Hi Frank!
After my big 6C33CB-Circlotron, I am starting now with a 833A-amp. SE-design, class A2, driven (just as my 6C33C-tubes) by a mu-follower, this time only much more powerful. I am planning to use a PL 519 as a driver tube, it can provide easily much more than half an Ampere of current for the grid. And pentode operation assures me lowest possible output impedance. Plate voltage of the 833A should be around 1 - 1.2 kV - everything still higher gets difficult to manage for an amateur. Heater supply (10 V, 10 A) DC stabilised, I've read somewhere (and would guess from my experience) that AC heating inevitably causes hum. Plate circuit will be Parafeed. Plate voltage stabilised also by a (bigger than usual...) standard three-tube stabiliser (glow tube reference (OB2, 108 V), amplifier tube (PL 81), regulator tube (EL 519)). Such a stabilisation also is the easiest way to get the hum out - if you don't want to use lots of heavyweight and expensive chokes. I don't like much of iron in an amp, and be it only for the reasons of weight.... For this reason, I also would like to do without an interstage xfmr.
If you are interested, I would keep you informed about the evolution of the project.
For your 304 TL tubes, these are still more powerful (if I'm right) and still look more impressive
than the 833A. But I think, just for evaluation, the data for the 833A could be taken as minimum requirements for such a project. And - since data for such an operation of these tubes are hardly to obtain, building a test circuit should be the best way to see, if it works satisfactorily or not. The crucial point, I think is, if the tube still can work with comparatively low voltages in the 1 kV-range.... Everything higher will get a REAL challenge and also will make your budget explode.....
Good luck....
Uli
Here's a pic of my 6C33CB-Circlotron....
After my big 6C33CB-Circlotron, I am starting now with a 833A-amp. SE-design, class A2, driven (just as my 6C33C-tubes) by a mu-follower, this time only much more powerful. I am planning to use a PL 519 as a driver tube, it can provide easily much more than half an Ampere of current for the grid. And pentode operation assures me lowest possible output impedance. Plate voltage of the 833A should be around 1 - 1.2 kV - everything still higher gets difficult to manage for an amateur. Heater supply (10 V, 10 A) DC stabilised, I've read somewhere (and would guess from my experience) that AC heating inevitably causes hum. Plate circuit will be Parafeed. Plate voltage stabilised also by a (bigger than usual...) standard three-tube stabiliser (glow tube reference (OB2, 108 V), amplifier tube (PL 81), regulator tube (EL 519)). Such a stabilisation also is the easiest way to get the hum out - if you don't want to use lots of heavyweight and expensive chokes. I don't like much of iron in an amp, and be it only for the reasons of weight.... For this reason, I also would like to do without an interstage xfmr.
If you are interested, I would keep you informed about the evolution of the project.
For your 304 TL tubes, these are still more powerful (if I'm right) and still look more impressive
than the 833A. But I think, just for evaluation, the data for the 833A could be taken as minimum requirements for such a project. And - since data for such an operation of these tubes are hardly to obtain, building a test circuit should be the best way to see, if it works satisfactorily or not. The crucial point, I think is, if the tube still can work with comparatively low voltages in the 1 kV-range.... Everything higher will get a REAL challenge and also will make your budget explode..... Good luck....
Uli
Here's a pic of my 6C33CB-Circlotron....
I had the pleasure of listening to a pair of 304TL SE amps and was impressed. They are definitely suitable for SE use if you can manage the tremendous power supply requirements. One very nice thing about them is that they have a high enough mu that you can pretty easily build a two stage amp. The one I listened to (Jeffrey Jackson's) ran 1500V on the plates, and needed (IIRC) about -15 or -20V bias to get the plates a nice cherry color.
There are two challenges:
(1) lighting the filaments is not trivial. I don't recall exactly, but for some reason 5V@20A comes to mind. So, for DC you would be looking for transformers something like 10V@30A. You'll need two. And some good filter chokes as well. And a good fraction of a Farad of capacitance. Good luck with that.
(2) SE output transformers for this beast are not trivial. The good news is that you do not need a very high load as with some transmitting triodes. People are running 20k on 75TL's. The 304TL is basically four 75TL's in parallel, so a 5k load is fine. The hard part is handling all of the DC current (200mA? I don't remember what Jeffrey was running.) That and 50W capability means it will be large.
Jeffrey uses transformers with an extra 'cancel' winding. The DC filament current is run through the extra winding to cancel some or all of the DC flux from SE operation. Parafeed is another option and might work well. Of course you would still need mammoth plate chokes or CCS's.
Don't need much of a schematic if you're imaginative. Assuming you build all the power supplies you just need a fairly meaty driver; there is a LOT of miller capacitance. But, as I said, it doesn't need to deliver a whole lot of volts, so it's do-able. Any big transmitting tube wants a low DC resistance at its grid, so either interstage coupling or direct coupling with a choke loaded driver is pretty much mandatory.
-- Dave
There are two challenges:
(1) lighting the filaments is not trivial. I don't recall exactly, but for some reason 5V@20A comes to mind. So, for DC you would be looking for transformers something like 10V@30A. You'll need two. And some good filter chokes as well. And a good fraction of a Farad of capacitance. Good luck with that.
(2) SE output transformers for this beast are not trivial. The good news is that you do not need a very high load as with some transmitting triodes. People are running 20k on 75TL's. The 304TL is basically four 75TL's in parallel, so a 5k load is fine. The hard part is handling all of the DC current (200mA? I don't remember what Jeffrey was running.) That and 50W capability means it will be large.
Jeffrey uses transformers with an extra 'cancel' winding. The DC filament current is run through the extra winding to cancel some or all of the DC flux from SE operation. Parafeed is another option and might work well. Of course you would still need mammoth plate chokes or CCS's.
Don't need much of a schematic if you're imaginative. Assuming you build all the power supplies you just need a fairly meaty driver; there is a LOT of miller capacitance. But, as I said, it doesn't need to deliver a whole lot of volts, so it's do-able. Any big transmitting tube wants a low DC resistance at its grid, so either interstage coupling or direct coupling with a choke loaded driver is pretty much mandatory.
-- Dave
Frank
You are heading do approximately the same road I am but I have changed to push pull. The one piece of advice that everyone gave me and I took them up on it is to join the yahoo group http://groups.yahoo.com/group/GM70/ . They are a yahoo group dedicated to the design and building of tube audio amp using large transmitter tubes.
And the other piece of advice I what to give you is if you are convinced on devoting the large quantity of time and money to this amp then dont let anyone change your mind so stick to it and good luck.
Nick
You are heading do approximately the same road I am but I have changed to push pull. The one piece of advice that everyone gave me and I took them up on it is to join the yahoo group http://groups.yahoo.com/group/GM70/ . They are a yahoo group dedicated to the design and building of tube audio amp using large transmitter tubes.
And the other piece of advice I what to give you is if you are convinced on devoting the large quantity of time and money to this amp then dont let anyone change your mind so stick to it and good luck.
Nick
Nick: yes, I did join the boatanchor group. My problem is that it has been too many years since my technical education, and I don't know the theory part. I do have fairly good construction skills.
Question: is there a simple way to find the "operating point" for a tube, in my case , the 304 TL? I have been told that this is important, but I do not understand what this means. For example, I have heard people say that they use a plate voltage of 2000 vdc and the output transformer with a ratio of 5000 ohms on the tube side and 4 or 8 ohms on the speaker side. If the plate voltage is selected and the output transformer is known, what other variables remain? Does it have anything to do with the tube bias setting? I thought that class A uses zero bias?
Thanks in advance, Frank
Question: is there a simple way to find the "operating point" for a tube, in my case , the 304 TL? I have been told that this is important, but I do not understand what this means. For example, I have heard people say that they use a plate voltage of 2000 vdc and the output transformer with a ratio of 5000 ohms on the tube side and 4 or 8 ohms on the speaker side. If the plate voltage is selected and the output transformer is known, what other variables remain? Does it have anything to do with the tube bias setting? I thought that class A uses zero bias?
Thanks in advance, Frank
Well I'm assuming by the way you are talking is that you dont have the spec sheet for the tube so here you go.
http://www.vacuumresearchcorp.com/pdfs/valves/gatevalv.pdf?link=valves/gatevalv.pdf
But unfortunately I am kind of in the same boat as you that is I know rf amps but not audio and the suggestion that I was given by alot of the people on this forum is to buy Morgan Jones book (valve amplifiers) it's a 600+ page bible on tube audio.
I have mine coming in tomorrow
As for your bias point there isn't one for class A that I see on the spec sheet so you would probably have to play around with the bias at lowered plate voltage to find the right voltage so at max signal input you don't draw to much current through the plate.
As I have learned about tube audio it's not a easy undertaking to build a amp thats never been built before It's a lot of trial and error oh and time.
Hope this help a little
Nick
http://www.vacuumresearchcorp.com/pdfs/valves/gatevalv.pdf?link=valves/gatevalv.pdf
But unfortunately I am kind of in the same boat as you that is I know rf amps but not audio and the suggestion that I was given by alot of the people on this forum is to buy Morgan Jones book (valve amplifiers) it's a 600+ page bible on tube audio.
I have mine coming in tomorrow
As for your bias point there isn't one for class A that I see on the spec sheet so you would probably have to play around with the bias at lowered plate voltage to find the right voltage so at max signal input you don't draw to much current through the plate.
As I have learned about tube audio it's not a easy undertaking to build a amp thats never been built before It's a lot of trial and error oh and time.
Hope this help a little
Nick
Oh shoot the link I posted for the spec sheet is wrong here is the right one.
http://eshop1.chem.buffalo.edu/images/304-TL/304TL.pdf
I honestly think the spec sheet should answer alot of you question and if there is something on it that you dont know just ask us on this forum and we will help you out.
The big problem is the tube isn't really made for class A if you have 4 it would be a whole lot easier to do push pull class ab1.
But dont think I'm trying to change your mind just telling you what I learned.
Nick
http://eshop1.chem.buffalo.edu/images/304-TL/304TL.pdf
I honestly think the spec sheet should answer alot of you question and if there is something on it that you dont know just ask us on this forum and we will help you out.
The big problem is the tube isn't really made for class A if you have 4 it would be a whole lot easier to do push pull class ab1.
But dont think I'm trying to change your mind just telling you what I learned.
Nick
angelfj said:
Usually it's done by either:
1) Obtain plate curves for the tube (showing plate current vs. plate voltage for a variety of different grid voltages.) Look at the curves and draw tentative load lines until you find something you like.
2) Pick something from the manufacturer's data sheets.
3) Pick something recommended by someone else.
As far as I know, plate curves are not available for these tubes, at least not ones that are suitable for class A operation.
The data sheets definitely don't offer any suggested operating points for this kind of application.
If you're lucky, you might find someone that can tell you how they ran theirs ...
But all is not lost. These tubes like (need) to be run with their plates cherry hot, so what you do is pick a plate voltage (1500, 2000, you might try running lower) then bias up the tubes until the plates glow visibly in a well lit room. You'll know it when you see it. Apply signal to the grid and enjoy!
Seriously, once you pick a load - say 5k in this case - there's only two things left to adjust: the plate voltage and the grid bias. Make both adjustable and you can decide for yourself what sounds best. You need them both to be adjustable anyway so that you can bring the amp up slowly. Variacs are your friends. Surround yourself with them.
-- Dave
First, please accept my apology for this long post and to those (I know you are out there, who see all of this as very elementary) please be patient with me.
I am trying to establish the operating point as suggested in previous posts. However, if I go to the tube characteristics for the 304TL, as published by Eimac, they are not the same as the typical representation for a triode. I found an excellent reference on the Internet, and this guy did a wonderful job of explaining, step by step, how to establish the operating point.
see: http://www.valveheart.com/
In order to follow the example, I seem to be missing some critical information, for the 304TL, which would normally be included in the characteristics for other tubes/manufacturers.
1. Rp = plate resistance
It was suggested that for triodes , the load resistance, Rl, also referred to as the output transformer primary impedance, can be calculated as 2 - 6 x's Rp. So, if I do not know what is the Rp, I can not calculate Rs, which is critical for calculation or a graphical check for the operating point.
2. The example uses a graphical representation of the tube operation. The horizontal axis represents plate voltage, and the vertical axis represents plate curent. Superimposed on this graph are curves for several values of grid bias. The Eimac data is very different. There are two types of curves. The first is a graph of power output (watts) horizontal axis VS grid driving power (watts)) vertical axis , for a fixed value of plate voltage (Eb). Each of these graphs show several curves for different values of Ec (cathode voltage???). The second type of graph, and there is only one, has plate voltage on the horizontal axis and grid volts on the vertical axis. Several curves are drawn for both different levels of plate current and grid current. So, it seems that these graphs are not suitable for the graphical determination of the operating point.
Can someone offer a little help? I have to believe that establishing the operating point fpr a 304TL, to be used in a SE, class A amp has been done before. In the spirit of not re-inventing the wheel, please induldge by curiosity. My only limitations are the voltage and capacity of my plate transformer, which will be limited to approx. 2000 vdc @ 500 ma.
Thanks in advance, Frank
I am trying to establish the operating point as suggested in previous posts. However, if I go to the tube characteristics for the 304TL, as published by Eimac, they are not the same as the typical representation for a triode. I found an excellent reference on the Internet, and this guy did a wonderful job of explaining, step by step, how to establish the operating point.
see: http://www.valveheart.com/
In order to follow the example, I seem to be missing some critical information, for the 304TL, which would normally be included in the characteristics for other tubes/manufacturers.
1. Rp = plate resistance
It was suggested that for triodes , the load resistance, Rl, also referred to as the output transformer primary impedance, can be calculated as 2 - 6 x's Rp. So, if I do not know what is the Rp, I can not calculate Rs, which is critical for calculation or a graphical check for the operating point.
2. The example uses a graphical representation of the tube operation. The horizontal axis represents plate voltage, and the vertical axis represents plate curent. Superimposed on this graph are curves for several values of grid bias. The Eimac data is very different. There are two types of curves. The first is a graph of power output (watts) horizontal axis VS grid driving power (watts)) vertical axis , for a fixed value of plate voltage (Eb). Each of these graphs show several curves for different values of Ec (cathode voltage???). The second type of graph, and there is only one, has plate voltage on the horizontal axis and grid volts on the vertical axis. Several curves are drawn for both different levels of plate current and grid current. So, it seems that these graphs are not suitable for the graphical determination of the operating point.
Can someone offer a little help? I have to believe that establishing the operating point fpr a 304TL, to be used in a SE, class A amp has been done before. In the spirit of not re-inventing the wheel, please induldge by curiosity. My only limitations are the voltage and capacity of my plate transformer, which will be limited to approx. 2000 vdc @ 500 ma.
Thanks in advance, Frank
To all:
Two questions regarding tube filament requirements:
1. which is better ac or dc?
2. how critical is the voltage?
For example the 304TL requires 10 volts @12.5 amps. I have located a transformer rated 12 volts @ 25 amps. $30 each , not center-tapped. Would 2 more volts kill the filaments? I was thinking of taking the transformers apart and removing turns, until I got 10 volts with a load to simulate the tube filaments. Or putting a variac in the primary. What do you think? What about not having a center tap. Is this a game stopper?
Thanks, Frank
Two questions regarding tube filament requirements:
1. which is better ac or dc?
2. how critical is the voltage?
For example the 304TL requires 10 volts @12.5 amps. I have located a transformer rated 12 volts @ 25 amps. $30 each , not center-tapped. Would 2 more volts kill the filaments? I was thinking of taking the transformers apart and removing turns, until I got 10 volts with a load to simulate the tube filaments. Or putting a variac in the primary. What do you think? What about not having a center tap. Is this a game stopper?
Thanks, Frank
Hello,
I would tentativly go DC, in order to avoid hum - and you have to use the right filament voltage - I would not even try 20% overvoltage short term. If I would have to light this beast, I would go SMPS right away. You could use a standard 9V supply and adjust to max. voltage. The only problem could be that hiccup supplies might not start on cold filaments, current limiting would be better. Or, use a 12V supply, adjust to min. voltage and drop some .5V or so in a resistor. That will work with hiccup overcurrent protection as well.
Marcus
I would tentativly go DC, in order to avoid hum - and you have to use the right filament voltage - I would not even try 20% overvoltage short term. If I would have to light this beast, I would go SMPS right away. You could use a standard 9V supply and adjust to max. voltage. The only problem could be that hiccup supplies might not start on cold filaments, current limiting would be better. Or, use a 12V supply, adjust to min. voltage and drop some .5V or so in a resistor. That will work with hiccup overcurrent protection as well.
Marcus
On most eimac tubes the filament voltage at beam on state is not to vary more then 5% also you should not apply full filament voltage at turn on.
There are maximum current limits on the filament and they can only be exceeded applying full voltage to a cold filament.
There is also a minimum heating time before plate voltage can be applied usually 3 to 5 minuates for transmitter tube.
Nick
There are maximum current limits on the filament and they can only be exceeded applying full voltage to a cold filament.
There is also a minimum heating time before plate voltage can be applied usually 3 to 5 minuates for transmitter tube.
Nick
I don't think you'll be happy with AC because of the very high mu.
Also, the 304TL is essentially four 75TL's in one bottle, all connected in parallel. Well, not everything is always in parallel. The 10V@12.5A connection puts one pair of filaments in series with the other pair. The consequence is that with DC two of the sections will be running with 5V more/less bias than the other two. Using the 5V@25A connection puts all of the filaments in parallel so that all four sections bias up the same.
-- Dave
Also, the 304TL is essentially four 75TL's in one bottle, all connected in parallel. Well, not everything is always in parallel. The 10V@12.5A connection puts one pair of filaments in series with the other pair. The consequence is that with DC two of the sections will be running with 5V more/less bias than the other two. Using the 5V@25A connection puts all of the filaments in parallel so that all four sections bias up the same.
-- Dave
It seems that most people are recommending DC filament supply only for SE. 304TL in class A. The reason they say is that ac filament supplies introduce too much noise. I am certainly no expert, but this seems logical.
Regarding independent filaments, I believe that these 5-volt filaments are electrically isolated. I had plan to connect them in series and us a 10 volt DC supply. This will be based on a salvaged welding transformer that can easily deliver 30 -40 amps at 12 volts. However, I don't believe that the tube cares how the cathode is heated.
Regarding independent filaments, I believe that these 5-volt filaments are electrically isolated. I had plan to connect them in series and us a 10 volt DC supply. This will be based on a salvaged welding transformer that can easily deliver 30 -40 amps at 12 volts. However, I don't believe that the tube cares how the cathode is heated.