I have been playing with a design for a friend using 5B/255M valves. These are basically a loctal version of 807 beam power tetrodes. My initial design and breadboard was a PP amplifier and it performed very well - but I thought I should knock up a SE version to compare with, in case he prefered that. The SE was not bad but was clearly worse than the PP. To try and see a little of what was going on here I measured the distortion in both modes and found the PP to be much, much lower for 2nd harmonic and about half as much third harmonic too, so I'm not surprised that the PP sounds better (in this case).
So then I thought I would try the out of fashion Tetrode Class AB2 PP just for fun. So I tweaked it to Class AB2 added a smidgeon of feedback (6dB) and sat down to listen.... I was very surprised! It sounds really good. Actually really really good! Close to the A1 pp and better than the SE! So to measuring again. Surprise it has a nice distortion spectrum! Second higher than A1 PP but much less than SE, third nicely less than second... hmm...
My speakers are fullrange about 7ohms impedance in a reflex cabinet and about 95dB sensitivity. Source battery powered CD player. All amplifiers fixed bias and IT interstage. Driver tubes 6N6P. PP amps are PP throughout.
Here is the graph of the measurements.
So then I thought I would try the out of fashion Tetrode Class AB2 PP just for fun. So I tweaked it to Class AB2 added a smidgeon of feedback (6dB) and sat down to listen.... I was very surprised! It sounds really good. Actually really really good! Close to the A1 pp and better than the SE! So to measuring again. Surprise it has a nice distortion spectrum! Second higher than A1 PP but much less than SE, third nicely less than second... hmm...
My speakers are fullrange about 7ohms impedance in a reflex cabinet and about 95dB sensitivity. Source battery powered CD player. All amplifiers fixed bias and IT interstage. Driver tubes 6N6P. PP amps are PP throughout.
Here is the graph of the measurements.
Attachments
Hi James,
Interesting post, and one which basically tallies with both what I've read, and what I've measured myself in the past. The 807 is one of my fave tubes (I have boxes of them in all flavours). The STC 807 datasheet (5.5M) might be of interest if you haven't already seen it. Excellent data with lots of graphs of different output stage configurations and distortions for 2H and 3H
If you have the facilities, it's very interesting to note the higher harmonic content in the different configurations, and how they change with level.
I'd be interested in seeing your schem for the A1 PP if that's possible.
BTW: do you have a source for 6N6P's? I've been trying to get my hands on some to try for a while now.
Cheers
Interesting post, and one which basically tallies with both what I've read, and what I've measured myself in the past. The 807 is one of my fave tubes (I have boxes of them in all flavours). The STC 807 datasheet (5.5M) might be of interest if you haven't already seen it. Excellent data with lots of graphs of different output stage configurations and distortions for 2H and 3H
If you have the facilities, it's very interesting to note the higher harmonic content in the different configurations, and how they change with level.
I'd be interested in seeing your schem for the A1 PP if that's possible.
BTW: do you have a source for 6N6P's? I've been trying to get my hands on some to try for a while now.
Cheers
Crossover distortion?
Interesting. Were you able to measure distortion at lower levels with the AB2 configuration? And, what Ra-a did you use? (I have some 5B/255Ms, you see.)
Interesting. Were you able to measure distortion at lower levels with the AB2 configuration? And, what Ra-a did you use? (I have some 5B/255Ms, you see.)
Hi Brett,
I hadn't seen that data set for the 807. Nice document! Thanks for the reference.
I had shown the schematic for the A1 pp in this thread
Simple IT PP . The input pot into the input transformer is not an optimum design...but it works after a fashion. I'm having some multiple tapped autotransformers built to tryn as a TVA (I guess AVA, or ATVA) for the real version.
On test I ended up changing the OPT tap to make it Ra-a=7K instead of the 11K shown in the diagram. Otherwise it's as shown. The pp balance pot was adjusted for best sound which was also minimum distortion ( as near as makes no difference) and it was with about 8% dc imbalance...
I have a great source for all sorts of valves. It's a chap in Germany, here is the URL
Ask Jan First . Jan is a great chap and very helpful. Please mention my name if you use him.
Hi EC8010
The Ra-a was 7.5K also in AB2. I didn't measure any other harmonics as my kit at home doesn't do it. I guess I could take it into work and put the Spectrum analyser on the output... Work is a bit busy at present so it might be a couple of weeks before we get to that. I'm curious as to the extent of crossover distortion and to the upper harmonic signature too. Through my speakers I doubt I pushed it beyond 10 watts at most. I was surprised at how relaxed it sounded and how good it's imaging was. I'm very impressed with the 5B/255Ms. BTW I didn't run them very hot - only 450volts on the anode and 250 on the screen grid. Also after 40watts the 3rd harmonic climbed away but the second had peaked and dropped down by about 30% after that...
ciao
James
I hadn't seen that data set for the 807. Nice document! Thanks for the reference.
I had shown the schematic for the A1 pp in this thread
Simple IT PP . The input pot into the input transformer is not an optimum design...but it works after a fashion. I'm having some multiple tapped autotransformers built to tryn as a TVA (I guess AVA, or ATVA) for the real version.
On test I ended up changing the OPT tap to make it Ra-a=7K instead of the 11K shown in the diagram. Otherwise it's as shown. The pp balance pot was adjusted for best sound which was also minimum distortion ( as near as makes no difference) and it was with about 8% dc imbalance...
I have a great source for all sorts of valves. It's a chap in Germany, here is the URL
Ask Jan First . Jan is a great chap and very helpful. Please mention my name if you use him.
Hi EC8010
The Ra-a was 7.5K also in AB2. I didn't measure any other harmonics as my kit at home doesn't do it. I guess I could take it into work and put the Spectrum analyser on the output... Work is a bit busy at present so it might be a couple of weeks before we get to that. I'm curious as to the extent of crossover distortion and to the upper harmonic signature too. Through my speakers I doubt I pushed it beyond 10 watts at most. I was surprised at how relaxed it sounded and how good it's imaging was. I'm very impressed with the 5B/255Ms. BTW I didn't run them very hot - only 450volts on the anode and 250 on the screen grid. Also after 40watts the 3rd harmonic climbed away but the second had peaked and dropped down by about 30% after that...
ciao
James
Spectrum analyser
Hello James,
very few spectrum analysers have the resolution to look a long way down on their own, but if you hang the spectrum analyser off the oscilloscope output of the distortion analyser, it adds the dynamic ranges together. This subterfuge allows me to look reliably down to -90dB, and occasionally further, limited primarily by oscillator purity.
Of course, at this point, you start thinking, "Wouldn't it be nice to have an AP System 2, or even a 1, surely they are due to turn up second-hand soon. Please?"
Hello Brett,
my (extremely lethargic) printer is starting to think about the possibility of printing the first of the 43 pages of 807 data sheet. It looks like a real gem - thanks for the link.
Hello James,
very few spectrum analysers have the resolution to look a long way down on their own, but if you hang the spectrum analyser off the oscilloscope output of the distortion analyser, it adds the dynamic ranges together. This subterfuge allows me to look reliably down to -90dB, and occasionally further, limited primarily by oscillator purity.
Of course, at this point, you start thinking, "Wouldn't it be nice to have an AP System 2, or even a 1, surely they are due to turn up second-hand soon. Please?"
Hello Brett,
my (extremely lethargic) printer is starting to think about the possibility of printing the first of the 43 pages of 807 data sheet. It looks like a real gem - thanks for the link.
Re: RAIN FORREST.
Hello Frank,
that's terribly 80s of you! In the 90s we worried about skin cancer due to the hole on the ozone layer. This year we simply worry about WWIII. Whoops, wrong thread.
fdegrove said:
Hello Frank,
that's terribly 80s of you! In the 90s we worried about skin cancer due to the hole on the ozone layer. This year we simply worry about WWIII. Whoops, wrong thread.
One of my favourites too. I have just one, brand new in box made by RCA in 1972. Anyway, I once saw a cct that used a pair of 807's running at about 700 volts or so and zero bias and had *all three* grids of each tube strapped together and driven with a transformer (grid current BTW) and it was supposed to get 120 watts at 2% distortion. It was a modulator for an AM transmitter.That's all I can remember now. Never seen that approach used by anyone.Brett said:
Re: It takes a worried man...
My three principal worries are
a) Trying to find 8116s/YL1071s at sensible prices
b) The outrageous level of tobacco tax in England
c) Why can't I "get more" smilies??
7N7
EC8010 said:
My three principal worries are
a) Trying to find 8116s/YL1071s at sensible prices
b) The outrageous level of tobacco tax in England
c) Why can't I "get more" smilies??
7N7
Hi EC8010
At work I have access to an HP 3585A analyser and an AP1. So when I can finally get my amps into the lab I can produce some proper measurements and, hopefully, some piccies and graphs to explore the performance properly.
Unfortunately I dumped a large testing programme onto the lab a couple of weeks ago, so now I have to be patient! I got desperate and bid on a Lindos set that came up here in Italy - fortunately I lost 🙂 My distortion meter is a beaten up old ferrograph and it does read low. Suprisingly the valve amps had very very good wow and flutter figures though
I echo your comments about needing an AP at home! Have you seen the list of test kit Jim Hagerman has? I'm seriously jealous
ciao
James
At work I have access to an HP 3585A analyser and an AP1. So when I can finally get my amps into the lab I can produce some proper measurements and, hopefully, some piccies and graphs to explore the performance properly.
Unfortunately I dumped a large testing programme onto the lab a couple of weeks ago, so now I have to be patient! I got desperate and bid on a Lindos set that came up here in Italy - fortunately I lost 🙂 My distortion meter is a beaten up old ferrograph and it does read low. Suprisingly the valve amps had very very good wow and flutter figures though
I echo your comments about needing an AP at home! Have you seen the list of test kit Jim Hagerman has? I'm seriously jealous
ciao
James
Ciao
io sto usando da qualche mese un SE di 807, rispetto allo schema c'è qualche modifica ma non nella sostanza.
Ho usato un carico un po' basso, ma ho un paio di 8K da provare al più presto
ciao
Filippo
www.audiofanatic.it
io sto usando da qualche mese un SE di 807, rispetto allo schema c'è qualche modifica ma non nella sostanza.
Ho usato un carico un po' basso, ma ho un paio di 8K da provare al più presto
ciao
Filippo
www.audiofanatic.it
Attachments
Test gear
The main thing I remember about the Ferrograph was that the knobs skinned your knuckles when you lifted it by a handle, but it was the only test set I was allowed to take home at the time (the BBC EP14 was verboten). How time has moved on. 7N7 now has my EP14, and I have an MJS401D. Actually, you were lucky in losing the Lindos auction. The LA100 is a very splendid piece of test gear (especially for tape machines) but its electronic balancing objects to being abused by being connected to unbalanced audio. I did this, and found clock noise on the audio when I looked with a Tek 2445B oscilloscope. Also, unlike the MJS and its Neutrik derivatives, it can't measure distortion at an arbitrary frequency.
James D. said:
The main thing I remember about the Ferrograph was that the knobs skinned your knuckles when you lifted it by a handle, but it was the only test set I was allowed to take home at the time (the BBC EP14 was verboten). How time has moved on. 7N7 now has my EP14, and I have an MJS401D. Actually, you were lucky in losing the Lindos auction. The LA100 is a very splendid piece of test gear (especially for tape machines) but its electronic balancing objects to being abused by being connected to unbalanced audio. I did this, and found clock noise on the audio when I looked with a Tek 2445B oscilloscope. Also, unlike the MJS and its Neutrik derivatives, it can't measure distortion at an arbitrary frequency.
Ciao Filippo,
Forgive my replying in english. My Italian is almost non-existent despite living in Roma for one year...
Your circuit is a nice SE one. Have you tried the 807 in triode mode as well as tetrode mode? If you have, what did you find of the difference in sound?
I like your site and have often referred to it for ideas for circuits.
Hi EC8010,
I must admit that I like the ferrograph set - always did and it only cost me a £8 as it is so beaten up! I knew the Lindos wasn't right for what I needed - The MJS401D is a great set. Used one many times at Thames TV - Teddington. I must get some decent test kit again! I've always been lazy on test kit as I've always used stuff from work. Now my sons want to participate I need decent stuff at home...
ciao
James
Forgive my replying in english. My Italian is almost non-existent despite living in Roma for one year...
Your circuit is a nice SE one. Have you tried the 807 in triode mode as well as tetrode mode? If you have, what did you find of the difference in sound?
I like your site and have often referred to it for ideas for circuits.
Hi EC8010,
I must admit that I like the ferrograph set - always did and it only cost me a £8 as it is so beaten up! I knew the Lindos wasn't right for what I needed - The MJS401D is a great set. Used one many times at Thames TV - Teddington. I must get some decent test kit again! I've always been lazy on test kit as I've always used stuff from work. Now my sons want to participate I need decent stuff at home...
ciao
James
Hi James
also my english is pretty rough...
i didn't try the triode connection because the HT is too high. I choose an high HT for achieve a class A2 driving with an high polarization. Actually the cathode voltage is approax 15V, and the high swing of the PCL82 can drive the 807 into the positive grid area and the low R load on the anode of the driver permits also a good current if the grid of the 807 need it.
ciao
Filippo
www.audiofanatic.it
also my english is pretty rough...
i didn't try the triode connection because the HT is too high. I choose an high HT for achieve a class A2 driving with an high polarization. Actually the cathode voltage is approax 15V, and the high swing of the PCL82 can drive the 807 into the positive grid area and the low R load on the anode of the driver permits also a good current if the grid of the 807 need it.
ciao
Filippo
www.audiofanatic.it
Hi Filippo,
Good point about the high B+ on the 807. I have been told that in triode mode one can run the screen grid over it's normal voltage rating but I wouldn't like to go too far over it!
I like the PCL82 valve! My office system is an ECL82 SE design and works very well.
ciao
James
Good point about the high B+ on the 807. I have been told that in triode mode one can run the screen grid over it's normal voltage rating but I wouldn't like to go too far over it!
I like the PCL82 valve! My office system is an ECL82 SE design and works very well.
ciao
James
James D. said:
The manufacturers permitted 400V for 807 in triode mode.
I have tested 13E1 (Vg2 max 300V) at 500V (and 200mA) without any problems and have recently been running 829s (Vg2 max 225V) at 370V and 40W as triodes, again with no trouble.
I have replaced the 829s with QQV07-50s (Vg2 max 250) and they are perfectly happy at 370V and 115mA each.
I am shortly going to be looking into some higher voltage types. I'll report back (I hope).
7N7
- Status
- Not open for further replies.