More DC-Filament Musings

[jbefumo]

Been looking at various DC filament supply components from TentLabs and others, and they all strike me as a bit of overkill for anything but high-end audiophile applications. Seems like it should be possible to just use an appropriate transformer, say, 115:9V with a full wave bridge, or maybe  some kind of step-up solution fed from the 6.3VAC supply, without the need for regulators, heat sinks, etc.  But would that be worth the bother?  I tend to obsess about hum ever since trying to sell a nearly new Mesa Blue Angel some decades back, and the first guy who came to look at it, turned every dial up to 10, listened to the hum, and declared that the amp was defective.  I brought it to a local tech who told me the amp was fine, and nobody in their right mind would expect anything else. Nevertheless, that's kind of one of the reasons I don't try harder to sell my amps -- just don't need the money badly enough to have anyone breaking my chops afterwards...


Still, it's become something of an acid test for me to be able to turn everything all the way up and near nothing whatsoever.


Joe

[kagliostro]

If DC is required a friend of mine simply rectify and level it (diode bridge + electolytic cap)

then a resistor to drop (under load) the voltage to 6.3v (or near)

Franco

[jbefumo]

So ...  thinking aloud ... I would pick a transformer that will give the closest voltage I need to rectify to 6.3V, say 9V-12V, at the required current, plus an allowance for heat in the dropping resistor.  Figure out the total maximum load of all my tubes, and use that to calculate the voltage drop I need and the power capacity of the dropping resistor -- then just fiddle with it until I get just the right voltage under load ... right?  Trying to be cautious here -- I have boxes of costly stuff I've bought, only to discover I had mis-specified and then had to replace -- of course, now I have all those extra parts, and have to design yet another amp around each of them ;^)

If DC is required a friend of mine simply rectify and level it (diode bridge + electolytic cap)

then a resistor to drop (under load) the voltage to 6.3v (or near)

Franco

[silverfox]

Picking a transformer with a higher voltage like 9 or 12 volts won't be the answer. When you rectify the 6.3 volts you'll end up with a little higher voltage out of the rectifier anyway and you have to drop the voltage a little- by adding a diode to the circuit. The actual voltage you get for the filaments can be a little lower. I've even seen 5 VDC used.

I've had good results with a ground lift circuit. Couple of resistors create a voltage divider and a capacitor in the middle of the divider. Then connect the filament center tap or artificial ground to the middle of the divider. One end of the divider goes to ground, the other to B+. Lots of examples around and usually the lift voltage is under 50 VDC.

In the attached sch. 'A' connects to +430 volts DC and after reduction through the divider circuit results in abot 35 VDC being superimposed on the center tap of the filaments. I really don't understand how it works but the ground state of the filaments is being lifted above the noise level.

As far as DC filaments and lifted grounds go, they typically are used on, the first two gain stages of high gain amps, and HiFi circuits, instrumentation and in general, noise sensitive electronic circuits.

silverfox.

[Willabe]

Scroll down to the part on elevated heaters. Very easy to do only a few parts and works great. 

I like to go with ~70dcv to 80dcv. 
 
http://www.valvewizard.co.uk/heater.html

[trobbins]

There are many parts of an amp that can add in hum and noise.  Although an AC heater is one likely contributor, it also may not be your problem with an amp at all, and if you don't add in the DC heater parts properly then you may end up not going forwards at all, but actually end up worse off due to the hum and rectifier noise from the heater DC supply.

It is much better advise to diagnose and identify where the hum is mainly coming from for starters - if you have the incentive, then there are many many forum posts on tracking down hum/noise, as well as articles and books that go in to that topic.

[jbefumo]

Actually, my amp is quite quiet even with A/C heaters -- just a matter of making it as quiet as possible

There are many parts of an amp that can add in hum and noise.  Although an AC heater is one likely contributor, it also may not be your problem with an amp at all, and if you don't add in the DC heater parts properly then you may end up not going forwards at all, but actually end up worse off due to the hum and rectifier noise from the heater DC supply.

It is much better advise to diagnose and identify where the hum is mainly coming from for starters - if you have the incentive, then there are many many forum posts on tracking down hum/noise, as well as articles and books that go in to that topic.

[jbefumo]

Was just now at the ValveWizard heater page 8^)

Scroll down to the part on elevated heaters. Very easy to do only a few parts and works great. 

I like to go with ~70dcv to 80dcv. 
 
http://www.valvewizard.co.uk/heater.html

[trobbins]

A suggestion is to use a 6v battery to power heaters to see how much reduction you could end up with.  Then compare with the simplest changes  - a humdinger pot, and then an elevated heater.

[kagliostro]

+ 1 for Tim


Franco

[Ugly Distortion]

I looked into the Tentlabs but haven't tried yet, they have two, I think regulated and unregulated, but double check that. The regulated would be more trouble (and money) than it's worth in an MI amp. Pete Millet also has a board/kit available, again regulated. The smaller Tentlabs is a nice size to stick in an amp, even if you remove it it woudl be a nice little board to have on hand. It supposedly can take just 6.3V in and give that out, not sure how it avoids loss. I emailed and it's 1.5amps in open air, more like 1A inside a chassis. But it's all a pretty pricey way to go compared to DIY.

http://www.hificollective.co.uk/catalog/tentlabs-tube-heater-supply-p-2189.html
http://www.hificollective.co.uk/components/tl_heatersupply.html
http://www.pmillett.com/dc_filament_supply.htm

[Willabe]

A suggestion is to use a 6v battery to power heaters to see how much reduction you could end up with.  Then compare with the simplest changes  - a humdinger pot, and then an elevated heater.

Yes, agree, +1!   

[Willabe]

There are many parts of an amp that can add in hum and noise.  Although an AC heater is one likely contributor, it also may not be your problem with an amp at all, and if you don't add in the DC heater parts properly then you may end up not going forwards at all, but actually end up worse off due to the hum and rectifier noise from the heater DC supply.

It is much better advise to diagnose and identify where the hum is mainly coming from for starters - if you have the incentive, then there are many many forum posts on tracking down hum/noise, as well as articles and books that go in to that topic.

+1 on this too. 

[shooter]

fwiw, if you wanna pursue dc filaments for whatever logic, the schematic below is a re-draw from an vox30C2

[Heinz]

Just a crazy idea I saw some time ago (but I don't remember where). If your power amp uses self-biasing with a large cathode resistor and has a high enough idle current you can use a filament of a preamp tubes as the cathode resistor for the power amp. This reduces the load on the heater supply and gives you free DC heating. Obviously, the idea cannot be applied to many amps but I still find it intriguing

[kagliostro]

Interesting Heinz

Few time ago I've seen a schematic that uses filament bias, not that I understand it very much, however here it is



---

If the chosen way is the regoulated way, in the web you can find a lot of low cost circuits like this

http://www.amazon.it/COLEMETER-Convertitore-Regolabile-23V-30V-Elettronica/dp/B00E370I7G/ref=sr_1_1?ie=UTF8&qid=1454948565&sr=8-1&keywords=dcdc

Input from 4v to 35v output from 1.23v to 30v @ 2A (3A with heat sink)

A friend uses this kind of circuits on the CNC controllers he assembles

Franco

[DummyLoad]

Just a crazy idea I saw some time ago (but I don't remember where). If your power amp uses self-biasing with a large cathode resistor and has a high enough idle current you can use a filament of a preamp tubes as the cathode resistor for the power amp. This reduces the load on the heater supply and gives you free DC heating. Obviously, the idea cannot be applied to many amps but I still find it intriguing

here is a pair...


http://el34world.com/charts/Schematics/files/silvertone/silvertone1474.pdf


http://el34world.com/charts/Schematics/files/silvertone/silvertone_1433_manual.pdf

[Heinz]

here is a pair...
http://el34world.com/charts/Schematics/files/silvertone/silvertone1474.pdf
http://el34world.com/charts/Schematics/files/silvertone/silvertone_1433_manual.pdf

Cool! I had no idea there were production models using this technique. It's not quite clear from the schematics which tubes are heated with the DC current but I bet the input stage is one of them.

[DummyLoad]

here is a pair...
http://el34world.com/charts/Schematics/files/silvertone/silvertone1474.pdf
http://el34world.com/charts/Schematics/files/silvertone/silvertone_1433_manual.pdf

Cool! I had no idea there were production models using this technique. It's not quite clear from the schematics which tubes are heated with the DC current but I bet the input stage is one of them.

there are 2 12AX7 filaments shown in schematic on the 1433. the 12AX7 in the lower chassis and the 12AX7 in the top chassis. all the other tubes are NOT 12AX7. so has to be the two. 

on the 1474 the 12AX7 summing amp/cathodyne and 6CG7 are 6.3VAC powered. the 12AX7 reverb drive/recovery and 12AX7 1st stage preamps are cathode bias string for 6L6G. series-parallel network with 39R 5W and 270R 5W. still not as clear as on the 1433.

it would be 2 of the three for 27V, but not all three. i'm assuming it's the two mentioned because the 12AX7 pair chosen are the most susceptible to noise.

--pete

[Ed_Chambley]

I make up a few of these with all components on the diode.  It is not regulated, but works fine on everything.  Normally with guitar amps the AC on the heaters is not a problem, but in studio it can be an issue if you must have the amp inside with you.

Checking floor noise with everything on "11" is rather ridiculous for someone to have expectations of silence.

This has worked well enough and I have also used the same the same basic idea for powering relays.

[PRR]

Don't put in as much gain as a Mesa.

Power tubes do not need DC.

90% of any hum will come at the LOWEST level stages: guitar preamp, volume post-amp, and reverb recovery amp.

Regulators are silly, but you can't tell new kids old truths.

Heaters tend to be high current low voltage because that reduces hum on AC heat. DC supplies don't hum, but are far more efficient if worked medium voltage medium current. (0.6V rectifier drop looms large with 6.3V loads; and three bottles is near an Amp which is too much for a 1N400x to be reliable). Wire two to four of your lowest-level stage bottles series heater and throw 24V or 48V 0.150A at them. See most any PCB-based modern tube amp. Peavey stacked 'em quite high.

That DC must be *clean*. At least one stage of R-C filtering with large caps. 120+Hz rectifier ripple buzz is much more annoying than pure 60Hz hum. Especially if you can run a "small" speaker which cuts-off below 100Hz. ("My" VT-40 with open-back cabinet had a horrid humm, but you could not hear it 2 feet away from the baffle.)

[jbefumo]

Interesting, thanks.


I'm not actually committed to DC filaments, unless I can make it happen without a lot of pain and suffering -- in any case, would probably just reserve it for the first preamp stages, if at all.  I know a lot of my ideas are aimed at solving problems that may not even exist, just to see if I can come up with something better .... Hell, if I took all the money and effort I've expended since really getting into this (Summer, 2009), I could have bought one each of every great vintage amp ever, and be playing like [pick you favorite great player]....


When I was younger it was about swapping cams, and carbs, and cylinder heads .... at least amps don't get you so greasy, though they CAN get your attention with a bit of a jolt from time to time.

fwiw, if you wanna pursue dc filaments for whatever logic, the schematic below is a re-draw from an vox30C2

[jbefumo]

Well, far from a kid any more, but ... must be that physics background -- always looking for high-tech tricks to solve simple problems.


As an aside -- haven't thought about this in years -- when I was an undergrad at Brooklyn College, I worked part time at their Dynamitron (high-resolution, medium energy proton accelerator) lab.  I participated in upgrading it from tube rectifiers to solid state, which upped its power from 8.75 MeV to 11.5 MeV --- now if only I could fit THAT into my amp!  I could rule them world!

Don't put in as much gain as a Mesa.

Power tubes do not need DC.

90% of any hum will come at the LOWEST level stages: guitar preamp, volume post-amp, and reverb recovery amp.

Regulators are silly, but you can't tell new kids old truths.

Heaters tend to be high current low voltage because that reduces hum on AC heat. DC supplies don't hum, but are far more efficient if worked medium voltage medium current. (0.6V rectifier drop looms large with 6.3V loads; and three bottles is near an Amp which is too much for a 1N400x to be reliable). Wire two to four of your lowest-level stage bottles series heater and throw 24V or 48V 0.150A at them. See most any PCB-based modern tube amp. Peavey stacked 'em quite high.

That DC must be *clean*. At least one stage of R-C filtering with large caps. 120+Hz rectifier ripple buzz is much more annoying than pure 60Hz hum. Especially if you can run a "small" speaker which cuts-off below 100Hz. ("My" VT-40 with open-back cabinet had a horrid humm, but you could not hear it 2 feet away from the baffle.)

[jbefumo]

Thanks -- downloaded and saved!


I should probably add that I'm nearly deaf as it is, and until I manage to get my hearing aids serviced, I can't hear the difference anyhow -- It's all about the challenge of building a better mousetrap ....


Re: My 'Bad JuJu' thread -- I'm at the point of coming to grips with the fact that the Blackface Super Reverb sound is pretty much what I've been chasing after all along .... but then, it's not the destination, but the journey that counts!

I make up a few of these with all components on the diode.  It is not regulated, but works fine on everything.  Normally with guitar amps the AC on the heaters is not a problem, but in studio it can be an issue if you must have the amp inside with you.

Checking floor noise with everything on "11" is rather ridiculous for someone to have expectations of silence.

This has worked well enough and I have also used the same the same basic idea for powering relays.

[Shack]

lol, I was gonna post this question too.........I lost my copy of the Torres book, inside tube amps, but.....I bought a bridge rectifier and a 4700 uf cap for something in his build a class A single ended amp project.....and couldnt remember what it was for. Now I know and still have those parts, maybe ill use it in a project to see how it works.

[jbefumo]

Actually, the problem turned out to be my guitar ;^)


I was going to add a d/c supply to the preamp tubes, then decided: what for? Just leaving well enough alone.

There are many parts of an amp that can add in hum and noise.  Although an AC heater is one likely contributor, it also may not be your problem with an amp at all, and if you don't add in the DC heater parts properly then you may end up not going forwards at all, but actually end up worse off due to the hum and rectifier noise from the heater DC supply.

It is much better advise to diagnose and identify where the hum is mainly coming from for starters - if you have the incentive, then there are many many forum posts on tracking down hum/noise, as well as articles and books that go in to that topic.