I am very much impressed by Thomas Mayer's designs and try out filament bias. His implementations use a lot of iron at considerable expense. My goal is to build a filament bias amp on the cheap (relatively speaking). Got the idea after reading this post.
The 4P1L, triode strapped, has considerably more gain than most DHTs (mu = 11ish), which considerably cuts down on the power required and dissipated by the filament bias. I am using six of Pete Milett's DC filament power supplies, which will allow biasing the tubes individually.
B+ is fed by Tubecad PS-2 supplies, one per channel. Using ASC motor runs as the ultrapath caps. Thinking the CCS for the driver stage may be overkill and that a 4k resistor would suffice. Probably also need a small cap to bypass the VR tube.
Going to have a ton of heat inside the chassis will all of those regulators and the filament bias - not sure how I will approach this yet.
The 4P1L, triode strapped, has considerably more gain than most DHTs (mu = 11ish), which considerably cuts down on the power required and dissipated by the filament bias. I am using six of Pete Milett's DC filament power supplies, which will allow biasing the tubes individually.
B+ is fed by Tubecad PS-2 supplies, one per channel. Using ASC motor runs as the ultrapath caps. Thinking the CCS for the driver stage may be overkill and that a 4k resistor would suffice. Probably also need a small cap to bypass the VR tube.
Going to have a ton of heat inside the chassis will all of those regulators and the filament bias - not sure how I will approach this yet.
Perhaps you could use that perforated orange material that Wavebourn uses to get rid of the heat. I presume it is metal of some sort.
http://www.diyaudio.com/forums/tubes-valves/183329-one-more-4p1l-se-6.html
ray
http://www.diyaudio.com/forums/tubes-valves/183329-one-more-4p1l-se-6.html
ray
I'm listening to my 300b SET as we speak. It has 4P1L input and 4P1L driver stage. Both stages are identical - they have filament bias and a 126C Hammond interstage. The 126c is rated for 15mA, 105H so the 4P1L run at around 12.5mA. I haven't tried the 126B which is rated at 30mA, 44H. I have a LL1660/18mA to try also. The 126C works fine and sounds very good, and I'd be interested in anything better.
Filament bias is very easy with the 4P1L, and actually you don't need big heatsinks. I opted for a bias of around 13v. The cathode resistors are VERY important in filament bias. I use discontinued Dale wirewound ones. Use the highest quality wirewounds you can, and rate them at three times what they are passing - they get hot. Don't use thick film, generic ceramic ones or alu clad ones - go for best quality. I bought a few 10 ohm Dale ones at 20W - best I found. So you juggle around the bias voltage partly to suit the resistors you are using. I was shooting for about 170v on the plate.
Sound quality is amazing. There is great clarity and delicacy. This is far and away the best 300b SET I've built. I was about to give up on SETs since I have a 2a3 all balanced PP amp that I love - 26 and 4P1L diff pairs into 6C4C outputs. There is an argument for using a 26 globe input stage - it's a lovely euphonic sound - but the 4P1L is a tad clearer and has better treble.
Go for it - this is a stunning tube and I'm using it for anything and everything. I wasn't bothered by the microphonics in the input stage - yes, it's microphonic but I have a heavy chassis and it doesn't get excited if you don't touch it. No problem in the driver stage.
Filament bias is very easy with the 4P1L, and actually you don't need big heatsinks. I opted for a bias of around 13v. The cathode resistors are VERY important in filament bias. I use discontinued Dale wirewound ones. Use the highest quality wirewounds you can, and rate them at three times what they are passing - they get hot. Don't use thick film, generic ceramic ones or alu clad ones - go for best quality. I bought a few 10 ohm Dale ones at 20W - best I found. So you juggle around the bias voltage partly to suit the resistors you are using. I was shooting for about 170v on the plate.
Sound quality is amazing. There is great clarity and delicacy. This is far and away the best 300b SET I've built. I was about to give up on SETs since I have a 2a3 all balanced PP amp that I love - 26 and 4P1L diff pairs into 6C4C outputs. There is an argument for using a 26 globe input stage - it's a lovely euphonic sound - but the 4P1L is a tad clearer and has better treble.
Go for it - this is a stunning tube and I'm using it for anything and everything. I wasn't bothered by the microphonics in the input stage - yes, it's microphonic but I have a heavy chassis and it doesn't get excited if you don't touch it. No problem in the driver stage.
Attachments
EvilMose
Let us know how the 10uF caps work - you don't need them and I don't use them, but it would be interesting to hear if they improve the sound for you - maybe try them on clip leads?
Experiment with starved filaments - you quote 650mA but try 600mA, then 580 and 550mA. Less microphonic and may sound better. Affects the cathode resistor calcs a little.
You have 14 and 23 ohms cathode resistors - consider 15 and 25 using good quality 10 ohm resistor in combinations. Quality here may be audible. Everything counts in filament bias. I use Rod Coleman's regs, and if you want the best sound of all use a choke input supply (or small cap like 220uF). Hammond 159Y is perfect.
What interstage are you thinking of? It's a case of juggling current and inductance. Hammond 126C has 15ma and 105H, 126B has 30ma and 44H. LL1671 has 30ma and 35H. LL1660/18mA looks good at 100H - best on paper. The 4P1L likes higher currents like 15-20mA. It works at less but sounds a little thinner with more treble emphasis. It may be possible even to use the LL1660/18mA as a 2:4.5 step-up and try eliminating a stage. Alt T for the 1660/10mA quotes 33H here so the 18ma version would probably be around 25H. Low but maybe possible.
If you want "on the cheap" the Hammonds are really good for the price. Not far at all behind the Lundahls in sound quality. They're bifilar wound and a good size. The colour coding is a bit vague - I connect brown to HT and green to ground. I tried reversing the secondaries and the sound was worse.
You'll be surprised at how good this will sound.
Andy
Let us know how the 10uF caps work - you don't need them and I don't use them, but it would be interesting to hear if they improve the sound for you - maybe try them on clip leads?
Experiment with starved filaments - you quote 650mA but try 600mA, then 580 and 550mA. Less microphonic and may sound better. Affects the cathode resistor calcs a little.
You have 14 and 23 ohms cathode resistors - consider 15 and 25 using good quality 10 ohm resistor in combinations. Quality here may be audible. Everything counts in filament bias. I use Rod Coleman's regs, and if you want the best sound of all use a choke input supply (or small cap like 220uF). Hammond 159Y is perfect.
What interstage are you thinking of? It's a case of juggling current and inductance. Hammond 126C has 15ma and 105H, 126B has 30ma and 44H. LL1671 has 30ma and 35H. LL1660/18mA looks good at 100H - best on paper. The 4P1L likes higher currents like 15-20mA. It works at less but sounds a little thinner with more treble emphasis. It may be possible even to use the LL1660/18mA as a 2:4.5 step-up and try eliminating a stage. Alt T for the 1660/10mA quotes 33H here so the 18ma version would probably be around 25H. Low but maybe possible.
If you want "on the cheap" the Hammonds are really good for the price. Not far at all behind the Lundahls in sound quality. They're bifilar wound and a good size. The colour coding is a bit vague - I connect brown to HT and green to ground. I tried reversing the secondaries and the sound was worse.
You'll be surprised at how good this will sound.
Andy
Last edited:
Sorry, I dont't catch any advantage of filament biasing. Instead, I see only disadvantages: Lots of heat, owed to the dissipations caused by both cathode and filament currents, need of well filtered, or better regulated DC supply for the filament(s) also, in order to avoid extra hum etc.
Cheers!
Cheers!
Hi Kay,
here a short article about filament bias on my blog:
VinylSavor: Filament Bias, Part 1: Concept
Thomas
here a short article about filament bias on my blog:
VinylSavor: Filament Bias, Part 1: Concept
Thomas
Hi Kay,
You have to look at filament bias in the context of other forms of bias. As the saying goes "democracy is the worst form of government except all the others that have been tried". And the same could be said of filament bias. Look at the alternatives:
Grid bias: Needs a cap input. Battery grid bias actually doesn't sound better - even without a cap - than properly implemented filament bias.
Cathode bias: Larger resistor. Distortion if you don't bypass it, awful sound from bypass caps (even polypropylene) if you do bypass it.
LEDs etc: When I tried them they didn't sound better than filament bias.
Yes, it requires a bit more work, but not that much in fact with the right valve (26, 4P1L) where the filament current/filament voltage/bias voltage is low. Forget it with bias voltages of 20-30v etc. especially when you add higher filaments and currents to that, though it has been done by enthusiasts. You have to use a good DC supply anyway and heatsink it, so the only extra work is a cathode resistor and a higher secondary voltage for the filament transformer. It can help, though to use a choke input, though it's not essential.
So for sound quality alone, I and many others have found it's well worth it. Never having to worry about bypassed cathode resistors is worth it in itself.
Andy
You have to look at filament bias in the context of other forms of bias. As the saying goes "democracy is the worst form of government except all the others that have been tried". And the same could be said of filament bias. Look at the alternatives:
Grid bias: Needs a cap input. Battery grid bias actually doesn't sound better - even without a cap - than properly implemented filament bias.
Cathode bias: Larger resistor. Distortion if you don't bypass it, awful sound from bypass caps (even polypropylene) if you do bypass it.
LEDs etc: When I tried them they didn't sound better than filament bias.
Yes, it requires a bit more work, but not that much in fact with the right valve (26, 4P1L) where the filament current/filament voltage/bias voltage is low. Forget it with bias voltages of 20-30v etc. especially when you add higher filaments and currents to that, though it has been done by enthusiasts. You have to use a good DC supply anyway and heatsink it, so the only extra work is a cathode resistor and a higher secondary voltage for the filament transformer. It can help, though to use a choke input, though it's not essential.
So for sound quality alone, I and many others have found it's well worth it. Never having to worry about bypassed cathode resistors is worth it in itself.
Andy
Ale Moglia needs to join this thread and talk about transformers and anode loads. He's getting distortion from the 126C at higher voltages, so may only be suitable for preamp use at lower voltages out. It does work in the driver stage of my 300b SET but I haven't tried any other option yet like the LL1660/18mA.
http://www.bartola.co.uk/valves/2012/07/27/4p1l-with-126c-ot-load-stage/
http://www.bartola.co.uk/valves/2012/07/27/4p1l-with-126c-ot-load-stage/
A CCS/gyrator performs brilliantly with a 4P1L. Here is my final test for a 126C OPT based on what Andy is using. Found the best point at -15V and Va close to 170V for 15mA. THD is 0.05% for 10Vrms. Sweet...
126C OPT | Bartola Valves
EDIT: I was actually wiring the OPT inverting and apparently Andy mentioned that there is a thread about this. Distortion was very high above 8Vrms. When wired correctly (secondary green to ground) then performs brilliantly)
Cheers,
Ale
126C OPT | Bartola Valves
EDIT: I was actually wiring the OPT inverting and apparently Andy mentioned that there is a thread about this. Distortion was very high above 8Vrms. When wired correctly (secondary green to ground) then performs brilliantly)
Cheers,
Ale
Last edited:
Ah - great! This is something I found out by trial and error - wired the 126C wrong way around and sounded dreadful. Then switched the outputs. The thread about it was probably by me. HT on primary is brown, ground on secondary is green.
So seems you can run 15ma through the 126C, and it's OK for a driver stage. My ears aren't wrong!
Andy
So seems you can run 15ma through the 126C, and it's OK for a driver stage. My ears aren't wrong!
Andy
I've been ruminating on what plate chokes or transformers could be used with the 4P1L. I have a 50H plate choke, 750 ohms to try out. I don't know what current it's gapped for but can experiment. Also a LL1667/15mA plate choke. I don't have a LL1668/25mA but that looks good with 100H.
As for Lundahl interstages:
LL1671 - 30mA - 35H
LL1671 - 20mA - 52H
LL1621 - 20mA - 30H
LL1635 - 20mA - 30H
LL1660 - 25mA - 72H
LL1692A - 18mA - 125H
LL1692A - 25mA - ?
These are all about 100 euros from JAC Music. I believe you can specify the gap.
What I don't know is if the gaps are user-adjustable. I have some Lundahls that I'd love to re-gap to use the 4P1L
Then there's Edcor - whole bunch of useful options there, though I don't know the audio quality. And no doubt other interstage offerings.
Andy
As for Lundahl interstages:
LL1671 - 30mA - 35H
LL1671 - 20mA - 52H
LL1621 - 20mA - 30H
LL1635 - 20mA - 30H
LL1660 - 25mA - 72H
LL1692A - 18mA - 125H
LL1692A - 25mA - ?
These are all about 100 euros from JAC Music. I believe you can specify the gap.
What I don't know is if the gaps are user-adjustable. I have some Lundahls that I'd love to re-gap to use the 4P1L
Then there's Edcor - whole bunch of useful options there, though I don't know the audio quality. And no doubt other interstage offerings.
Andy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.