against my best headphone amplifier, the Audeze Deckard, all discrete, direct coupled, solid, humongous Class A, that sold for a whopping $700! It has an internal linear power supply. Amazingly, the two amps proved extremely close, with the Deckard sound a bit warmer and more robust, with a more solid bass

Measurements showed the two amps to be neck and neck. Both are flat response, -1 dB at 10 Hz. I think the difference in bass quality can be attributed to stiffness of power supply. These days with 18 ohm headphones you have to have a good backbone to keep them happy.

So I took the Liquid Spark to my bench and added two 1000 microfarad/50 volt caps across +18 v and ground and -18 volt and ground. The result: totally amazing. The cheaper amp now sounds warmer and fuller with even more solid bass. The $68 amp now sounds very much as good as the $700 model! I can’t believe I’m listening to a $68 amp!

It’s the Power supply – folks!

[Gregory Pastic]

I know very little about electronic design, but I remember Chris Russel of Bryston stressing that same point about the power supply way back in the early 90’s.  I think they were one of the first companies to use outboard power supplies for their pre-amps too.
Now I’m curious enough about Bob’s enthusiasm for this little headphone amp to buy one myself!  I’ll have to find someone locally to do the mod for me though…LOL

[Bob Katz]

There are a few reasons why companies are using outboard supplies. One of them is very practical: it costs to get Underwriter’s Labs approval for gear with wall AC inlets. So you can buy an external wall glomper and be UL approved.

another good reason is for high gain preamps. Locating the power transformer externally reduces the chances of hum induction from the transformer field.

[Luis Berrios]

Great. Now I have an assignment.

[TiKkO Rome]

Could you share a little more detail about the modification? I’d love to try this out for myself.

[Bob Katz]

Yeah, get two 1000 uF at 50 volt (or 25 volt) electrolytic caps and solder one (properly polarized or you’ll make blue smoke) between the +18 V point and ground on the little marked strip and the other between -18 v and ground. Use normal precautions: Measure voltage on the terminals first, then turn off power. If you are not experienced at basic circuit design, do not attempt this mod yourself.

[Larry Oppenheimer]

I want to second this motion and extend it to say that the power supply can affect things in all sorts of devices in ways you would never imagine. In an earlier life, I was a bench tech, first at Lexcion (I’m talking WAY back), and, later, for an importer/distributor of European electronic instruments. I can’t tell you the bizarre problems that I eventually traced back to power supply problems. Really subtle stuff like missing menu commands, intermittent glitchy sounds, and on and on. Again, these were *digital* devices!

And, again, for power supplies in amps. I spent four years at Meyer Sound, and one thing John Meyer talked about was that some amplifier companies would cite the maximum voltage output of their amps in order to indicate their headroom and ability to respond to high-peak transients, but, he pointed out, they often could only *maintain* that output for a few milliseconds or even just some number of microseconds, where a transient could have a high level last for up to tens of milliseconds. Thus, such amps could reproduce the transient *level,* but not necessarily accurately reproduce the actual *transient* itself.

[Gregory Pastic]

Is there a particular capacitor brand you would suggest, or are they all the same quality?

[Bob Katz]

Larry, that is such a valuable and important point! Regarding transients, you can measure static or steady state response and harmonic distortion, but that doesn’t tell you how well the amplifier will hold up during transients. My power supply modification for the Liquid Spark headphone amplifier really made a sonic difference. I did not measure a before/after, I only listened. Perhaps to run a low frequency square wave and look at the tilt would tell us if an amplifier’s DC supply is a little saggy.

[Phil Koenig]

Not just amplifiers…

Clock generators can suffer increased jitter.  (I spent the final 10 years at my day job designing clock circuits for microwave data radios, and jitter reduction represented a major portion of my efforts).

[James Johnston]

There’s a number of issues here.  The results may be different for standard full-wave rectifiers vs. switching supplies, for instance, but in both cases, more storage is almost always a win, the question is “how much more matters”.

For standard rectifier supplies, you can get very annoying 120 cycle-shaped limiting with a weak power supply.

For switchers, the same problem can occur, but not at the “power supply output” rather at the first stage after line rectification happens, if there isn’t enough capacitance to hold up the regulator control range.

So where the capacitor needs to go, and what size and voltage it requires, can vary.

However, adding on millifarad of the right voltage across each pole of the supply to center should never hurt (for low voltage supplies, that is!).

My thought, though, if you’re going to do this (old analog research hardware designer in me is coming out here) go large. Capacitors are cheap.

[Bob Katz]

Gregory asked “what brand” of cap I would recommend. It’s not an audio coupling cap so I think the ESR of the electrolytic is not that important. I usually buy Panasonic electrolytics from Digikey. Any well-known, reliable brand of electrolytic should be fine. I do recommend getting 105 degree caps as opposed to 85 degree caps, they are more rugged and will last longer under heat.