Mic PSU - Regulated or unregulated?

[shot]

I went through a lot of threads here and elsewhere on net trying to figure out pros and cons of unregulated vs regulated power supplies.
In almost all applications regulated is the preferred way, be it linear or SMPS, but it seems that microphones prefer having unregulated PSUs instead.
On the other hand, we have few threads here that deal with repurposing chinese mic power supplies and I haven't noticed anyone says anything bad about them.

So what is the deal?
Why do 47, c12, 67, 49 etc. clones all have unregulated PSUs, but done without tube rectification? If they were to be historically correct shouldn't they have tubes all the way, even in PSU?
And what is wrong with building regulated PSU using LM350, TL783 and similar regulators?
Isn't it after all that we need clean and stable voltage and enough current, no matter what?



Luka

 

[Khron]

How about the absolute lack of double-blind testing, along with confirmation bias as well as (blind) vintage-philia?

 

[rock soderstrom]

Why do 47, c12, 67, 49 etc. clones all have unregulated PSUs, but done without tube rectification? If they were to be historically correct shouldn't they have tubes all the way, even in PSU?

Not all of the mentioned classics also had tube rectification in the original. For example the U67 PSU was always with diode rectifier, I think the C12 or N12k PSU were also based on semiconductors, as well as the U47M as far as I know. (which ones were with tube rectifier at all?)

And what is wrong with building regulated PSU using LM350, TL783 and similar regulators?

Nothing wrong with it. Some guys on the internet say unregulated power supplies sound better, but from my point of view this must be proven, see Khrons post! I am skeptical. Some regulators (e.g. 780X, LM317 for filament based biasing) tend to introduce noise, but with following CRCRC filter this will be fine.

Isn't it after all that we need clean and stable voltage and enough current, no matter what?

True, with clean being more important than stable from my point of view.

Microphone builders today are more conservative, rebuilding proven classic concepts. When the originals were created, mic builders were more progressive, using the latest technology they could get. Hence the sand rectifiers...

 

[gyraf]

mics generally consume very little power, making it easy to passive the supply all the way

 

[rock soderstrom]

mics generally consume very little power, making it easy to passive the supply all the way

That's true, especially for B+. Another argument for the all-passive solution is long-term stability. Active components tend to die a little earlier than resistors from my point of view.

 

[maarvold]

Adding simple zener regulation to a pair of cathode followers took the noise floor from "it's still bugging me just a little" to "ok now". Cathode followers are not great at power supply noise rejection.

 

[rubinstu]

I'm relatively new to the mic game, but here are my thoughts on this from a purely engineering perspective. No regulated power supply regulates perfectly, and no un-regulated power supply is totally wonky under a varying load. In theory, if your load (i.e. the mic preamp, etc.) does not change in use, the power supply type should not matter at all. So, a Class-A amp in theory draws the same amount of current regardless off the source -- in theory. In practice, a strong burst of signal may draw more power, which changes the operating point of the amp, which causes some distortion. The problem which analyzing this is that, whether people understand it or admit or not, this is what you want. These subtle distortions are what give the mics their character. I think a blind A-B comparison of supplies would be really interesting.

 

[shot]

Since I plan to build another 47 mic for myself, I'll definitely try and build a regulated PSU for it. That one could be used on both newly built and on the mic I already have. Just to be able to swap PSUs and do a test. That wouldn't be a double blind test, but at least I could measure behavior of different PSUs on the same mic.
But, it'll take some time for me to do this. No spare time nowadays at all.

 

[kingkorg]

Since I plan to build another 47 mic for myself, I'll definitely try and build a regulated PSU for it. That one could be used on both newly built and on the mic I already have. Just to be able to swap PSUs and do a test. That wouldn't be a double blind test, but at least I could measure behavior of different PSUs on the same mic.
But, it'll take some time for me to do this. No spare time nowadays at all.

Make sure you push those mics hard, i doubt something like this would be obvious on a guitar or vocal take.

 

[shot]

Make sure you push those mics hard, i doubt something like this would be obvious on a guitar or vocal take.

When you say "push hard", you mean like recording something loud?

 

[kingkorg]

Yes. Even loud vocal at usual distance don't realy swing plate voltage much, i presume high level signal would reveal more when it comes to psu regulation.

 

[Cqwet Dbdfte]

Subjectively a nice 2nd harmonic descending series may sound better. The ear makes it own distorsion.
The capsule may likely also produce its own artifacts.
If a stiff power supply with microvolt noise, or less, does not give "positive" results, maybe the power supply is not the issue.
The parts available in 1947 did not include reasonably sized, low ESR ~220uF caps.
You could the test the mike preamp without the capsule with a scope on the PSU and a suitable signal source, maybe some challenging music, and see how much gets into the power supply.
For a low power draw like a mike, this would not be a very challenging job. If a using a cathode follower with a gain tube and the cathode and plate resistors have the same values, their currents would cancel and minimize power supply fluctuations.
Again the right answer may be subjective.

 

[Delta Sigma]

Is everyone referring to B+, H+/-, or both?

B+; I can see passive vs regulated not making much of a difference. Especially since it doesn't matter if it drifts a bit. There's no real current draw, so there's no sag like guitar amps. Wonder if it's easier to clean up a regulated B+ though...

Heater supply; my assumption is that regulated makes more sense, especially if it provides bias like M49, U67, etc. It's been a long time since I learned PS design - though I'm trying to find time to re-learn no - but I think RC + regulator + RC would achieve better results for cheaper than RLC + RC. Those chokes are pricey, heavy, and take up a lot of real estate. Plus, a regulated supply simplifies powering relays.

 

[tomas.borgstrom]

I have only seen a tube rectifier in one microphone PSU and it was produced by Pearl in the 50's.

It might be confirmation bias but I can hear a little bit slower transient response in an unregulated PSU. Could it be because of very fast and sublte voltage swings that you can't detect on a regular voltmeter? The sag you hear on a guitar amp is easily audible.

 

[Cqwet Dbdfte]

One could put a cap on the power supply and run the signal into another amplifier and then measure and listen to the noise on the +B.