Dual power supply for tube mics

[Murdock]

What are the possible drawbacks (other than wasted energy when only one mic is used) for using a dual power supply for tube mics?
And what should one keep in mind when designing/building one?
A passive power supply for two mics is probably more problematic than a regulated one, right?
I know there is the Neumann NG-2 which is a double supply in one housing for two U47. It states that it has seperate rectifiers and "filter blocks" (Siebketten) but shares one power transformer. I could not find a schematic. But I recon one could not use only one U47 with that one as the voltages would than probably be too high as the load is less. Right?
Altough I found a description from someone who sold his: "this supply can operate one mic at a time or two with no change in audio or supply voltage as this supply uses two separate supplies, and two separate power transformer windings"...
So is that the solution when using a passive supply? Using one transformer with seperate windings for each supply?
That would probably not be needed when going active, right?
I sketched out some circuits in spice for heater and B+. Simulations show that it should work for either one or more mics. It's "inspired" by a few other circuits. So I just copy and pasted some circuit parts and modified it a bit.
One uses a mosfet and one a crude zener regulation. Both seem to "work". Altough my understanding of Spice is limited so I don't know how to see the noise specs...
The Heater supply does kinda work, but the range is limited... I tested it with a load of 100mA and 500mA. Voltage dropped 0.4V when using 500mA.
What do you think?

 

[Tubetec]

Some Neumann supplies used gas discharge tubes for stabilising the HT , they might draw 5-15ma , the 0.7-1.5ma drawn by the mics wont make much difference .

Using a seperate winding for each mic makes sense in the case of the dual U47 supply ,they still share power from the same primary and core so there would still be some small variation with only one mic connected compared to two , but it might be more or less irrelevent with all the filtering and series resitancance and a larger transformer , one of the downsides of the single rail tube mic means loads of power is dumped in the series resistance to the heater , as we know .

Its seems to me that if your making a supply for two mics , going to the trouble of at least using a seperate HT winding ,bridge ,filters on each mic is better , ultimately it means a better figure for crosstalk because theres less chance of interaction between the two mics .

As far as heaters go , why not use a seperate regulated rail per channel ,

Think a bit like the old fashioned dual mono set up , rather than stereo , so each mic has its own seperate grounding network all the way back to source .
The smart way to do it might be using the low profile dual primary.dual secondary wound power transformers , so some thing like 110-110 to to dual 120 or 150v secondaries for HT ,
And then 110-110v to dual heater windings of the appropriate voltage for LT .
It means two off the shelf transformers get the job done which is a major plus , your good either side of the pond also .

The one thing I dont like about the low profile transformers is there almost always designed for pcb mounting , even a short drop to the floor can cause the transformer to get ripped off the pcb with the possibillity of destroying the transformer pin terminination in the process ,
I'd favour a situation where the transformer is supporting the much lighter pcb ,
In otherwords frame mount the transformer to the chassis in the conventional way , design the PCB with the correct orrientation so the components fit neatly off to one side ,
The far end of the PCB can have aditional support added if needed .

One final thought ,
anytime you have more than one transformer in a box its worth considering how orientation of the core effects how they interact magnetically , using good layout and perhaps some extra metal sheilding here and there might be worthwhile .
Energising one transformer and using the primary of the second as a search coil into a decent multimeter you can zero in on 50hz radiation , and orientate the components for minimum interaction .

 

[Murdock]

Thanks for your input!
But is crosstalk really such an issue with a shared power supply?
There are loads of stereo tube preamps with common supply, e.g. Gyraf G9, Jontes Therapy Preamp, etc.
Not to deny your concern but just would like to understand more the pitfalls of such a design. Especially when using regulators.
I can totally understand the concern when going "passive". But do these problems also occur when going active?

 

[Khron]

I would argue the amount of crosstalk depends entirely (?) on where the common point is. The further "upstream" (ie. closer to the transformer, topologically speaking), the lower the crosstalk should be.

 

[Murdock]

I would argue the amount of crosstalk depends entirely (?) on where the common point is. The further "upstream" (ie. closer to the transformer, topologically speaking), the lower the crosstalk should be.

So you also say, that separate supply circuits is favorable? Or where would the common point be with a shared circuit? At the connectors?

What about those stereo tube mics (SM2, C24, USM69,...). They all also share a common supply. But crosstalk is probably not that of an issue because the capsules are so close together anyway?

 

[rock soderstrom]

I would argue the amount of crosstalk depends entirely (?) on where the common point is. The further "upstream" (ie. closer to the transformer, topologically speaking), the lower the crosstalk should be.

This sounds realistic, but the crucial question is rather how far "upstream" is really necessary and at what point is it simply good enough?

I think we will need a real world test to answer this question. The whole thing is certainly also dependent on which microphones we want to feed. A cathode biased circuit is probably less sensitive than one with fixed bias derived from the heater voltage. I will try it with 2x U67 clones...

 

[Khron]

So you also say, that separate supply circuits is favorable? Or where would the common point be with a shared circuit? At the connectors?

What about those stereo tube mics (SM2, C24, USM69,...). They all also share a common supply. But crosstalk is probably not that of an issue because the capsules are so close together anyway?

This sounds realistic, but the crucial question is rather how far "upstream" is really necessary and at what point is it simply good enough?

This was my point / logic

 

[rock soderstrom]

I will try it with this circuit. I will set up B+ in a similar way.

That should be good enough?

 

[Khron]

Probably gonna wanna go with an LM337, for the negative heater voltage Unless the circuit is altered to work with a positive heater..?

 

[rock soderstrom]

Probably gonna wanna go with an LM337, for the negative heater voltage Unless the circuit is altered to work with a positive heater..?

You are right!

 

[rock soderstrom]

I'm not 100% sure yet if that's how it works. I wouldn't worry about it for a C12, but the U67s are a bit problematic because of the biasing. I might give each tube its own LM337, but then it should really work! We will see.

 

[Tubetec]

Well my logic for doing both heater and HT circuit seperately for each mic is ,
two small off the shelf transformers will probably work out cheaper than a special one with HT and LT windings , A single transformer will have interactions between HT and LT windings , in the olden days they often arranged screening on the bobbin to isolate windings , a modern power transformer might not have that .
Being able to arrange two transformers(HT and LT) for minimum interaction is another plus .
Yeah sure its over the top , but the extra cost in terms of components is small , the same PCB for a standard mono mic could be used twice , rather than having to go to the trouble of designing (down in terms of performance) a seperate stereo psu board .

 

[rock soderstrom]

Well my logic for doing both heater and HT circuit seperately for each mic is ,
two small off the shelf transformers will probably work out cheaper than a special one with HT and LT windings , A single transformer will have interactions between HT and LT windings , in the olden days they often arranged screening on the bobbin to isolate windings , a modern power transformer might not have that .
Being able to arrange two transformers(HT and LT) for minimum interaction is another plus .
Yeah sure its over the top , but the extra cost in terms of components is small , the same PCB for a standard mono mic could be used twice , rather than having to go to the trouble of designing (down in terms of performance) a seperate stereo psu board .

You are in the right thread!

 

[Murdock]

I'm not 100% sure yet if that's how it works. I wouldn't worry about it for a C12, but the U67s are a bit problematic because of the biasing. I might give each tube its own LM337, but then it should really work! We will see.

Would also be great if you could test if two LM337 really are necessary or if one shared will suffice.

 

[Lampie519]

I think the current is no problem here but the gound path. Never have 2 devices that do not share the same signal path be powered by the same DC power source. Now with microphones you will not get ground loops but currents can be different over ground. Therefor i would suggest to use seperate secondaries at least. I don't say it will not work just to make sure it will as trouble shooting is a pain this way.