sound practices reading club /45

Only the second instalment of the phono preamps topic and we go off on an expedient tangent. Last week jc morrison told us ‘Personally, I despise the effect voltage regulators have on tube preamps.’ So today we will check them out. From the same sound practices issue 3, we are reading two articles (I am pretty sure Joe meant it that way). First up, there is One tube Regulator by Mike Vans Evers. In just two pages, Mike does a good job of explaining how his regulator works and how you can adapt it t your needs. There is a bonus line stage schematic.

teaser quote: ‘Tube regulators were invented in the 1930s to replace the bulky battery power supplies used with scientific instruments. There were many types, some simple, some complex. The regulator presented in this article is of the simple variety: one tube.’

The second article is Classic Voltage Regulators by Alan Douglas. In just two and a half pages, Alan does a good job of how series regulators work and how they scale up to industrial size. There are several recommendations for (US) industrial brands.

teaser quote: ‘Why a regulated power supply? What’s wrong with a simple transformer, rectifier, and “brute force” filter? Maybe nothing. But regulated supplies have several advantages over simple filtered ones’

my take
What both articles make perfectly clear is that a regulator is nothing but an amplifier. Depending how the regulator is coupled to the audio circuit, and what the power supply rejection ratio is towards this regulated supply, one is listening to this extra amplifier singing along. I am convinced that one should apply the same design aesthetic—for the circuit topology—towards regulators and amplifiers. The requirements for a regulator make it hard to avoid loop feedback, but let’s show some style. Ending up with a regulator circuit that one would never, ever feel proud of as a power or line amp: it does not have to be that way.

What I like to add is that a regulator performs three different jobs, two of which are optional:

1. reduce hum and buzz. This seems to be always part of the deal.
2. stabilise the DC voltage. Optional. When the DC is stabilised, the regulator is amplifying a voltage reference. In the first article this is the zener, in the second a VR tube. We see from the first article that the problem has moved to obtaining a stable and quiet voltage reference, which may need extra filtering to be up to the job. Typically these references are regulators themselves. And thus the beat goes on. Also when DC is stabilised, some careful calculations need to be done that take the fluctuations of the AC coming from the wall socket into account. Depending on where you live, this can be quite a bit. Both too little or too much difference between the input and output voltage of a regulator can damage it. When the DC is not stabilised, the regulator simply takes XY% of the input voltage; filters it and then buffers or amplifies it to the regulator output. It just wobbles along with the AC line voltage.
3. cycle the currents of your audio stage. Optional. Certainly with single ended circuits, the power supply closes the audio current loop. Usually this is a capacitor and you will be listening to its non-linearities. The second article points this out. Both articles intent that the regulator takes over this job. The schematics pictured may have a cap over the output but that one is to help out with high, maybe even only ultrasonic, frequencies. Thus the music is going through the regulator and you will hear it. You may like it, you may not. However, if you apply RC or LC decoupling after the regulator, you will (depending on the time constant) for part or all of the audio spectrum not hear the regulator. But you are listening to a cap again. The regulator will still reduce hum and buzz, and optionally stabilise the DC voltage, see above.

practical note: in the first article, the output cap of the line stage says ‘the more the merrier.’ How big the cap should be is dependant on what follows the line stage, but even for 600 Ohm headphones 50µF is enough. For coupling to a power amp, a way smaller cap will be better. Big time constants simply result in crap treble. The Kepco 815 schematic should not be followed blindly. It already says ‘simplified circuit,’ and when I looked at the C- output circuit, the shown resistances, voltages and currents do not add up.

bonus tracks
Want to read more on regulators? Then move on to SP issue 6. The articles above deal with series regulators, John Camille, in 211 Amplifier: Part 2, explains shunt regulators very thoroughly.

Now go and read the articles, see you next week.