djn, this is your guy over there in the US:
https://youtu.be/Gd8PSpeWCxo
Really skilled and tooled up for delicate ally work.
Cheers
Matt
https://youtu.be/Gd8PSpeWCxo
Really skilled and tooled up for delicate ally work.
Cheers
Matt
As long as the Power Transformer can handle the current of the heaters and any additional current the 807's might produce, I'd go for it.
Simulate it out in a spice program- simetrix is what i prefer- just to be sure currents, voltages and power will be in range of the transformers.
The finals should work okay with the anode supply in between the first two columns from RCA's datasheet, but you need roughly double the grid bias voltage and grid drive signal, which may be asking too much from the original drive circuit. And if you alter the drive circuit, then you must be prepared to re-optimize the feedback network.
Attachments
First post on forum - my experience with 807
Como vai FullRangeMan, hello fellow DIYers,
I'm in a similar situation as you are, running full range speakers (Fostex FE204 Voigt pipes - around 95dB), and tried them with a host of 807 amps, both PP and SE, through many incarnations and mods.
I tried the 807 out first in a Heathkit W2, which is PP with feedback. Modified the W2 for 807: connected it up in what might be called a "fake-triode mode" or "voltage compensated triode". I never seen such operation reported before by users, but there is mention of such strapping in the 807STC datasheets: G2 running to the plate though a volatge dropping resistor - which is kind'of triode as the electrodes are tied together, but G2 is kept at at lower voltage by the voltage dropping resistor. In reality, I've no idea how the tube operates in such strapping, but indeed, it does sound like a triode. B+ is about 490V. Amp was super stable, and has absolutely glorious sound, which opened up more by decreasing the feedback. The W2/807 plays with incredible authority. 70s music and rock come through so life-like it really grabs you. Super relaxed, glorious, huge sound with plenty of details - haven for Led Zeppelin addicts. I get the same feel listening to it as when being at a live concert. You have the palpable feeling that there is a huge crowd around you, and feel their anticipation adding to yours.
I ran a pair of Russian Zaerix brand 807s - which I heard might be military-grade 807s. Also tried US NOS RCA, which is reportedly the finest of 807s, and I got some premium ones. To my mentors dismay (he worshiped those US NOS 807s) the Russkies sounded much fuller, much more musical than his treasured RCAs. However, since, I have not found any Zaerix 807 - grab them if any ever turns up.
Afterwards, I have modified my Darling amp (I have two of them anyway): ditched the tubes, but held on to the CLCLC filtered 450VDC, and the Hammond 125ESE OPTs. Yes, those entry level buggers!!! 10K pri, 10H pri inductance. Quite a laugh, really. almost 10 times less inductance than desired. I tried Edcor as well. To me they are both similar, the tiny Edcor winning at dynamics, but the bigger size Hammond is producing more relaxed, more spacious sound with bigger low end. Used 6SD7GT as driver tube (which sets up just as a 717A, but can be pushed harder than that, has a separate supressor, and is not good for GHz service), cap coupled to 807 in tetrode mode. The amp was idling about 22W. I reserved the Zaerix 807 for special occasions, and ran the US NOS in the SE 807s. After 8-9 years, with average 2-8 hours a day on time (sometimes I don't turn it off for a whole week) the tubes tested 95% of their new values. I would say they truly last forever @22WD! The amp sounded very good, just as my souped up Double DC Darling amp. Same character as a SE triode but way more power. No feedback. Shrillness, distortion? Zero shrillness, not even when pushed hard. Low end full and with control. It made me realize that the single ended triode sound is not as much attributed to the triode itself, but the single ended operation. Power was enough not only for the sensitive Voigt pipes, but the relatively less sensitive Vandersteens as well. Playing even base heavy stuff such as Dr Alban's 45RPM No coke on the Vandersteens really moved and shook the walls - well, it was like hearing it the first time. Wall-demolishing psychedelic techno party without subwoofers or substance abuse. It's really quite something to hear that kind of music from a quality rig. The poor Kirby salesman who just finished cleaning my carpet asked me what that curious thing between the speakers was, so I gave him a demo. He was literally crying tears and suddenly forgot about selling the vacuum cleaner and wanted to buy my entire setup... haha. I ruined him for life.
Years went by, and I did countless mods, tried many things out, but the amp stayed with same tube complement.
About a year ago I received the recommendation that 807s are a sleeper at low voltages and at 450 they do nothing but sleep. So, the 807 went to the design board - also by necessity, as the lytics were already clocking 15 years in the amp so it was due for a major overhaul. (Ironic - power tubes lasting longer than the caps!) Went for 700V B+, CLCLC supply with second LC section separate for each channel. Separate power supplies for driver and VR regulated supply for the 807 screens. I'm routinely used to 450V, and this is my first venture beyond 500V. Needless to say, I was sweating like crazy when fired it up the first time. Of course, did that with a variac and with rubber gloves. ; ). Nothing blew up, I did not get zapped, and during the past months I totally got used to the idea of 700V. I switch it on without a second thought.
Indeed, 700V was a huge eye opener! The 807 is a fine amp at 450VDC, but it goes through a super-transformation around 700V. I did a sweep with the high voltage - increased the B+ from 400V to 820V over a period of 3 hours, and spotted how it sounds. (Current of 807 is set by screen, which was kept 275V the whole time. That is, with a set screen voltage you get roughly the same current whether 400V or 750V is on the plate.)
400-580V range: very nice sound,
600-660V range: strained and hollow sound,
700+V: tremendous jump in sound quality in every aspect, and at
800V something magical happens: the tube leaps onto another level. Not a hint of strain in the sound, ambience is eerie - background noise DISAPPEARS, speakers DISAPPEAR entirely, a sudden shift from recorded material to the real world. It was utterly spooky and unexpected. Yet, I had no idea whether the whole thing will blow up or not, so shut it off at 820V. When you look at the datasheets, they do not go past 750V or so! However, my seasoned veteran friend who had plenty of experience with them said no worries, they can take 800V like a breeze.
Yet, I draw the line at 700VDC at the moment. There, with 10K pri OPT, the 807 prefers 26mA current, that's where it sounds the best. I'm pretty sure though that with a different load the current / voltage test will give different results, and you can find the right load that makes it sound right at 600V. (I'd guess 13K.)
At 700V/26mA/10Kpri OPT it is insanely dynamic, and to my huge surprise, it has very good bottom end. Yes, I can hear that a decent OPT would tighten it up significantly. Clearly, the amp is much more refined in the mids and highs than in the base. But still, that comparably loose base is just as tight, detailed and alive as that of a very good amp.
The amp does not have a genre preference, it shows the albums without further ado - to a much greater degree than anything I had before, or run across. It's ultra hi def, but without shoving it in your face. Zero fatigue. Imaging is the most stunning aspect: sound fills the room, no matter where you are: sit, stand or lay on the floor, turn towards speakers or show your back to them. It sounds as if people are in your house playing music. Music fills the room, does not seem to stick to the speakers. Gives tremendous freedom that you are not rooted to a bloody chair in the middle of the room, and others have to wait for their turn.
I would like "more" base on many of the records, though. Unfortunately, if it's not there, it will not fake it for you. However, material that has base really comes through. This amp produced the only credible organ reproduction I ever heard - and I'm an organ junkie, spent countless hours in European cathedrals listening to magic pouring from Arp Snitger miracles. I was shocked, as I never expected that credible organ reproduction is possible without an actual cathedral or at least a XVth century chapel with yard thick stone walls. Shocked me to my core, to have the breath of life coming from a measly 10H primary OPT, without any feedback, no subs! And it was not just my ears - I could feel the cover on the sofa ripple under me when the pedals worked. (Life-like organ volume levels in a 15ftx25ft room. I was amazed the house did not collapse.) Throughout the ordeal the Fostex drivers cones stayed in the same place, barely any noticeable movement. I've heard the same speakers with a PX-25 SE amp produce similar quantity base on Dead Can Dance, and then it was followed by 1in cone excursion - I feared for the destruction of the poor drivers.
Before you go crazy about the exact operating points for the 807, I can tell that the driver operating points impacted the sound much more. If your 807 doesn't sound as you like it - tweak the driver op points. A pentode driver is a good start, allows tremendous freedom of tweaking.
And do not go nuts about feedback or no feedback. To me, no feedback is a no brainer - even little feedback sounds closed and restrained to my ears, and it spoils the base. My experience is that feedback does not have any benefits at all when you have a full range system or have speakers which are phase and time aligned. However, you might have different hearing, and others might have time/phase misaligned speakers where feedback can help. Try it out. Easy to put in or to remove the feedback loop. I've tried using a variable resistor, to see what varying levels of feedback do. You can go back and forth from very little feedback to generous feedback. Go with a pot 20 times the size of your feedback resistor, and place the pot in series with the resistor: increasing resistance you will lower feedback. That was a real eye opener for me, listening to feedback change in real time. About the shrillness: I never experienced it, so no idea what people are talking about. Although I always use grid stoppers and what I found very important: cathode stoppers. Never ever leave your 807 cathodes unattended!
G2 stopper is also critical, but when you run VR regulation on G2, then I found no stopper is necessary, just run the VR tube directly to g2. Plate stopper also recommended, but works for me without just the same.
For driver stage: carbon copy the Yamamoto A08 driver stage (717A, zener regulated screen), and start with that. Spice it with a beefy CLCLC power supply. That's where I started, and it's the beginning of a wonderful road.
Have fun, Janos
Como vai FullRangeMan, hello fellow DIYers,
I'm in a similar situation as you are, running full range speakers (Fostex FE204 Voigt pipes - around 95dB), and tried them with a host of 807 amps, both PP and SE, through many incarnations and mods.
I tried the 807 out first in a Heathkit W2, which is PP with feedback. Modified the W2 for 807: connected it up in what might be called a "fake-triode mode" or "voltage compensated triode". I never seen such operation reported before by users, but there is mention of such strapping in the 807STC datasheets: G2 running to the plate though a volatge dropping resistor - which is kind'of triode as the electrodes are tied together, but G2 is kept at at lower voltage by the voltage dropping resistor. In reality, I've no idea how the tube operates in such strapping, but indeed, it does sound like a triode. B+ is about 490V. Amp was super stable, and has absolutely glorious sound, which opened up more by decreasing the feedback. The W2/807 plays with incredible authority. 70s music and rock come through so life-like it really grabs you. Super relaxed, glorious, huge sound with plenty of details - haven for Led Zeppelin addicts. I get the same feel listening to it as when being at a live concert. You have the palpable feeling that there is a huge crowd around you, and feel their anticipation adding to yours.
I ran a pair of Russian Zaerix brand 807s - which I heard might be military-grade 807s. Also tried US NOS RCA, which is reportedly the finest of 807s, and I got some premium ones. To my mentors dismay (he worshiped those US NOS 807s) the Russkies sounded much fuller, much more musical than his treasured RCAs. However, since, I have not found any Zaerix 807 - grab them if any ever turns up.
Afterwards, I have modified my Darling amp (I have two of them anyway): ditched the tubes, but held on to the CLCLC filtered 450VDC, and the Hammond 125ESE OPTs. Yes, those entry level buggers!!! 10K pri, 10H pri inductance. Quite a laugh, really. almost 10 times less inductance than desired. I tried Edcor as well. To me they are both similar, the tiny Edcor winning at dynamics, but the bigger size Hammond is producing more relaxed, more spacious sound with bigger low end. Used 6SD7GT as driver tube (which sets up just as a 717A, but can be pushed harder than that, has a separate supressor, and is not good for GHz service), cap coupled to 807 in tetrode mode. The amp was idling about 22W. I reserved the Zaerix 807 for special occasions, and ran the US NOS in the SE 807s. After 8-9 years, with average 2-8 hours a day on time (sometimes I don't turn it off for a whole week) the tubes tested 95% of their new values. I would say they truly last forever @22WD! The amp sounded very good, just as my souped up Double DC Darling amp. Same character as a SE triode but way more power. No feedback. Shrillness, distortion? Zero shrillness, not even when pushed hard. Low end full and with control. It made me realize that the single ended triode sound is not as much attributed to the triode itself, but the single ended operation. Power was enough not only for the sensitive Voigt pipes, but the relatively less sensitive Vandersteens as well. Playing even base heavy stuff such as Dr Alban's 45RPM No coke on the Vandersteens really moved and shook the walls - well, it was like hearing it the first time. Wall-demolishing psychedelic techno party without subwoofers or substance abuse. It's really quite something to hear that kind of music from a quality rig. The poor Kirby salesman who just finished cleaning my carpet asked me what that curious thing between the speakers was, so I gave him a demo. He was literally crying tears and suddenly forgot about selling the vacuum cleaner and wanted to buy my entire setup... haha. I ruined him for life.
Years went by, and I did countless mods, tried many things out, but the amp stayed with same tube complement.
About a year ago I received the recommendation that 807s are a sleeper at low voltages and at 450 they do nothing but sleep. So, the 807 went to the design board - also by necessity, as the lytics were already clocking 15 years in the amp so it was due for a major overhaul. (Ironic - power tubes lasting longer than the caps!) Went for 700V B+, CLCLC supply with second LC section separate for each channel. Separate power supplies for driver and VR regulated supply for the 807 screens. I'm routinely used to 450V, and this is my first venture beyond 500V. Needless to say, I was sweating like crazy when fired it up the first time. Of course, did that with a variac and with rubber gloves. ; ). Nothing blew up, I did not get zapped, and during the past months I totally got used to the idea of 700V. I switch it on without a second thought.
Indeed, 700V was a huge eye opener! The 807 is a fine amp at 450VDC, but it goes through a super-transformation around 700V. I did a sweep with the high voltage - increased the B+ from 400V to 820V over a period of 3 hours, and spotted how it sounds. (Current of 807 is set by screen, which was kept 275V the whole time. That is, with a set screen voltage you get roughly the same current whether 400V or 750V is on the plate.)
400-580V range: very nice sound,
600-660V range: strained and hollow sound,
700+V: tremendous jump in sound quality in every aspect, and at
800V something magical happens: the tube leaps onto another level. Not a hint of strain in the sound, ambience is eerie - background noise DISAPPEARS, speakers DISAPPEAR entirely, a sudden shift from recorded material to the real world. It was utterly spooky and unexpected. Yet, I had no idea whether the whole thing will blow up or not, so shut it off at 820V. When you look at the datasheets, they do not go past 750V or so! However, my seasoned veteran friend who had plenty of experience with them said no worries, they can take 800V like a breeze.
Yet, I draw the line at 700VDC at the moment. There, with 10K pri OPT, the 807 prefers 26mA current, that's where it sounds the best. I'm pretty sure though that with a different load the current / voltage test will give different results, and you can find the right load that makes it sound right at 600V. (I'd guess 13K.)
At 700V/26mA/10Kpri OPT it is insanely dynamic, and to my huge surprise, it has very good bottom end. Yes, I can hear that a decent OPT would tighten it up significantly. Clearly, the amp is much more refined in the mids and highs than in the base. But still, that comparably loose base is just as tight, detailed and alive as that of a very good amp.
The amp does not have a genre preference, it shows the albums without further ado - to a much greater degree than anything I had before, or run across. It's ultra hi def, but without shoving it in your face. Zero fatigue. Imaging is the most stunning aspect: sound fills the room, no matter where you are: sit, stand or lay on the floor, turn towards speakers or show your back to them. It sounds as if people are in your house playing music. Music fills the room, does not seem to stick to the speakers. Gives tremendous freedom that you are not rooted to a bloody chair in the middle of the room, and others have to wait for their turn.
I would like "more" base on many of the records, though. Unfortunately, if it's not there, it will not fake it for you. However, material that has base really comes through. This amp produced the only credible organ reproduction I ever heard - and I'm an organ junkie, spent countless hours in European cathedrals listening to magic pouring from Arp Snitger miracles. I was shocked, as I never expected that credible organ reproduction is possible without an actual cathedral or at least a XVth century chapel with yard thick stone walls. Shocked me to my core, to have the breath of life coming from a measly 10H primary OPT, without any feedback, no subs! And it was not just my ears - I could feel the cover on the sofa ripple under me when the pedals worked. (Life-like organ volume levels in a 15ftx25ft room. I was amazed the house did not collapse.) Throughout the ordeal the Fostex drivers cones stayed in the same place, barely any noticeable movement. I've heard the same speakers with a PX-25 SE amp produce similar quantity base on Dead Can Dance, and then it was followed by 1in cone excursion - I feared for the destruction of the poor drivers.
Before you go crazy about the exact operating points for the 807, I can tell that the driver operating points impacted the sound much more. If your 807 doesn't sound as you like it - tweak the driver op points. A pentode driver is a good start, allows tremendous freedom of tweaking.
And do not go nuts about feedback or no feedback. To me, no feedback is a no brainer - even little feedback sounds closed and restrained to my ears, and it spoils the base. My experience is that feedback does not have any benefits at all when you have a full range system or have speakers which are phase and time aligned. However, you might have different hearing, and others might have time/phase misaligned speakers where feedback can help. Try it out. Easy to put in or to remove the feedback loop. I've tried using a variable resistor, to see what varying levels of feedback do. You can go back and forth from very little feedback to generous feedback. Go with a pot 20 times the size of your feedback resistor, and place the pot in series with the resistor: increasing resistance you will lower feedback. That was a real eye opener for me, listening to feedback change in real time. About the shrillness: I never experienced it, so no idea what people are talking about. Although I always use grid stoppers and what I found very important: cathode stoppers. Never ever leave your 807 cathodes unattended!
G2 stopper is also critical, but when you run VR regulation on G2, then I found no stopper is necessary, just run the VR tube directly to g2. Plate stopper also recommended, but works for me without just the same. For driver stage: carbon copy the Yamamoto A08 driver stage (717A, zener regulated screen), and start with that. Spice it with a beefy CLCLC power supply. That's where I started, and it's the beginning of a wonderful road.
Have fun, Janos
Janos - your results are quite exciting! Who would think good ol' 807 is capable of such tricks...
Low distortion of 807 at high plate voltages can be deduced from data sheet distortion curves. In fixed bias AB1, the third harmonic at maximum power decreases as plate voltage and power output increase:
Ua=500 V, Pout=46 W, D3=2.6%
Ua=600 V, Pout=56 W, D3=2.2%
Extrapolating, it could be something like this:
Ua=800 V, Pout=75 W, D3≈1.5%? Or, at half-power, 37.5 W, D3=0.75%! And this is without any negative feedback!
For comparison, 807 triode-connected:
Ua=400 V, Pout=15 W, D3=3.1%.
My experience with global negative feedback is similar to yours.
I also believe (no definitive proof) that having good cathode current reserve is beneficial for sound. In your case, only 25% of cathode current capability is used.
The last but not least, running 807 at your operating point is conservative with regard to tube reliability and longevity.
The only shortcoming I see is a requirement for custom output transformer.
Low distortion of 807 at high plate voltages can be deduced from data sheet distortion curves. In fixed bias AB1, the third harmonic at maximum power decreases as plate voltage and power output increase:
Ua=500 V, Pout=46 W, D3=2.6%
Ua=600 V, Pout=56 W, D3=2.2%
Extrapolating, it could be something like this:
Ua=800 V, Pout=75 W, D3≈1.5%? Or, at half-power, 37.5 W, D3=0.75%! And this is without any negative feedback!
For comparison, 807 triode-connected:
Ua=400 V, Pout=15 W, D3=3.1%.
My experience with global negative feedback is similar to yours.
I also believe (no definitive proof) that having good cathode current reserve is beneficial for sound. In your case, only 25% of cathode current capability is used.
The last but not least, running 807 at your operating point is conservative with regard to tube reliability and longevity.
The only shortcoming I see is a requirement for custom output transformer.
807 current requirements
Ryssen - Actually, already built in an alternative high voltage supply that will be about 430V for 1. tube-saver mode for background listening 2. for operation by persons other than me and 3. to be able to listen to the amp if I'm taking the 700V supply off for tweaking. (The 700V supply is on a separate chassis.)
There is also a #4 reason, a new idea, which was the main reason: to use the 430V for warm-up and cool down. The 430V is fed as if it was a filter section preceeding the 700V filter sections, and placed two 1kV diodes in series to brace for the 700V - so 430V can feed forward, but 700V cannot leak backwards and fry the first filter cap plus burn out the lower voltage power transformer. This has not been tested yet, and I'm not sure whether it will work.... will report on this later when I have my test bench up and running again. For now, workbench is off line (extensive home remodeling).
sser2 - Thanks for the data sheet distortion figures! Indeed, I also suspect that distortion keeps dropping as high voltage is increased beyond the datasheet specs. Higher voltage accelerates electrons much faster, keeping them on course much tighter - that's how I imagine it.
My suspicion is that the current reserve is indeed paramount for dynamics. My last mod on the 807 SE amp was to turn it to PSE (to lower the demands on the OPT primary inductance). Indeed, that aspect worked out: I have more copious lower end but it made a double current demand on the power supply. Even though I have a monster of a power transformer feeding the 700V (520VAC at 700mA - so with PSE I'm still using only a fraction of the iron), but the explosive dynamics is gone. The amp still does not run out of reserves when pushed hard, but the unrestrained dynamics is no more.
Also noticed that when you have a normal size power supply, even a small bleeder resistor robs dynamic reserve and power. Because of that, I installed switchable bleeders on the caps that see 700V. Incidentally, there's no sonic difference whatsoever if the bleeders are on or off - the gigantic source makes such a tiny difference not noticeable. But doubling the load was a major difference.
Indeed, there's no easy solution. I'll probably have to go back to SE: sounded much better than PSE. Another alternative is that the screens could be lacking juice since 4 screens are very close to the limits of a single VR source (up to about 36mA), while 2 screens are well within the bounds. I have not been that much suspicious of the screen yet, as VR cutoff sounds very nasty (similar to extreme motor boating), and there's not even a hint of that. However, it's known that VR regulation gets worse as the current demand increases.
I'll have the summer to figure out whether the limiter in PSE is the high voltage supply or the screen supply (or maybe OPT core saturation? 125ESE should handle 80mA, and gets 52mA, versus the 26mA of SE).
But, just as you said, a custom OPT is not easy to come by. I estimate that a 11-12K pri iron would work, I'll get Electraprint to wind one. I have my eyes on the partial silver secondary version, and I'll go for oversized core size. Just to make sure it has plenty of reserves. Will have to save for that one though, so that's not happening this summer yet.
Ryssen - Actually, already built in an alternative high voltage supply that will be about 430V for 1. tube-saver mode for background listening 2. for operation by persons other than me and 3. to be able to listen to the amp if I'm taking the 700V supply off for tweaking. (The 700V supply is on a separate chassis.)
There is also a #4 reason, a new idea, which was the main reason: to use the 430V for warm-up and cool down. The 430V is fed as if it was a filter section preceeding the 700V filter sections, and placed two 1kV diodes in series to brace for the 700V - so 430V can feed forward, but 700V cannot leak backwards and fry the first filter cap plus burn out the lower voltage power transformer. This has not been tested yet, and I'm not sure whether it will work.... will report on this later when I have my test bench up and running again. For now, workbench is off line (extensive home remodeling).
sser2 - Thanks for the data sheet distortion figures! Indeed, I also suspect that distortion keeps dropping as high voltage is increased beyond the datasheet specs. Higher voltage accelerates electrons much faster, keeping them on course much tighter - that's how I imagine it.
My suspicion is that the current reserve is indeed paramount for dynamics. My last mod on the 807 SE amp was to turn it to PSE (to lower the demands on the OPT primary inductance). Indeed, that aspect worked out: I have more copious lower end but it made a double current demand on the power supply. Even though I have a monster of a power transformer feeding the 700V (520VAC at 700mA - so with PSE I'm still using only a fraction of the iron), but the explosive dynamics is gone. The amp still does not run out of reserves when pushed hard, but the unrestrained dynamics is no more.
Also noticed that when you have a normal size power supply, even a small bleeder resistor robs dynamic reserve and power. Because of that, I installed switchable bleeders on the caps that see 700V. Incidentally, there's no sonic difference whatsoever if the bleeders are on or off - the gigantic source makes such a tiny difference not noticeable. But doubling the load was a major difference.
Indeed, there's no easy solution. I'll probably have to go back to SE: sounded much better than PSE. Another alternative is that the screens could be lacking juice since 4 screens are very close to the limits of a single VR source (up to about 36mA), while 2 screens are well within the bounds. I have not been that much suspicious of the screen yet, as VR cutoff sounds very nasty (similar to extreme motor boating), and there's not even a hint of that. However, it's known that VR regulation gets worse as the current demand increases.
I'll have the summer to figure out whether the limiter in PSE is the high voltage supply or the screen supply (or maybe OPT core saturation? 125ESE should handle 80mA, and gets 52mA, versus the 26mA of SE).
But, just as you said, a custom OPT is not easy to come by. I estimate that a 11-12K pri iron would work, I'll get Electraprint to wind one. I have my eyes on the partial silver secondary version, and I'll go for oversized core size. Just to make sure it has plenty of reserves. Will have to save for that one though, so that's not happening this summer yet.
I´m planning on a 807SE for 400v,I once built a GM70 amp that run of 900v
I was allmost scared...so I´ll stay safe with 400v.
My experience with 807 around 400V: The screen supply counts the most. I got best sound with VR tube regulated supplies. I use 2 OD3 tubes in series to get about 310VDC, and run the 807 self-bias mode. That is, the screens will be VR regulated voltage minus bias voltage, around 285V. Tried lower screen voltages as well, both VR regulated (OC3-OD3 series gives around 225V), zener regulated (anemic compared to VR), and plain RC filtered (mediocre sound). Also, when you look at the premium theater amps of the 50s, they ran the 807 screens with VR regulation. (Such as the fabled IPC amp.)
Higher screen voltage gives more lively, dynamic sound. Also, the higher the screen U, the lower the pri impedance you need for the OPT, and lower the Rp. Give it a try with 36mA current and 10K load, with a variable cathode resistor - that way, you can set current for optimal sound.
Also, I got as advice from my 807 expert friend to keep the 807 screen a couple volts under 300V for best sound, and not any above - that's asking for short life and trouble. This agrees with my findings as well.
You can try triode mode as well. I'd tie the screen trough a 2K to the plate to give some slack for the screen. 10K-12K load for OPT will work. That will sound very nice, and saves you the trouble of building a screen regulator.
And for safety - I'd only go with self bias. Cathode bypass caps is a must. 807 self biased, unbypassed is completely anemic, distorted and powerless. Fixed bias - I stay away from there, that's asking for long term stability issues. And to me, self bias sounds more natural.
Thanks for the catch, 5K it is, not 10! Playing around with screen voltage, cathode current, tetrode - triode mode, optimum load will be between 3K-6K at 400V.
I'm glad you also have positive experience with 807. What input / phase splitter are you using?
Not yet. What sonic difference is between that and the regular global feedback? I suspect that as you can get by with using lower feedback resistor it should have a cleaner sound, but the phase effects at the frequency extremes will be just the same as regular global feedback at same level of feedback. The Quad II is a very neat amp, though. ; ) I'm currently rebuilding an Ampex 5616. Keeping it relatively close to stock, but sticking in a pair of OD3s instead of the big bleeder resistor for screen/driver regulation.
I built a 807 AB1 PP amplifier with centre tapped secondary OT, the cathodes cross-connected to the secondary. It gives 3-4 dB local NFB around the OT, this is why I am asking. B+ is 425V, g2 tied to anode with 1k resistor, no global NFB, fixed bias. I am thinking about ways of improvement, and cathode bias is one.
I think cathode bias will be an improvement. I have converted several Quicksilvers to cathode bias from fixed bias, and the result was much nicer tone. I religiously use cathode bias for everything I build. I think tubes like it better - and they can also adjust for changing line voltages.
De nada ; ) I lived in Brazil, have fond memories.
Interesting info on Zaerix! Did not know that they also branded British tubes.
My OPT has a Ultra liniear tap,could I try that to via a switch between "triod" and "UL"?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.