I've been using nothing but DHTs as input tubes for about 3 years now, and my 300b amp has gone through various configurations using mostly 46, 10Y and 26 tubes in the input. I've gone through various topologies also. The worst by far was resistor loads in anode and cathode with cathode bypasses. Battery grid bias was an improvement, but I got best results with filament bias. My attempts at direct coupling the first stages ran into issues with cathode resistors and bypasses, and I gradually lost interest and started preferring all-DHT 2a3 PP amps as an alternative.
When I started using the 4P1L my interest in the 300b SET revived. Here was a tube that worked exceptionally well in filament bias. As a triode it's mu was around 10 and it's anode resistance around 2k. Perfect! And the filaments were around 2v at 650mA, so quite easy to provide a DC filament supply for. They sound best at around 15mA or more, and I had some interstages I could try out so I rigged up an experimental amp. After starting with a 26 input stage I went over to two similar stages of 4P1L. This seemed to give the best tone to the instrumental sound - the 300b is on the warmer side and the 4P1L on the leaner side. The 4P1L had better treble than the 26 also. It is a little microphonic in the input stage but not grossly so, and it doesn't hum.
Having decided to use two similar stages with filament bias, I then started to play around with interstages. I like the 126C from Hammond in the driver stage at present - it is a rounder and fuller sound. In the input I like the LL1660/18mA which is a leaner sound with better treble detail. The combination seems to have good tone. I tried using LL1692A into LL1660/18mA but strangely this sounded too lean and I lost a bit of the magic. The LL1660 is better to my ears than the LL1692A, but not by much. It has better low level and treble detail at the expense of a slight loss of smoothness.
So I ended up with a very satisfying sounding 300b SET. I enclose a draft schematic to kick off some discussion. The voltages are approximate and need refining - I'm still at the tweaking stage. This has become my daily listening amp and gives me a lot of pleasure. It's transparent and detailed as you'd expect from DHTs and has particularly good tone and low level detail with orchestral instruments - you can clearly hear the individual tone quality of clarinets, bassoons, oboes and so on, and percussion in jazz groups is excellent - resonant thwack to the drum skins and a nice crisp metallic sound to the cymbals.
I owe a huge debt of gratitude to Thomas Mayer, Rod Coleman, Ale Moglia and Paul Leclercq for helping with this project, and to all members of the London Audiocircle for general support over the years.
Andy
When I started using the 4P1L my interest in the 300b SET revived. Here was a tube that worked exceptionally well in filament bias. As a triode it's mu was around 10 and it's anode resistance around 2k. Perfect! And the filaments were around 2v at 650mA, so quite easy to provide a DC filament supply for. They sound best at around 15mA or more, and I had some interstages I could try out so I rigged up an experimental amp. After starting with a 26 input stage I went over to two similar stages of 4P1L. This seemed to give the best tone to the instrumental sound - the 300b is on the warmer side and the 4P1L on the leaner side. The 4P1L had better treble than the 26 also. It is a little microphonic in the input stage but not grossly so, and it doesn't hum.
Having decided to use two similar stages with filament bias, I then started to play around with interstages. I like the 126C from Hammond in the driver stage at present - it is a rounder and fuller sound. In the input I like the LL1660/18mA which is a leaner sound with better treble detail. The combination seems to have good tone. I tried using LL1692A into LL1660/18mA but strangely this sounded too lean and I lost a bit of the magic. The LL1660 is better to my ears than the LL1692A, but not by much. It has better low level and treble detail at the expense of a slight loss of smoothness.
So I ended up with a very satisfying sounding 300b SET. I enclose a draft schematic to kick off some discussion. The voltages are approximate and need refining - I'm still at the tweaking stage. This has become my daily listening amp and gives me a lot of pleasure. It's transparent and detailed as you'd expect from DHTs and has particularly good tone and low level detail with orchestral instruments - you can clearly hear the individual tone quality of clarinets, bassoons, oboes and so on, and percussion in jazz groups is excellent - resonant thwack to the drum skins and a nice crisp metallic sound to the cymbals.
I owe a huge debt of gratitude to Thomas Mayer, Rod Coleman, Ale Moglia and Paul Leclercq for helping with this project, and to all members of the London Audiocircle for general support over the years.
Andy
Attachments
Last edited:
Nice project!!!
I need to add one stage to my 300B 10Y DHT SE Amp, but here is hard to find 4P1L ...
Thanks for idea Andy!
I need to add one stage to my 300B 10Y DHT SE Amp, but here is hard to find 4P1L ...
Thanks for idea Andy!
Quality of Ka bypass capacitor is concerned to sound quality much, the worst is electric, then bipolar. the higher value of this C, the worse sound of high frequencies.
I think if you use film or MKP capacitor for this situation then you may have a different feeling.
I you performe fixed bias for all circuit, so you don't need any more Ka capacitor. The sound you get maybe so good.
Last edited:
To answer the question of cathode bypass and fixed bias on 300b.
I tried all kinds of cathode bypasses - polypropylenes, electrolytics (awful), sandwiches with teflon bypasses etc etc. I used some clip leads and went through all my stock in various combinations. Everything put a veil over the sound, so I just gave up.
Tried fixed bias on the 300b and it did improve the bass, but I couldn't get a certain hardness out of the treble - tried polypropylene last cap in the DC supply, various voltage regs etc. Didn't try choke input. Didn't try more sophisticated shunt regs. I was getting variations in sound so it did seem this could be optimised, but I left it alone for the time being. You're right to suggest it.
In terms of topology, you could do this in various combinations of boxes since the stages are transformer coupled. Presently I have the first stage as a preamp, and 2 stages in the amp. I also considered 2 stages in the preamp and one stage in the amp - this would allow me to put the power supply in the amp with the output stage. Right now I have a separate power supply. I also have a separate filament supply for the input stage because it has choke input, and I want to keep the filament chokes away from the signal. They do have a field around them. This was crucial when I used the 26 in the first stage, but seems the 4P1L is less prone to hum. I run the 300b off DC, but this could be run off AC as well.
In terms of filament bias, there are a small number of DHTs that can be used without the DC supply getting ridiculous. 26 is easy and can be run off a 12v DC supply even. 01A is OK. 4P1L is good. Basically you want low filament voltage and current for the tube, plus a low voltage bias point. 10Y is not impossible since the bias voltage is lower than most, but you have to add on 7.5v for the filaments, and the filament current means big transformers. Sounds good in filament bias as Thomas has proved. I used to use that but found the 4P1L much easier to use, and also it sounds extremely good. More gain and lower plate resistance too. Where it gets a bit out of hand is 46, which is a great tube otherwise. At the recommended operating point of -33v you'd need a filament supply of at least 40v at 2.5 amps. The cathode resistor would be 13.5 ohms with 82 watts through it - would need a 200 watt part at least! With the 4P1L the filament supplies are quite sensible - 650mA and anything up to around 24vDC. Cathode resistor should be OK at 20W. These resistors get hot and should be 2 or 3 times the rating.
I'll see if I can put together some pics.
I tried all kinds of cathode bypasses - polypropylenes, electrolytics (awful), sandwiches with teflon bypasses etc etc. I used some clip leads and went through all my stock in various combinations. Everything put a veil over the sound, so I just gave up.
Tried fixed bias on the 300b and it did improve the bass, but I couldn't get a certain hardness out of the treble - tried polypropylene last cap in the DC supply, various voltage regs etc. Didn't try choke input. Didn't try more sophisticated shunt regs. I was getting variations in sound so it did seem this could be optimised, but I left it alone for the time being. You're right to suggest it.
In terms of topology, you could do this in various combinations of boxes since the stages are transformer coupled. Presently I have the first stage as a preamp, and 2 stages in the amp. I also considered 2 stages in the preamp and one stage in the amp - this would allow me to put the power supply in the amp with the output stage. Right now I have a separate power supply. I also have a separate filament supply for the input stage because it has choke input, and I want to keep the filament chokes away from the signal. They do have a field around them. This was crucial when I used the 26 in the first stage, but seems the 4P1L is less prone to hum. I run the 300b off DC, but this could be run off AC as well.
In terms of filament bias, there are a small number of DHTs that can be used without the DC supply getting ridiculous. 26 is easy and can be run off a 12v DC supply even. 01A is OK. 4P1L is good. Basically you want low filament voltage and current for the tube, plus a low voltage bias point. 10Y is not impossible since the bias voltage is lower than most, but you have to add on 7.5v for the filaments, and the filament current means big transformers. Sounds good in filament bias as Thomas has proved. I used to use that but found the 4P1L much easier to use, and also it sounds extremely good. More gain and lower plate resistance too. Where it gets a bit out of hand is 46, which is a great tube otherwise. At the recommended operating point of -33v you'd need a filament supply of at least 40v at 2.5 amps. The cathode resistor would be 13.5 ohms with 82 watts through it - would need a 200 watt part at least! With the 4P1L the filament supplies are quite sensible - 650mA and anything up to around 24vDC. Cathode resistor should be OK at 20W. These resistors get hot and should be 2 or 3 times the rating.
I'll see if I can put together some pics.
Member
Joined 2009
Paid Member
Power supplies:
300b filaments: 2x9v transformer, Schottky diodes, 20,000uF caps, Rod Coleman filament supply board, available from him
4P1L filaments: 2x24v transformer, Schottky diodes, 280mH choke (2.4A), 20,000uF caps, Rod Coleman filament supply board
Amp HT: 390v secondary on mains transformer, 2x 6BY5G damper diodes in parallel in a hybrid Bridge, 10uF cap, 5uF choke, 80uF, dropper resistor, 80uF. All caps polypropylene.
4P1L driver stage: 180v glow tubes (voltage regs) off same HT
Preamp HT: 230v secondary on transformer, 80 rectifier, 2uF input cap, 10H choke, 80uF cap, 10H choke, 40uF cap. All caps polypropylene.
Andy
300b filaments: 2x9v transformer, Schottky diodes, 20,000uF caps, Rod Coleman filament supply board, available from him
4P1L filaments: 2x24v transformer, Schottky diodes, 280mH choke (2.4A), 20,000uF caps, Rod Coleman filament supply board
Amp HT: 390v secondary on mains transformer, 2x 6BY5G damper diodes in parallel in a hybrid Bridge, 10uF cap, 5uF choke, 80uF, dropper resistor, 80uF. All caps polypropylene.
4P1L driver stage: 180v glow tubes (voltage regs) off same HT
Preamp HT: 230v secondary on transformer, 80 rectifier, 2uF input cap, 10H choke, 80uF cap, 10H choke, 40uF cap. All caps polypropylene.
Andy
Thanks Andy.
Filament Bias can work very well - all amps sound better without the electrolytic capacitor bypassing the cathode resistor. The payback is that the filament supply must be VERY quiet - the current noise in the heating supply is directly impressed on the cathode resistor AND mixed into the cathode current within the filament itself - as dictated by Kirchoff's Law.
I can prepare special high voltage versions of the Filament Regulators that can handle the ~25V rectified dc supplies noted on Andy's circuit. They are self-assembly kits. These are all ultra-low noise and have caused a lot of surprise by how much they improve the sound in any DHT - (see 'New DHT Heater' thread, last few pages)
Filament Bias can work very well - all amps sound better without the electrolytic capacitor bypassing the cathode resistor. The payback is that the filament supply must be VERY quiet - the current noise in the heating supply is directly impressed on the cathode resistor AND mixed into the cathode current within the filament itself - as dictated by Kirchoff's Law.
I can prepare special high voltage versions of the Filament Regulators that can handle the ~25V rectified dc supplies noted on Andy's circuit. They are self-assembly kits. These are all ultra-low noise and have caused a lot of surprise by how much they improve the sound in any DHT - (see 'New DHT Heater' thread, last few pages)
Hi Rod,
I credited Rod and Thomas Mayer because they've really put filament bias on the map. Thomas revived it from some dusty archive which everyone else had forgotten and made it work using chokes in the filament supply - excellent idea, and I use choke input in my filament supplies to the first stages. Rod happened to have developed a really good DC filament supply at about the same time and then quite a few people including myself started using filament bias with Rod's nice clean DC supply with their 26 preamp projects. It's pretty ideal for a 26, but it's really good for a 4P1L as well - this is where Ale, Anatoliy and some others came in and convinced us to use this rather forgotten Russian tube that was originally an old Grerman Wermacht design. Though it must be said that Klaus (klausmobil) had circuits for the 4P1L driving 6C4C dating back years and years if anyone had the sense to try them out. Turns out it's just brilliant, and it's even kicked out my beloved 26s. The combination of all this has taken the sound of my amplifiers up a whole level.
It's funny that every now and then you think you've reached the end stop with building an amp, and then a whole lot of new ideas come out and you're jumping with joy again! Never ceases to amaze me that there's always something better round the corner. And most of all this tube stuff dates back to the 1920s. Astonishing.
Andy
I credited Rod and Thomas Mayer because they've really put filament bias on the map. Thomas revived it from some dusty archive which everyone else had forgotten and made it work using chokes in the filament supply - excellent idea, and I use choke input in my filament supplies to the first stages. Rod happened to have developed a really good DC filament supply at about the same time and then quite a few people including myself started using filament bias with Rod's nice clean DC supply with their 26 preamp projects. It's pretty ideal for a 26, but it's really good for a 4P1L as well - this is where Ale, Anatoliy and some others came in and convinced us to use this rather forgotten Russian tube that was originally an old Grerman Wermacht design. Though it must be said that Klaus (klausmobil) had circuits for the 4P1L driving 6C4C dating back years and years if anyone had the sense to try them out. Turns out it's just brilliant, and it's even kicked out my beloved 26s. The combination of all this has taken the sound of my amplifiers up a whole level.
It's funny that every now and then you think you've reached the end stop with building an amp, and then a whole lot of new ideas come out and you're jumping with joy again! Never ceases to amaze me that there's always something better round the corner. And most of all this tube stuff dates back to the 1920s. Astonishing.
Andy
Last edited:
Well done Andy!
This reminds me when I did my first side by side comparison between my 4p1l preamp against the 26. I preferred the 4p1L but soon tried to convince myself that my 4p1l design and build was improved from my original 26.
Definitely will incorporate some of your ideas in next designs...
Ale
This reminds me when I did my first side by side comparison between my 4p1l preamp against the 26. I preferred the 4p1L but soon tried to convince myself that my 4p1l design and build was improved from my original 26.
Definitely will incorporate some of your ideas in next designs...
Ale
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.