I'm appalled by the number of DC power supplies I have to build to
feed the filaments of the DHTs I use (directly heated triodes), which
need one supply per tube ideally. I'm using up loads of voltage regs
in the process, and wondered which ones "members of the panel" favour.
Typical supply is a 60VA 0-12, 0-12 transformer, then on each
secondary Schottky bridge, 15,000uf cap, 12v voltage reg and an
adjustable reg confugured as a current source. Typical requirement is
7.25v at 1.25A (10Y), but there's also 2.5v at 1.75A (46) and 5v at
1.25A (300b).

I'm using LM1084 at present, but sneaking in some 78S12 because
they're cheaper. LM1084s have gone up recently, so also interested in
cheaper suppliers than Farnell.

So - which voltage regs are your fave choices here? Good quality
output important, but cost also a consideration. This looks right up
Jim's street for starters.

In article
,
Andy
Evans wrote:


So - which voltage regs are your fave choices here? Good quality output
important, but cost also a consideration. This looks right up Jim's
street for starters.

FWIW I tried various types of 'IC voltage regulator/stabiliser chip' some
years ago and decided I wasn't keen on any of the common types. Too prone
to oscillations or excess noise, etc. Newer ones may be better, but I lost
interest in using them. :-)

So I have tended to use variations on the kind of topology shown on

http://www.audiomisc.co.uk/Armstrong/700/730PSU.gif

As you can see, the zener sets the nominal voltage, and a capacity
multiplier pass device smooths over the noise and gives a slow start/stop.

That example was for about +/-20V up to about 100mA, used for the rails in
a preamp. But the same topology has been used many times by my old research
group for voltages in the few V region at currents up to a few amps, so may
serve your purpose. [1] Just alter the components to suit. Make the pass
transistor a form of darlington pair if needed. I found a single device was
usually fine for currents of the order of an amp or two. But if you need
high current the single-pack darlingtons made for cheaper SS amps might
well do that well if you wanted. But these may need to be checked for
oscillation problems.

Main advantages of the topology are low noise and a gradual windup and
rundown rather than coming on or going off with a crack. Kinder for your
valve heaters, perhaps. You may also like the fact that it has no overall
feedback. :-)

No idea if it will suit you, but since you mentioned my name...

For power amps, though, I always just used a decent transformer and large
caps, then designed the amp to reject power line variations. So no need for
any active smoothing/stabilisation/regulation.

Slainte,

Jim

[1] e.g. for the PSU for 5-12V 1A Gunn diodes where noise from the PSU
needs to be minimal as it would introduce noise sidebands to the 94GHz
output. If you want I can see if I can dig out a diagram of the variable
and fixed versions used for that, but they are in essence the above.

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.audiomisc.co.uk/index.html
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html

Andy Evans wrote:

I'm appalled by the number of DC power supplies I have to build to
feed the filaments of the DHTs I use (directly heated triodes), which
need one supply per tube ideally. I'm using up loads of voltage regs
in the process, and wondered which ones "members of the panel" favour.
Typical supply is a 60VA 0-12, 0-12 transformer, then on each
secondary Schottky bridge, 15,000uf cap, 12v voltage reg and an
adjustable reg confugured as a current source. Typical requirement is
7.25v at 1.25A (10Y), but there's also 2.5v at 1.75A (46) and 5v at
1.25A (300b).

I'm using LM1084 at present, but sneaking in some 78S12 because
they're cheaper. LM1084s have gone up recently, so also interested in
cheaper suppliers than Farnell.

So - which voltage regs are your fave choices here? Good quality
output important, but cost also a consideration. This looks right up
Jim's street for starters.

Dare you use a switching regulator ?

Graham

Thanks Jim - I'll look at this solution. Were you thinking something
like a TIP50 device? I have a box full, so that would be handy for
lower currents. What would be good for 2A and over (say up to 5A
rating)?

Did you try the LM1084 types? I guess you didn't like the 78 types and
the 317.

In article , Jim Lesurf jcgl@st-
and.demon.co.uk scribeth thus
In article
,
Andy
Evans wrote:


So - which voltage regs are your fave choices here? Good quality output
important, but cost also a consideration. This looks right up Jim's
street for starters.

FWIW I tried various types of 'IC voltage regulator/stabiliser chip' some
years ago and decided I wasn't keen on any of the common types. Too prone
to oscillations or excess noise, etc. Newer ones may be better, but I lost
interest in using them. :-)


Shouldn't you valve types be using DC from batteries;?...

--
Tony Sayer

"Andy Evans" wrote in message
...
I'm appalled by the number of DC power supplies I have to build to
feed the filaments of the DHTs I use (directly heated triodes), which
need one supply per tube ideally. I'm using up loads of voltage regs
in the process, and wondered which ones "members of the panel" favour.
Typical supply is a 60VA 0-12, 0-12 transformer, then on each
secondary Schottky bridge, 15,000uf cap, 12v voltage reg and an
adjustable reg confugured as a current source. Typical requirement is
7.25v at 1.25A (10Y), but there's also 2.5v at 1.75A (46) and 5v at
1.25A (300b).

I'm using LM1084 at present, but sneaking in some 78S12 because
they're cheaper. LM1084s have gone up recently, so also interested in
cheaper suppliers than Farnell.

So - which voltage regs are your fave choices here? Good quality
output important, but cost also a consideration. This looks right up
Jim's street for starters.

**The LM317 is as good as the 1084, but, if you want really good regulation,
you still can't beat the venerable LM723. More inconvenient, but still one
of the best performing regulators at reasonable price. For my money, I like
the LM317/337 regs. Very cheap, easy to use and v ery high performance.

Trevor Wilson

if you want really good regulation,
you still can't beat the venerable LM723.

Note that I really need an absolute minimum of 2A current handling, so
LM723 is out and LM317 is too marginal - tubes need between 1.25 and
1.75A.

No virtue in batteries, and a pig to implement. I have tried, but a
good current source sounds just as good as long as you can get the hum
right down.

Keep the suggestions coming!! Andy

"Andy Evans" wrote in message
...
if you want really good regulation,
you still can't beat the venerable LM723.

Note that I really need an absolute minimum of 2A current handling, so
LM723 is out and LM317 is too marginal - tubes need between 1.25 and
1.75A.

**It is a trivial exercise to boost the current of any regulator. Just check
out the manufacturer's data sheet.


No virtue in batteries, and a pig to implement. I have tried, but a
good current source sounds just as good as long as you can get the hum
right down.

Keep the suggestions coming!! Andy

**Use a LM317 and a series pass device. You could use a LM338K, but they're
too expensive for a cheapskate like me. For a filament supply, they'll do
fine and they're easy to implement. In truth, I have used them in the past.
I just don't like to shell out the big Bucks for them.

Trevor Wilson

"Andy Evans" wrote in message
...
if you want really good regulation,
you still can't beat the venerable LM723.

Note that I really need an absolute minimum of 2A current handling, so
LM723 is out and LM317 is too marginal - tubes need between 1.25 and
1.75A.

The LM723 was designed to work with a series-pass transistor/

No virtue in batteries, and a pig to implement. I have tried, but a
good current source sounds just as good as long as you can get the hum
right down.

Heavy filamented DHTs were comonly run from raw AC in the old days so I
don't know why you think you need to keep the hum "right down". Valve
filaments should really be supplied from a voltage, rather than a current,
source. The only advantage of feeding them with a current source is that it
reduces the switch-on surge, but again with heavy filaments that really
isn't an issue.

David.

On Jan 16, 12:14*pm, Andy Evans wrote:
I'm appalled by the number of DC power supplies I have to build to
feed the filaments of the DHTs I use (directly heated triodes), which
need one supply per tube ideally. I'm using up loads of voltage regs
in the process, and wondered which ones "members of the panel" favour.
Typical supply is a 60VA 0-12, 0-12 transformer, then on each
secondary Schottky bridge, 15,000uf cap, 12v voltage reg and an
adjustable reg confugured as a current source. Typical requirement is
7.25v at 1.25A (10Y), but there's also 2.5v at 1.75A (46) and 5v at
1.25A (300b).

I'm using LM1084 at present, but sneaking in some 78S12 because
they're cheaper. LM1084s have gone up recently, so also interested in
cheaper suppliers than Farnell.

So - which voltage regs are your fave choices here? Good quality
output important, but cost also a consideration. This looks right up
Jim's street for starters.

LM338K in the steel case will do all the business you require and last
forever. Maybe pricey but worth it for an IC you intend sticking into
the furnace that is a tube amp.

Good luck.

Andre Jute
Visit Jute on Amps at http://members.lycos.co.uk/fiultra/
"wonderfully well written and reasoned information
for the tube audio constructor"
John Broskie TubeCAD & GlassWare
"an unbelievably comprehensive web site
containing vital gems of wisdom"
Stuart Perry Hi-Fi News & Record Review

HTH.