I’ve been contemplating this project over years now while building guitar amps and just recently a stereo SE 6L6 amp as a tryout.
I’m a noob when it comes to hifi and have been lurking for a couple of years. I don’t have a listing room, will be in the living room, and haven’t acquired speakers yet. It will not be a pair of $$$ ones but something nice from the 70’s. Have a lot of good elements that I’ve been collecting over the years including a pair of Stentorian H.F 1012 but no cabinets nor tweeters. They are expensive so should probably be good? Maybe to good for me so I might sell them.
I have some matched quads of 807’s, eight matched 7193’s, 6SN7’s, ECC35’s and various pentodes.
I’m leaning to triode strapped 807’s and have a pair of Danbury OPT’s 25W, 6.6K/8ohm and a pretty massive PT 330-0-330V, some 10H chokes (one Lundahl massive choke that might be sufficient for the whole power rail in both half’s and of course a chassis.
Load lines aren’t my most read up topic so I do have some questions.
1) If I triode strap them and use a Vp of 360-390V, would a load of 3.3 or 6.6K be preferred?
Also leaning to a Mullard design in homage to the OPT’s since they were apparently the ones used in the old design and would prefer not to use global NFB.
2) a fun thing for looks would be to have as many top cap tubes as possible so I would like to use the 7193’s as the LTP. The children are older and wiser so i don’t expect them to fiddle with the amp when it’s up and running. Any considerations regarding that tube?
3) For the preamp would it just be to slab a 6SJ7/6J7/EF37A up front after the 100K input?
4) since I’m no hifi guy the RIAA really intimidates me. I do have a record player with an integrated RIAA. Would that be enough?
5) sounds like a good plan? What kind of sensitivity speakers would I be needed? Output probably around 8-12W/channel.
6) does anyone knows a smart and easy way to add a Bluetooth in the amp? If so, where in the circuit? Would the preamp pentode in combination with a LTP give the Bluetooth signal to much gain?
My wife would probably want to play her Spotify lists. I would really like to get rid of our Sonos system and go analog.
7) does anyone have a schematic that they will share that is about what I described🙂? I have collected various over the years so otherwise I can just take bits and pieces from them and make a scrapbook of the circuit.
Thanks for my time.
I’m a noob when it comes to hifi and have been lurking for a couple of years. I don’t have a listing room, will be in the living room, and haven’t acquired speakers yet. It will not be a pair of $$$ ones but something nice from the 70’s. Have a lot of good elements that I’ve been collecting over the years including a pair of Stentorian H.F 1012 but no cabinets nor tweeters. They are expensive so should probably be good? Maybe to good for me so I might sell them.
I have some matched quads of 807’s, eight matched 7193’s, 6SN7’s, ECC35’s and various pentodes.
I’m leaning to triode strapped 807’s and have a pair of Danbury OPT’s 25W, 6.6K/8ohm and a pretty massive PT 330-0-330V, some 10H chokes (one Lundahl massive choke that might be sufficient for the whole power rail in both half’s and of course a chassis.
Load lines aren’t my most read up topic so I do have some questions.
1) If I triode strap them and use a Vp of 360-390V, would a load of 3.3 or 6.6K be preferred?
Also leaning to a Mullard design in homage to the OPT’s since they were apparently the ones used in the old design and would prefer not to use global NFB.
2) a fun thing for looks would be to have as many top cap tubes as possible so I would like to use the 7193’s as the LTP. The children are older and wiser so i don’t expect them to fiddle with the amp when it’s up and running. Any considerations regarding that tube?
3) For the preamp would it just be to slab a 6SJ7/6J7/EF37A up front after the 100K input?
4) since I’m no hifi guy the RIAA really intimidates me. I do have a record player with an integrated RIAA. Would that be enough?
5) sounds like a good plan? What kind of sensitivity speakers would I be needed? Output probably around 8-12W/channel.
6) does anyone knows a smart and easy way to add a Bluetooth in the amp? If so, where in the circuit? Would the preamp pentode in combination with a LTP give the Bluetooth signal to much gain?
My wife would probably want to play her Spotify lists. I would really like to get rid of our Sonos system and go analog.
7) does anyone have a schematic that they will share that is about what I described🙂? I have collected various over the years so otherwise I can just take bits and pieces from them and make a scrapbook of the circuit.
Thanks for my time.
Attachments
Maybe this is a good starting point as a straw man.
This is an amplifier i would like to build which sounds similar to your plan, the Brimar 25P1. There is also a preamp circuit. 5B/255M is an 807 on a loctal base.
I think the OPT is a similar spec to yours.
This is an amplifier i would like to build which sounds similar to your plan, the Brimar 25P1. There is also a preamp circuit. 5B/255M is an 807 on a loctal base.
I think the OPT is a similar spec to yours.
I tried the 807 triode connected, search for 807 in title and my nickname.
Your transformer will give between 450 to 500 VDC, and the g2 of 807 does not like it. It works but premature failure could be the consequence. You can expect around 10 Watts in triode mode.
Eventually I overhauled my first prototype, with 350 V B+, UL, 12 dB negative feedback, 3.2k : 8R transformer. And a lot of 450 V capacitors in the PSU.
Your transformer will give between 450 to 500 VDC, and the g2 of 807 does not like it. It works but premature failure could be the consequence. You can expect around 10 Watts in triode mode.
Eventually I overhauled my first prototype, with 350 V B+, UL, 12 dB negative feedback, 3.2k : 8R transformer. And a lot of 450 V capacitors in the PSU.
Icsaszar, if I use a CT 330V PT it would probably be around 440 under load? Then a 5U4 with its 50V voltage drop around 390V.
Wouldn’t G2 in a triode strapped mode handle around 400V? I think I read that somewhere. 300V in pentode/UL mode but around 400V in triode mode since the plate and G2 would be the same?
If not a string of zener on the CT to drop the voltage to around 350V?
Wouldn’t G2 in a triode strapped mode handle around 400V? I think I read that somewhere. 300V in pentode/UL mode but around 400V in triode mode since the plate and G2 would be the same?
If not a string of zener on the CT to drop the voltage to around 350V?
Your power-transformer calculations are about right, I think. With cathode bias your 807s will only see about 360 volts from plate to cathode. The 807 grids will tolerate up to 400VDC without a problem.
The issue then becomes triode vs. pentode, also feedback vs. no feedback. 6.6K is a bit low for triode-strapped 807s or KT66s. The typicaly load in a early triode Williamson using those tubes was 9-10K, but that also assumed feedback was used. Your output transformers are, as has been pointed out, much more suited for an ultralinear amplifier with feedback. I'm a huge fan of the Williamson design myself, but the Mullard topology, like the Brimar amp referred to, certainly has its attractions--it's generally easier to stabilize.
Which brings us to the feedback question. There's NO reason to avoid it, either sonically or technically. Properly implemented it wil produce a much better amp than without it. If you choose NOT to use, you have to design a different amp because otherwise a Williamson or Mullard circuit will have way too much gain. So the decision has to be made before you build.
The caveat about feedback is that you really need an oscilloscope to make sure the amp is stable and operating properly. You can buy a decent USB-based scope unit for less than $150US. But that's an investment. But if you plan to continue with this hobby it's a very good investment to make.
Personally, I would not try to incorporate a phono stage. It makes the amp far more complicated, and introduces issues with noise etc. Use the one in your turntable or save up for a better outboard model.
There are any number of small bluetooth receivers on the market. Lots of people add them to old consoles. It should not be hard to incorporate one in the amp.
The issue then becomes triode vs. pentode, also feedback vs. no feedback. 6.6K is a bit low for triode-strapped 807s or KT66s. The typicaly load in a early triode Williamson using those tubes was 9-10K, but that also assumed feedback was used. Your output transformers are, as has been pointed out, much more suited for an ultralinear amplifier with feedback. I'm a huge fan of the Williamson design myself, but the Mullard topology, like the Brimar amp referred to, certainly has its attractions--it's generally easier to stabilize.
Which brings us to the feedback question. There's NO reason to avoid it, either sonically or technically. Properly implemented it wil produce a much better amp than without it. If you choose NOT to use, you have to design a different amp because otherwise a Williamson or Mullard circuit will have way too much gain. So the decision has to be made before you build.
The caveat about feedback is that you really need an oscilloscope to make sure the amp is stable and operating properly. You can buy a decent USB-based scope unit for less than $150US. But that's an investment. But if you plan to continue with this hobby it's a very good investment to make.
Personally, I would not try to incorporate a phono stage. It makes the amp far more complicated, and introduces issues with noise etc. Use the one in your turntable or save up for a better outboard model.
There are any number of small bluetooth receivers on the market. Lots of people add them to old consoles. It should not be hard to incorporate one in the amp.
Thanks for replying. I do have a scope and know the basics in how to use it. Mainly for troubleshooting…
I’m not totally against NFB but I am a bit afraid of building a Williamson amp due to the “huge” amount of feedback and a bit uncertain that my OT’s will be able to handle it and in a guitar amp design a high level of NFB will make the amp sound a bit sterile.
So a Mullard design with some dialed amount of NFB sounds easiest and hopefully a well sounding amp. Will keep the RIAA out of there🙂
I’m not totally against NFB but I am a bit afraid of building a Williamson amp due to the “huge” amount of feedback and a bit uncertain that my OT’s will be able to handle it and in a guitar amp design a high level of NFB will make the amp sound a bit sterile.
So a Mullard design with some dialed amount of NFB sounds easiest and hopefully a well sounding amp. Will keep the RIAA out of there🙂
That makes sense. You can fiddle with the gain of a Mullard design by adjusting the voltage amp stage (say, a triode-strapped EF86 instead of pentode). A Williamson doesn't really give you that flexibility if you want to maintain the integrity of the design.
On the other hand, a Williamson does not by any stretch sound "sterile"! It can have tremendous warmth and beauty. Mine certainly do--delicious, natural midrange, smooth highs and excellent bass. Just about any decent tube amp built after 1950 used 20dB of feedback, it's a bit of a "magic number." Unless you consider an Eico or Leak or Marantz or other classic amp to be "sterile," then you should let go of that idea. ;-)
On the other hand, a Williamson does not by any stretch sound "sterile"! It can have tremendous warmth and beauty. Mine certainly do--delicious, natural midrange, smooth highs and excellent bass. Just about any decent tube amp built after 1950 used 20dB of feedback, it's a bit of a "magic number." Unless you consider an Eico or Leak or Marantz or other classic amp to be "sterile," then you should let go of that idea. ;-)
The "STC 807 application report" tells us pretty much everything there is to know about the 807, including that they can operate at 400V in triode connection.
I've been running a pair at 370V myself for quite some time now without problems. Not sure if that applies to UL though, I believe the screens are better protected when the tubes are triode wired.
I agree that 6,6k is a bit on the low side for a pair of 807s but it should work, my PP monoblocks uses 5,8k transformers although at a lower voltage (295V) and in class A2. You could use that big Lundahl choke to drop the B+ a bit by using a small cap after the rectifier in a semi choke input fashion.
I've been running a pair at 370V myself for quite some time now without problems. Not sure if that applies to UL though, I believe the screens are better protected when the tubes are triode wired.
I agree that 6,6k is a bit on the low side for a pair of 807s but it should work, my PP monoblocks uses 5,8k transformers although at a lower voltage (295V) and in class A2. You could use that big Lundahl choke to drop the B+ a bit by using a small cap after the rectifier in a semi choke input fashion.
I guess what I'm saying is, don't ignore the Williamson. Your OPT's were surely designed to handle 20dB feedback. It's an incredible sounding amp, and folks here can help with tuning if you have a scope and post your results.
Choke Input power supply should be on with the OP’s power transformer for a 807 G2 compliant B+
My 2 cents, ymmv etc
330V x0.9 ~= 297
Subject to V drop via the choke of course, so probably a bit lower.
300V is maximum G2 for an 807.
Subject to V drop via the choke of course, so probably a bit lower.
300V is maximum G2 for an 807.
G2 max is 300v in pentode mode. According to RCA 400v is fine in triode. https://tube-data.com/sheets/049/8/807.pdf
Plenty of people run 807 at 400v triode with no problem.
Plenty of people run 807 at 400v triode with no problem.
The circuit I linked in post #2 runs the 5B/255M (807 equivalents) at 440V in UL with 390R as a screen grid stopper, so I think it boils down to the dissipation on the screen and staying within the limits.
I use those tubes in my Quad IIs, but I fitted a 1k screen grid resistor to be on the safe side. Va is 340V and Vg2 is 330V in that amp.
I use those tubes in my Quad IIs, but I fitted a 1k screen grid resistor to be on the safe side. Va is 340V and Vg2 is 330V in that amp.
It's a bit odd as it has 2 coupling caps whereas Mullards typically benefit from only having the one between the splitter and the output stage.
I am not sure what your point is. It is not a Mullard circuit; Brimar had their own tubes to push.
I think he means Mullard topology, where the first voltage amplification stage is DC-coupled (no cap) to the phase inverter.
They indeed also came with the Maplin Millennium 4-20 kit (a Williamson clone): https://livinginthepast-audioweb.co.uk/millennium/index.html