Phantom power current requirements, AC vs. DC heater voltage

[fazeka]

Newbie Chris, here.

As some of you may know, I have an Ampex 351 electronics "toob" preamp I am hacking, err... "modifying" for use as a microphone preamp (don't worry, it was incomplete and trashed before I got my grubby little hands on it, so I didn't quite start with a pristine vintage piece).

Anyway, I am modifying it according to notes by Greg Norman (Steve Albini, go here for his document from this page. Refer here for original schematic). He says that they use an outboard source for phantom power. However, I was hoping that there could be a way to supply phantom power from the stock power transformer. Because I was already planning to run all the 12A*7 filaments on the 12.6 VDC secondary, I had an idea to utilize the power transformer's 12.6 VAC secondary for a phantom power supply after looking at Jakob's G9 mic pre schematic here. I was hoping to create a "voltage quadrupler" to get the voltage up to 48 VDC. The thing is, I understand that I'll need anywhere from 10 to 20 mA for phantom power. I'm not sure ('cause I'm a newbie :wink, but I remember reading that for some applications four 12A*7 tubes (from original schematic) draw on average about 0.3 mA each (or 1.2 mA total). But, I am not sure if the power transformer in this preamp *isn't* overdesigned to deliver a minimum of 10mA. My questions are 1) is 10 ? 20 mA a reasonable requirement for phantom power, 2) does anyone know what the current ratings are on this particular secondary and, more importantly, 3) is there a way to calculate this info so I don't have to resort to asking the forum everytime and moreover so I can try and learn something for a change? :grin:

Finally, is there any harm or benefit regarding using DC for the filaments on all the preamp tubes instead of what's shown on Norman's schematic? I guess he based his modified schematic on the original schematic as V4 and V5 originally used AC on their filaments. I thought I remember reading something somewhere about preamp applications benefitting from DC filaments as it keeps the noise/hum down.

Thanks,
Chris

 

[CJ]

20 ma should be plenty for most mics.
I do not think the transformer will notice an extra 20 ma missing.
But does a quadrupler chew up 80 ma at 12 volts to make 20 ma at 48 volts?
The math would seem to indicate so.
The 12A? tubes I believe draw 300ma at 6.3 and 150 ma at 12 volts.
I would just build a doubler, which would give you about 24 times a bump from the caps, maybe 28 volts?
Should be plenty for most mics.
Mahalo!
cj
And get to work on the LA2!
:guinness:
Did you ever recover that wild blue Fender?

 

[PRR]

> I understand that I'll need anywhere from 10 to 20 mA for phantom power.

The maximum current available is 52V/3.4K, or 15mA, and that leaves zero power for the mike to live on. Phantom specs set a limit of 10mA per mike. Most mikes draw 3mA or less.

> I was hoping to create a "voltage quadrupler" to get the voltage up to 48 VDC.

Voltage quadruplers are pretty inefficient. You would need around 80mA from the 12VAC winding to get 10mA at the 48V side, and you need VERY low winding impedance and very large caps to get there. If you do "too good" (or unloaded), voltage could rise to 71V, so you need a regulator.

Hell, if you are not actually Recording, you are not using the bias oscillator, and that must eat 10mA when it runs. So you have 10mA to spare, just at 300VDC. Yank 1V-6, the bias osc tube, and tape a note over the socket so nobody puts it back. Drop a 25KΩ 5 Watt resistor from the bias switching relay to a couple 24V 0.5 Watt Zener diodes to ground. The top of the Zeners will be 48V, very clean, up to 10mA.

 

[fazeka]

OK, I don't know where I got that figure regarding the current consumption of the 12A*7 type tubes. It's understandable, I'm a newbie. :shock:
I think I meant to say 0.3A, not mA... :grin:

CJ:

Ah, makes sense about the 80 mA needed at 12V to make 20 mA at 48V. I'm still learning, ya know...

I am still needing a couple of pots (100k log RV4 2-watt mil-spec), the chassis and the trannies for my LA-2A's, so I can't get to them quite yet. :sad:
In due time, I guess.

Yes, I got the "wild blue" '62 Princeton fixed. Bad bias diode caused the 6V6s to draw 80 mA (IIRC). Those Ruskies had some nice cherry plates! :shock:
Now, I got them somewhere around 22 mA @ 325 V on a NOS Philips matched pair. :grin:
Also got a stash of some Mazdas and Brimars matched pairs! :green:
Then I got Larry Rodgers to build me a period correct cab (the other one was junque and was not even original!). So yeah, it?s rockin' now. :thumb:

Paul, you're right, I'm obviously not recording (i.e., using the record relays) with the pre as I have no transport. So there's no need for V6, and that makes sense. I should have thought about that, I guess that's why I'm not getting paid the big bucks. :shock:

Yee-ha, I get to add phantom power to my pre! Cool! And I learned something to boot.

Thanks, guys!

:guinness:

 

[PRR]

> I guess that's why I'm not getting paid the big bucks.

Hell, I don't get paid the big bucks either (and my nose was rubbed in that fact two different ways today).

That's why I look for the simplest, cheapest way.

A 48VDC wall-wart is simple, but not always easy to find, at a good price (i.e. very cheap).

Voltage multiplication has its place, but you usually need real over-kill to get anywhere near the theoretical multiplication. And in this case you really only have 6.3VAC to ground to play with: the lower 12V winding is center-tapped to ground, and the upper one really wants its midpoint grounded.

So looking around to see how much power this beast ate (to see if there was any place where another 48V*10mA= 0.5 Watts would not be noticed) I saw RECORD circuits. The head-driver may be quite low power, but the bias oscillator (actually the erase oscillator) has to be most of a watt, and when you do it with a tube it tends to be several watts. Also the wacky switched-in 1K resistor suggests that the bias-erase circuit eats about 66% of the power of everything else, so my gut feeling was that there was an easy 10mA to spare if you never actually erase or bias mag-tape with this box.

Also, yanking the oscillator tube saves 2 Watts of power transformer load and chassis heat just from losing the 12AU7 heater. That almost covers the 3 Watts my rather wasteful dropping-technique uses.

 

[fazeka]

Hey, Paul. I can dig simple and cheap. That's pretty much the story of my life! It's also how I like my women! :grin:

Thanks for the methodology on finding a way to provide phantom power to this beast. Entertaining *and* educational. Dang! Why couldn't more of my college courses have been like this? :grin:

So, let me understand this: there could be another (less "wasteful") way to provide phantom power to this circuit?

I also didn't get any criticism about my DC filament supply suggestion, so I guess you guys are OK with it?

Oh, and CJ, here's some pics of my Fender guitar amps. You can see the restored "wild blue" Princeton. The other shot is my '62 Bandmaster I actually acquired before "wild blue".

 

[PRR]

> my DC filament supply suggestion

You only need 3 tubes, 1, 4, 5, to get from Mic In to Line Out. You don't need tube 2, 3, or 6. You can just move 4 and 5 over to the 12VDC supply.

I would however (I did not study the modern mod) replace any copper or selenium heater rectifier with silicon, replace that old 4000uFd cap, and probably upgrade it to 20,000uFd.

 

[fazeka]

Hi, Paul.

I was going to replace the selenium rectifier with two 1N4007s, though I am curious as to replacing the 4000 uF with 20,000 uF? Does this provide better filtering? Apologies if this is wrong or obvious.

Thanks much,
Chris

 

[CJ]

Nice amps!
:guinness:

 

[Sinterklass]

I have a mod like that in progress myself, currently "on hold" while a list of more urgent projects gets done. My intent is to do the heaters a bit differently. I plan to use a standard rectifier bridge and regulator from the existing AC heater winding (record amp and bias tubes will be pulled, cutting current consumption in half). The original DC heater supply winding then becomes available for phantom power. I plan to take out the selenium rectifier and run a full wave voltage doubler instead (ignoring the CT of that winding), feeding a phantom power regulator. This way I get all voltages from the existing transformer, don't have to waste power and heat trying to derive phantom from B+ (which seems a bit dangerous to me anyway, especially if the shunt resistor opens).

You may want to bear in mind that there are resistors from the input XLR with a grounded CT for the input attenuator. This probably does no harm, but I would lift that ground.

Another sneaky trick you might want to try (I have done this before) is to disconnect the front panel heaphone jack and re-wire it as a DI input. You just run a (preferably shielded) wire from it to the "unbalanced bridge" select terminal on the switch. The input impedance here is a bit lower than I like, but it does work.

Sinterklass

 

[mikeyB]

Just racked up a load of telefunken type broadcast modules WSW 1A's x 4, V276 x2 and V376 x2. Front ends have JLM go betweens and the 276/376 have FET DI's. Rack was rushed for a session so I'll take pics when I get it back - few tweaks needed. Using N624 powersupply - chunky beast, but only 24V - not seen schematic for this so until I get it back, I dont know what the secondary ac voltage is. He's using seperate 48 supply but would like to derive it from the main psu. From points mentioned am I right to assume that 8 channels would require up to 80mA at 48V and would it be wise to make it nearer 120mA to be on the safe side. Hopefully would only have to double the voltage(but more likely triple). What value caps would I need to use in the circuit - from gyraff G9 (2 channel) uses 47uF on tripler chain and 100uF at the end. Would I be looking at somewhere around 220uF in the doubler/tripler and 470uF pre regulator?
There is just enough space in the module to use axial caps on tag board but if I need seperate transformer then so be it!! - i'll have to mod the case!!
Any thoughts/ advice much appreciated ;D

 

[Sinterklass]

I know of at least one console that derived its phantom power from the +24 volt rail by using a DC-DC converter.  I'm pretty sure these are available as standard modules, but I am not sure whether you would not be better off just adding another transformer.