regulated 12vdc on 6vac heaters?

[rdrgtr]

I recently came across some projects that called for a 12.6vdc (regulated) supply for 12ax7 heaters.  They were regulated with a 7812 (I think) regulator.  The projects were high gain preamps (boogie mkII c and soldano type).  The builders goal was to use dc heaters to reduce noise.  I guess the 12ax7 can take the dc, but the 12.6v part was a little puzzling.  If the dc heater voltage does reduce noise, why doesn't every amp use it?

[RicharD]

The 12A*7 family of tubes have 2 - 6V filaments wired in series.  1 filament is wired from pin 4 to pin 9 and the other is wired from pin 5 to pin 9.  When fed from a 6V source, you parallel pins 4 & 5.  This doubles the current.  When wired from a 12V source, you connect the supply to pins 4 & 5 and ignore pin 9.  There are times where you can ground pin 9, but lets not confuse thangs.  Filaments are basically a coil of wire that gets hot.  They don't care if the source is AC or DC.  Pure DC is silent whereas AC has frequency.  Wall current in Europe is 50Hz and it's 60Hz in the US.  It's low audio frequency..... hum.  With proper layout, AC filaments are just fine for a guitar amp.  When you get your low level audio wires running in parallel next to AC filament wires is where you get into trouble.  Look at how Fender ran their filament wires, up and away from everything else.  AC filaments vs DC filaments was purely a cost issue.  I've seen elderly "High End" equipment with AC filaments.  The Fairchild 670 had AC filaments and those thangs sell for $50,000.00 in barely working condition.  Back in the day when tubes were king and silicon was something secret at Bell Labs, DC was expensive.  You'd need another rectifier tube and more capacitors.  Regulated DC involved adding a regulator tube and a bias circuit on top of that.  A twisted pair of wires becomes a lot more appealing to the manufacturers.  Once silicon became the norm, DC filaments gained popularity because it was now affordable + with the advent of printed circuit boards, you didn't have to worry about noisy filament traces being next to the audio path.  I suspect Soldano was using DC filaments because of his love for very high gain circuitry.  An increase in gain ups the risk of your audio path picking up filament noise.  None the less, if you keep your filament wires up & away, AC is just fine.  My 6DJ8 preamp that hangs off my computer has AC filaments.  I had provisions for DC filaments but never got around to it mainly because it's not a problem.

-Richard

[PRR]

Any heater will work on DC. Batteries used to be common. The hard part used to be getting useful operation with wall-power. That was one reasn heater-cathode tubes were invented.

When using AC (hummy) heat, you want low voltage for low hum. But for a given power, low voltage is high current which is fat copper which is costly.

There's probably an optimum near 2.5VAC or 5VAC. That's what most of the early AC-compatible heaters were.

But then came a fad for putting radios in cars. 6.3V cars needed 6.3V heaters. This was also a period when a lot of new-improved tubes came out, and they mostly hewed to the 6.3V standard.

There's other applications where a higher voltage is better. When you put two triodes in one bottle, two heaters with four leads, you can bring it out on 3 pins and let the desugner wire it for either 6V or 12V use.

*Clean* DC heat won't hum. (Sloppy "DC" heat buzzes.) So voltage is not an issue as far as amplifier operation.

Most ways of making DC from wall-power have several voltage-drops. Rectifier drops 1V. FWB drops 2V. Regulators need 3V of excess raw voltage to be sure of delivering clean DC. Capacitors 16V and over traditionally had better space/cost-efficiency than 6V caps.

If we have 5V of voltage-losses, then making 6VDC is a lot of loss: 5V/11V. If we make 12VDC the relative loss is less: 5V/17V. 24V is even better: 5V/29V. That's pretty good, and if we went further we'd hit the 35V input limit on 1-buck regulators.

So while DC heat eliminates AC around preamps, and higher voltage (lower current) may allow a smaller wire (though we usually pick wire for mechanical strength), there's a lot more parts (to go bad) and more waste heat. Don't do it until you NEED to do it. Guitar-level signals and good twisted wire is low-enuff hum for most reasonable-gain gitar amps.

There is NO reason to regulate. Tube heater voltage should be "close", but tubes ain't as fussy as incandescent lamps. Do you (did you) regulate your 109V-127V wall-outlets to 120.0V to match your incandescent lamps? Nah, my 120V lamps burn-out a little early on my 124V but so what? Lamps are 1,000-hour, tubes are more like 10,000 hour. Instead of a regulator, use a C-R-C filter. This takes buzz off the heater feed, and gives you a resistor to tweak to adjust whatever raw DC voltage you get down to what the tubes want.

[eleventeen]

Couple of other aspects:

a 7812 is nominally a 1 AMP device, so while at first glance you could run 6 qty 12A_7 = 6 * .150 ma = .900 amp @ 12 volts (a la 2-channel AB763) it would still require a modestly sized heat sink; easily the footprint of a 9 pin tube. And...if that heatsink is the chassis sheetmetal and the chassis is also busy gathering all the heat from running other tubes, upside down, in a semi-enclosed cabinet, uhhh.....maybe not such a great concept. A 7812 will need a heatsink at or exceeding 1/4 amp, in my humble opinion.

The power tubes can be pretty confidently excluded from such a scheme because the push-pull nature of their circuit would cancel heater hum. Unless a tube had a heater-to-cathode short, which is rare...and nothing would fix that other than replacing the tube.

I've considered doing this = 12 vdc reg'd to the preamp tubes. It's not exactly a walk in the park. To pass decent current, three terminal regulators need dead minimum 2 volts (and preferably a little more) over their desired output voltage to work. If you have multiple 6.3 volt windings on your PT, fine. A 6.3 winding in series with an unused 5 v winding falls just shy. Let's just say that if the regulator drops out because the input voltage falls below what it need to work, it'll REALLY reduce your hum.

In the end, I'm not convinced that one would be able to discern a 12 v reg'd preamp heater amp from a 6.3 VAC one where really good anit-hum technique was used wiring up the preamp tubes. Maybe I would consider it in a no-budget-constraint ultra fetish hi-fi amp, but in a guitar amp? Probably not.

[PRR]

> power tubes can be pretty confidently excluded from such a scheme because the push-pull nature of their circuit would cancel heater hum.

It's not the push-pull. It is signal LEVEL.

The first stage accepts signal as small as 0.02V. The stage after the volume control typically works at 0.05V. The driver is often 1V and the power tubes 20V or more.

The first stage, the post-Volume stage, and the reverb recovery stage, are where 95% of the hum gets in. Everything else is typically 20 times higher level and thus hum has 20 times less impact.

> 7812 will need a heatsink at or exceeding 1/4 amp

Also depends on the voltage drop. If you must come down from 35V, even 50mA may cause shut-down. And as you say, dumping heat inside a tube amp is difficult. That's my other objection to regulators: proper use requires insight into several factors we never run into just doing tube-amp work. Is there enough after-ripple headroom? Will lead-length oscillate? How much heat in the chip, how much heat in the chassis, how big a fin is needed to stay below thermal cut-out? How hot is too hot? On test-bench? For a mid-day mid-summer outdoor gig?

KISS. Do your AC heat "RIGHT". For sane guitar amp work, the hum will be fine. If it isn't, find your 3 or 4 most hum-sensitve stages. Put those heaters in series, rig a DC supply at somewhat higher voltage, and C-R-C down to a good voltage.

[Tone Junkie]

Forgive me Im kind of a newb to these things but you guys always teach me alot by C-R-C do you mean cap-resister-cap. Bill

[sluckey]

by C-R-C do you mean cap-resister-cap

yes

[LooseChange]

But tell me more.

[sluckey]

But tell me more.

There ain't much to tell. Cap to smooth rectified voltage, resistor to drop voltage to desired level, another cap for additional smoothing. C-R-C.

There is another C-R-C. It's a brand of aerosol spray lubricant. Good chance PRR was not referring to this.   

[LooseChange]

Sorry man... I missed something up there. I get exactly what you are talking about.
I use CRC are the time... It cleans without residue and is really good stuff.