At lunch time today I was asked which valve would sound best at 600V - an 845 or triode mode 813?
I don't have experience with either valve (I have a pair of 845 but haven't done anything with them...). Looking at their curves I can't see that it makes a lot of difference... Has anyone any experience with these valave at an OP of, say, 600V and 100mA.
Actually I think at this OP the higher gain of the 813 means it probably wins but I could be completly wrong - again
Any help would be appreciated.
ciao
James
I don't have experience with either valve (I have a pair of 845 but haven't done anything with them...). Looking at their curves I can't see that it makes a lot of difference... Has anyone any experience with these valave at an OP of, say, 600V and 100mA.
Actually I think at this OP the higher gain of the 813 means it probably wins but I could be completly wrong - again
Any help would be appreciated.
ciao
James
That's an odd question, much akin to, "What's better for opening a can, a screwdriver or a pair of pliers?" Hey, how about a CAN OPENER?
These are fine tubes for HV/low current circuits. Why use an inappropriate component for a given set of conditions? If you want to use an 813, use a power supply with a comma in the output.
These are fine tubes for HV/low current circuits. Why use an inappropriate component for a given set of conditions? If you want to use an 813, use a power supply with a comma in the output.
Hi SY,
It is and it isn't an odd question...
It came about as we were discussing valve RIAA pre-amps and one of my engineers remarked that "using valves in an RIAA pre is odd - I can understand the choice in a power amp. but why for low level signals that need lots of equalisation?" A very fair question and one I was happy to take on (he's still not convinced...). But that lead onto the most outrageous valves used in an RIAA pre .. and hence to the 845 RIAA amp with the 845 run with a 600V B+. Then another collegue opinioned that the 813 in triode mode was more linear at 600V than an 845 so it would be a better choice. I think he was playing 'devils advocate' but I have no way of deciding one way or the other as the plate curves for both look very similar and I have no other data for an 813 in triode mode...
Curiosity drove the question - please don't kill the cat
BTW - what valve in triode mode would you recommend at 600V? I'm not sure what I would recommend at that voltage for a power amp or pre-amp. Maybe a Monkey style design would work best at that B+?
ciao
James.
It is and it isn't an odd question...
It came about as we were discussing valve RIAA pre-amps and one of my engineers remarked that "using valves in an RIAA pre is odd - I can understand the choice in a power amp. but why for low level signals that need lots of equalisation?" A very fair question and one I was happy to take on (he's still not convinced...). But that lead onto the most outrageous valves used in an RIAA pre .. and hence to the 845 RIAA amp with the 845 run with a 600V B+. Then another collegue opinioned that the 813 in triode mode was more linear at 600V than an 845 so it would be a better choice. I think he was playing 'devils advocate' but I have no way of deciding one way or the other as the plate curves for both look very similar and I have no other data for an 813 in triode mode...
Curiosity drove the question - please don't kill the cat
BTW - what valve in triode mode would you recommend at 600V? I'm not sure what I would recommend at that voltage for a power amp or pre-amp. Maybe a Monkey style design would work best at that B+?
ciao
James.
In my own triode amp, I'm using 6LF6s driven in enhancement mode (i.e., grounded grid, input to screen). But I've had good luck with 6550As triode strapped and EL34s ditto at similar plate voltages. There are no doubt a lot of other choices. I don't know what the max plate voltage rating of 300B is...
FWIW, I use tubes in my RIAA, but nothing commonly available. In such circuits, passive eq is the way to go, IME.
FWIW, I use tubes in my RIAA, but nothing commonly available. In such circuits, passive eq is the way to go, IME.
6LF6s looks interesting. I'd forgotten all my 509 and 519 experience too.
600volts is beyond the manufacturers recommended Va for triode mode 6550 or El34s so I've never run them that hot. 300B normal max is 450V so it makes a good Monkey at 600V or so.
I agree with the passive eq. on RIAA. At present mine is still transistor - though it does run off +/-60V rails to keep things nice and linear
I'm trying to source some 6F12s to build a replacement...ciao
James
James D. said:
Hi James, all
Did someone say 813 ?
I have tried the 845 and the 813 in the same amps , started with 845s and ended with
813s , but that is at high voltage, never tried either at low voltage, so can't really say which is "best" at 600V
To my ears most of the valves I have tried sounds best wide open, high voltage,
they seem to open up , but that is probably a taste thing.
But as a guess , at 600V , maybe the 845. Or maybe not.
Using transmitter valves in a RIAA pre sounds
, but why not go all the wayand use them at 1100 volts ?
813 as a Triode
I have been running a p-p 813 triode amplifier since 1997.
Admittedly I am using 830V (7k a-a). 813 is great value.
At 100mA ra is about 1.8k - i.e. the same as an 845, but becasue mu of 813 is 8, I need only about -75V bias which makes it much easier to drive.
The filament supply is a pain though: 10V @ 5A; I am therefore burning 100W just to light 'em up - and they do look lovely lit up.
I have attached triode curves for 813.
7N7
I have been running a p-p 813 triode amplifier since 1997.
Admittedly I am using 830V (7k a-a). 813 is great value.
At 100mA ra is about 1.8k - i.e. the same as an 845, but becasue mu of 813 is 8, I need only about -75V bias which makes it much easier to drive.
The filament supply is a pain though: 10V @ 5A; I am therefore burning 100W just to light 'em up - and they do look lovely lit up.
I have attached triode curves for 813.
7N7
Attachments
No doubt given the current 6LF6 pricing/availability, I'll be moving over to 509/519 when the time comes to retube.
IIRC, both the 6550 and the EL34 are rated to 600V plate (data sheets for 6CA7 indicate 800V max with no current for both plate and screen). The screen ratings are lower, but only for pentode and UL connections where plate and screen don't track.
IIRC, both the 6550 and the EL34 are rated to 600V plate (data sheets for 6CA7 indicate 800V max with no current for both plate and screen). The screen ratings are lower, but only for pentode and UL connections where plate and screen don't track.
Hi Slowmotion,
Thanks for your reply, I seem to remember that the 600V point was chosen as it worked better as a pre-amp at that point... I think I need to go re-read some of my bumf...
Hi 7N7,
Yeah the 813 looks very nice at lower voltages and it's extra gain against an 845 or 211 looks useful. Not to mention that it is cheaper too...
Hi SY,
Agree with the 600+V in pentode or tetrode mode but GE state max 500V in triode GE 6550A data
I can't find my EL34 data but I recall a similar limit in triode mode. Of course, being valves they would work well at the higher voltage but, I guess, have a shorter lifetime.
I have been doing some research on sg setups in triode mode and it seems that the audio pentode and beam tetrodes were engineered to allow sg to be connected to plate at very high voltages. This is not the case for most transmitter valves where much lower plate voltages should be used when triode connected i.e. they should be run close to the sg maximum not the plate maximum... I have not found any data on line output valves yet. I wonder if this info is accurate though...
ciao
James
Thanks for your reply, I seem to remember that the 600V point was chosen as it worked better as a pre-amp at that point... I think I need to go re-read some of my bumf...
Hi 7N7,
Yeah the 813 looks very nice at lower voltages and it's extra gain against an 845 or 211 looks useful. Not to mention that it is cheaper too...
Hi SY,
Agree with the 600+V in pentode or tetrode mode but GE state max 500V in triode GE 6550A data
I can't find my EL34 data but I recall a similar limit in triode mode. Of course, being valves they would work well at the higher voltage but, I guess, have a shorter lifetime.
I have been doing some research on sg setups in triode mode and it seems that the audio pentode and beam tetrodes were engineered to allow sg to be connected to plate at very high voltages. This is not the case for most transmitter valves where much lower plate voltages should be used when triode connected i.e. they should be run close to the sg maximum not the plate maximum... I have not found any data on line output valves yet. I wonder if this info is accurate though...
ciao
James
Triode-strapped transmitting valves
James,
Where does this information come from? The only maker's information that I have found is for 6146 (g2 max 250V) where Philips state that it can be run as a triode at 400V. 13E1 has a g2 limit of 300V and I have tested it flat out at 500V as a triode without problems.
Next week I have some transmitting valves coming. These are rated at 2kV anode and 400V g2. It is my intention to prepare triode curves for them since if they are linear as triodes I would like to run them at about 600V; this of course will require preparation of curves up to 1kV! Fireworks eh?
813 is more generous with its 1.1kV g2 limit; I expect that 1.5kV as a triode would be OK. I did consider this for a while, but these high voltages are no good for audio, since they mean enormous primary output loads, in the interests of getting a decent load line.
7N7
James D. said:
James,
Where does this information come from? The only maker's information that I have found is for 6146 (g2 max 250V) where Philips state that it can be run as a triode at 400V. 13E1 has a g2 limit of 300V and I have tested it flat out at 500V as a triode without problems.
Next week I have some transmitting valves coming. These are rated at 2kV anode and 400V g2. It is my intention to prepare triode curves for them since if they are linear as triodes I would like to run them at about 600V; this of course will require preparation of curves up to 1kV! Fireworks eh?
813 is more generous with its 1.1kV g2 limit; I expect that 1.5kV as a triode would be OK. I did consider this for a while, but these high voltages are no good for audio, since they mean enormous primary output loads, in the interests of getting a decent load line.
7N7
The information has come from various sources.
I recovered my notebooks from when I was taught about valves in the sixties and seventies by a group of experienced valve designers. I recovered some of my old text books at the same time.
On the web I started reading about valve rf amplifiers and tubes. I found two links particularly useful:
Screen Grids
and
Care and feeding.. .
The 813 looks more and more interesting... I must play with one soon...
ciao
Jmaes
I recovered my notebooks from when I was taught about valves in the sixties and seventies by a group of experienced valve designers. I recovered some of my old text books at the same time.
On the web I started reading about valve rf amplifiers and tubes. I found two links particularly useful:
Screen Grids
and
Care and feeding.. .
The 813 looks more and more interesting... I must play with one soon...
ciao
Jmaes
Afterthought
Another point about all this is that audio beam tetrodes and pentodes have a smaller ratio between maximum anode voltage and g2 voltage: the greatest I can think of off hand is Mullard's specification for the EL34 where 800V was quoted as Va max (Read: "OK boys, let's push that octal socket insulation to the limit) and 450V on g2.
6146's voltages were Va max 600V CCS and 750V ICAS, with g2 at 250V.
STC's 12E1 and 13E1 regulator beam tetrodes had 300V g2 limits and 800V anode limits (octal with top-cap and septar respectively), whilst nearly all of the more recent transmitting types set the g2 limit at 400V, with anode voltage limits generally starting at 2kV.
I suppose the ultimate issue with this is the risk of flash-over between g2 and g1 and between g2 and g3/beam plates since obviously the distances are much smaller than between the anode and the other electrodes.
I should be very interested to hear other views on this.
7N7
James D. said:
Another point about all this is that audio beam tetrodes and pentodes have a smaller ratio between maximum anode voltage and g2 voltage: the greatest I can think of off hand is Mullard's specification for the EL34 where 800V was quoted as Va max (Read: "OK boys, let's push that octal socket insulation to the limit) and 450V on g2.
6146's voltages were Va max 600V CCS and 750V ICAS, with g2 at 250V.
STC's 12E1 and 13E1 regulator beam tetrodes had 300V g2 limits and 800V anode limits (octal with top-cap and septar respectively), whilst nearly all of the more recent transmitting types set the g2 limit at 400V, with anode voltage limits generally starting at 2kV.
I suppose the ultimate issue with this is the risk of flash-over between g2 and g1 and between g2 and g3/beam plates since obviously the distances are much smaller than between the anode and the other electrodes.
I should be very interested to hear other views on this.
7N7
James,
I have taken a look at those papers; they are very interesting.
Most frustrating is the fact that I'll bet that all sorts of tests were done at the time and now the data have been lost - probably forever.
I once tried to contact RCA regarding experimental data for 813; what a waste of time! I received some rubbish about cheap and nasty-looking in-car music equipment. That is a terrible come-down for one of the great companies in electronics history. Friends who have visited the USA tell me that the RCA brand appears on all sorts of terrible rubbish.
Of course it would be unreasonable to expect the designers of VHF and UHF transmitting valves to have anticipated eccentrics triode-strapping their products.
It would be wonderful to find someone who worked on the design of these valves and pick his brains!
7N7
I have taken a look at those papers; they are very interesting.
Most frustrating is the fact that I'll bet that all sorts of tests were done at the time and now the data have been lost - probably forever.
I once tried to contact RCA regarding experimental data for 813; what a waste of time! I received some rubbish about cheap and nasty-looking in-car music equipment. That is a terrible come-down for one of the great companies in electronics history. Friends who have visited the USA tell me that the RCA brand appears on all sorts of terrible rubbish.
Of course it would be unreasonable to expect the designers of VHF and UHF transmitting valves to have anticipated eccentrics triode-strapping their products.
It would be wonderful to find someone who worked on the design of these valves and pick his brains!
7N7
6CA7 data sheets, in their full glory, are up at Triode Electronics's web site. They show the 800V anode voltage figure, but give no explicit max ratings in triode mode.
In any event, I've run both of these tube types (6CA7/EL34 and 6550) from different manufacturers at 550-600V with no reliability problems. It does help to keep the idle current low and the load impedance high- in my modded Fairchild 260s, the B+ runs about 550V, p-p load is 6600 ohms, and the EL34s idle at about 33-35 ma. I've been using these for a couple of decades, replacing the O/P tubes about every two years or so.
In any event, I've run both of these tube types (6CA7/EL34 and 6550) from different manufacturers at 550-600V with no reliability problems. It does help to keep the idle current low and the load impedance high- in my modded Fairchild 260s, the B+ runs about 550V, p-p load is 6600 ohms, and the EL34s idle at about 33-35 ma. I've been using these for a couple of decades, replacing the O/P tubes about every two years or so.
James D. said:
Simple. Dynamic range. A good valve, used optimally, produces only slightly more noise than the best transistors, but because its HT supply is >300V rather than 30V, it typically has >20dB more headroom compared to the transistor circuit. Although passive equalisation sounds best (IME), it is greedy for dynamic range. Ergo, the valve is the ideal device for the uncontrolled signals leaving LP. A good pre-amplifier displays its dynamic range by not being overloaded when the stylus collides with boulders, so it produces quieter clicks and bangs.
Hi 7N7,
I'm still working on the physics of beam tetrodes and pentodes to try and understand how to get the best out of them. Triodes are easy - in comparision! I can't help feeling that we have lost a wealth of knowledge over the last 60 years Although that could just be a reflection of my own ignorance 
The RF boys have there own stash of knowledge that is very interesting too.
Hi SY,
Thanks for pointer. Your amp sounds interesting, is the design on the web?
Hi EC8010,
That quote was from one of my engineers at work
I hasten to add...
I agree about the dynamic range aspect. My present mc RIAA pre is ss and uses +/- 60V rails - it has a very relaxed low distortion manner so I've not felt any need to move to a valve pre yet. Maybe I'll use my pair of 845 for that
I noticed that Sakuma used 845 at B+s from 350V to 1100V...
ciao
James
I'm still working on the physics of beam tetrodes and pentodes to try and understand how to get the best out of them. Triodes are easy - in comparision! I can't help feeling that we have lost a wealth of knowledge over the last 60 years
Although that could just be a reflection of my own ignorance The RF boys have there own stash of knowledge that is very interesting too.Hi SY,
Thanks for pointer. Your amp sounds interesting, is the design on the web?
Hi EC8010,
That quote was from one of my engineers at work
I hasten to add...I agree about the dynamic range aspect. My present mc RIAA pre is ss and uses +/- 60V rails - it has a very relaxed low distortion manner so I've not felt any need to move to a valve pre yet. Maybe I'll use my pair of 845 for that
I noticed that Sakuma used 845 at B+s from 350V to 1100V...
ciao
James
No, not yet, though the input stage topology is posted (as a drawing test) in the math thread. I'm trying to learn how to get a schematic drawn and into a postable format and work out some parts subs to replace stuff (like the tubes and FETs that I used) which is pretty much unavailable these days.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.