I would really like to build a tube amplifer but really know nothing about them. The tube I have is a GE, and on the glass it has "251.6" and "GT". On the black part of it, it has faded nubmers that I think are "58.43" and "C88-3" Can I use this tube as an amplifier. I also have a 13CW4.
If they will work, How do I use them? Does anyobody have a schematic for it? Are they simple to build and operate? Do I have to worry about heat? What kind of power supply is needed (voltage, amps, AC or DC, single or split)? And, finally, speaker wires shorting?
Thanks if anybody can help at all, Mike
If they will work, How do I use them? Does anyobody have a schematic for it? Are they simple to build and operate? Do I have to worry about heat? What kind of power supply is needed (voltage, amps, AC or DC, single or split)? And, finally, speaker wires shorting?
Thanks if anybody can help at all, Mike
I can't help you with the tube identification
but the tube i should be able to help with the schematic
you check out my website www.geocities.com/nickchua81
to get a rough idea of a tube amp
than from there figure the schematic for the tube you have
i am speculating the 13cw4 is nuvistor. Is it very small compared to a 9pin. If so then you got yourself a voltage amp then. I think Joel made a nuvistor amp drving el84 i think .
but the tube i should be able to help with the schematic
you check out my website www.geocities.com/nickchua81
to get a rough idea of a tube amp
than from there figure the schematic for the tube you have
i am speculating the 13cw4 is nuvistor. Is it very small compared to a 9pin. If so then you got yourself a voltage amp then. I think Joel made a nuvistor amp drving el84 i think .
Some questions;
- Do you know the tubes are good, ie did they come from a knowledgable/reputable source?
- Do you want to simply experiment with some tubes, or would you looking to build an amplifier that you can use, long term, in your system soon?
- How much power do you need?
- Budget?
- What speakers are you / are you intending to, interface with?
In a tube amp, a large part of the budget can go towards the transformers, so if you want something to use and keep, then I wouldn't spend money to optimise around a 25L6. As an experimenter / learning amp with surplus iron it would be fine. But why not build something to use and enjoy? If you're starting from scratch and have to purchase everything anyway build the best you can within your budget; the learning experience will be the same or better.
My suggestion would be something like the Magnavox 6BQ5 PP amps with Hammond (or good second hand) output transformers. About 15W/ch and good sounding too. Not hard to build, complicated or expensive.
- Do you know the tubes are good, ie did they come from a knowledgable/reputable source?
- Do you want to simply experiment with some tubes, or would you looking to build an amplifier that you can use, long term, in your system soon?
- How much power do you need?
- Budget?
- What speakers are you / are you intending to, interface with?
In a tube amp, a large part of the budget can go towards the transformers, so if you want something to use and keep, then I wouldn't spend money to optimise around a 25L6. As an experimenter / learning amp with surplus iron it would be fine. But why not build something to use and enjoy? If you're starting from scratch and have to purchase everything anyway build the best you can within your budget; the learning experience will be the same or better.
My suggestion would be something like the Magnavox 6BQ5 PP amps with Hammond (or good second hand) output transformers. About 15W/ch and good sounding too. Not hard to build, complicated or expensive.
A person I know who worked for Motorola in their electronics area gave me all of his parts when he retired, and one of them is a metal plate with 2 potentiometers, 1 a switched one, a small transformer that is just labeled "TR-118" "798 724" It has some wires, caps and resistores connected to the tube socket in p2p. Then, there is a 150V capacitor that has many wires and has 40, 30 and 20mfd and the tube. Out of the holes on the top of the metal are 4 cut off wires. Also on the botttom there ius this grey metal thing that has a screw in the bottom, a round shaft up the middle and 6 square things that look like fins.
Also, after the fact that the tube is a 251L6 was pointed out, i looked at it again and realized that the thing i had thought was the . was really the bottom of the L.
I would like to use this tube, to save some money, and also, because I just want a simple to build first-time amplifier. I don't need anything fancy. I have been using gainclones for a while now, and wanted to try something different.
As I said, I don't know too much about tubes. What kind of impedences can they drive? Also, how much power is needed to operate it? I remember reading that tubes work best for bass and midrange. Does this mean that ijt can drive a subwoofer? Possibly better than a 3886 gainclone?
Thanks for the help, Mike
Also, after the fact that the tube is a 251L6 was pointed out, i looked at it again and realized that the thing i had thought was the . was really the bottom of the L.
I would like to use this tube, to save some money, and also, because I just want a simple to build first-time amplifier. I don't need anything fancy. I have been using gainclones for a while now, and wanted to try something different.
As I said, I don't know too much about tubes. What kind of impedences can they drive? Also, how much power is needed to operate it? I remember reading that tubes work best for bass and midrange. Does this mean that ijt can drive a subwoofer? Possibly better than a 3886 gainclone?
Thanks for the help, Mike
Here's my email response
I thought I would post this so others know my thoughts and what we have discussed-
Hi Mike,
Like Brett said, the 25L6 is not a tube that I would use to build an amp. And as Colt45 said, you are not going to get much power from just one of these....but that's not what it's all about anyway! You really don't see this tube used much at all. The parts you were given are likely not good anymore. Caps like the multisection one you have can dry out, rendering them useless for filtering. The metal thing with the fins is a selenium rectifier. They were used in many Dynaco amps and its the first part that most folks replace when restoring an amp. I have read that they can explode when they do go bad, just like caps will if you exceed the voltage rating. I'm betting the transformer is the output transformer and this amp was run straight from the line voltage. It's not a safe way to do things since you are not isolated from the current out of the wall socket. That was done many years ago to cut costs, the cost of the power transformer. That's why you see many tubes used in TVs and old console radios that have heater voltages that can be run in series from 115 volts line in. 25, 30, and 50 volt heaters were far more common back then. I don't think you could ever find insurance coverage if you marketed a product built like that these days.
Most tube amps that you see built these days operate on B+ voltages much higher than your gainclone, like 200-500 volts neighborhood. And in the tube world, power costs $$. But it isn't power that many of us are after. You will read about many loonies, myself included, that spend a great deal of time and money on flea power directly heated triode amps with outputs in the 1-8 watt range. If you want to keep costs down on your first project, go with something simple, like a linestage. It should be a great buffer for you gainclone amp. Maybe someone on the DIYAudio BBS can suggest a nice simple design that runs on a B+ of about 165v and low current. That way I could send you the parts you would need, sans volume pot, jacks and, chassis. It's a great way to get started without dropping much $$. I think I spent around $600 on my first tube amp project (a simple 2A3) and it sounded pretty poor in comparison to one of the entry level kits out there. It was an expensive way to find out that I needed to learn more before calling to order parts!
Let me know what you decide. We'll set you up with something to play with!
Tom
I thought I would post this so others know my thoughts and what we have discussed-
Hi Mike,
Like Brett said, the 25L6 is not a tube that I would use to build an amp. And as Colt45 said, you are not going to get much power from just one of these....but that's not what it's all about anyway! You really don't see this tube used much at all. The parts you were given are likely not good anymore. Caps like the multisection one you have can dry out, rendering them useless for filtering. The metal thing with the fins is a selenium rectifier. They were used in many Dynaco amps and its the first part that most folks replace when restoring an amp. I have read that they can explode when they do go bad, just like caps will if you exceed the voltage rating. I'm betting the transformer is the output transformer and this amp was run straight from the line voltage. It's not a safe way to do things since you are not isolated from the current out of the wall socket. That was done many years ago to cut costs, the cost of the power transformer. That's why you see many tubes used in TVs and old console radios that have heater voltages that can be run in series from 115 volts line in. 25, 30, and 50 volt heaters were far more common back then. I don't think you could ever find insurance coverage if you marketed a product built like that these days.
Most tube amps that you see built these days operate on B+ voltages much higher than your gainclone, like 200-500 volts neighborhood. And in the tube world, power costs $$. But it isn't power that many of us are after. You will read about many loonies, myself included, that spend a great deal of time and money on flea power directly heated triode amps with outputs in the 1-8 watt range. If you want to keep costs down on your first project, go with something simple, like a linestage. It should be a great buffer for you gainclone amp. Maybe someone on the DIYAudio BBS can suggest a nice simple design that runs on a B+ of about 165v and low current. That way I could send you the parts you would need, sans volume pot, jacks and, chassis. It's a great way to get started without dropping much $$. I think I spent around $600 on my first tube amp project (a simple 2A3) and it sounded pretty poor in comparison to one of the entry level kits out there. It was an expensive way to find out that I needed to learn more before calling to order parts!
Let me know what you decide. We'll set you up with something to play with!
Tom
> gave me all of his parts when he retired
And was glad to get rid of "that junk".
I don't mean to laugh at you. You want to learn, and that's all good.
But I will offer a similar story. Pretend you never had a car. A friend gives you a carb from a 1953 Nash, a wheel from a 1967 Fiat, a seat from a 1987 Chevy, and some other car parts you don't know what they are. You want to know if you can build a car, and you probably mean "a good car", not just "something that moves".
You have a real odd-lot of tube-radio parts, some fairly specialized (the 25L6 is specialized for really cheap radios), all in unknown condition (loose parts from a tinkerer's bins are very often dead or half-dead parts left behind from repairs). You are missing some important parts to do anything useful. While a gainclone chip is nearly a complete amplifier in one blob, a tube is very much a "part", and you need a lot more parts before it does you any good.
> What kind of power supply is needed (voltage, amps, AC or DC, single or split)?
It Varies. Speaking very generally, you need a low-volt high-amp supply, like 6V 2A, to warm-up the "heaters". A tube won't do anything until its heater is red-hot. But the heater is just heat to loosen some electrons. You need a high-volt low-current, like 250V 0.050A, DC supply to suck those electrons through a load and do something useful. Yes, this is a real shock hazard and yes, tube gear runs very hot compared to most transistor/chip stuff. And then, because tubes are high-volt low-current and most loads are low-volt high-current, you need an audio transformer to make them work together. This transformer is often as big and more expensive than the power transformer.
I could sketch-up a 25L6 amplifier to make a watt or so. But you need lots more parts. While some of the parts you have might work, the general impression I get is that these are cheap parts and why go to a lot of trouble to make an amp that will sound cheap and give "tube amps" a bad name? By the time you buy good parts like power and output transformer and socket(s), there is no point in saving that $0.50 25L6 when a $15 tube will not only work better but possibly be easier to build. (The 26L6 has an unusual heater voltage, and you will need at least one more tube, which probably won't be 25V heater, so already your heater supply is getting complicated.)
Here is a simple one-tube amplifier, intended for headphones but will make a good sound in a loudspeaker. You probably won't understand the whole essay at first, and I don't say this is any great design, but it was intended to be about as simple and inexpensive as a tube audio power amplifier can be. It will give you an overall picture of what a tube amp project looks like. You can find far better projects on the web. Most of them assume you have some idea what you are doing already; you will need to read a bunch of them to get your bearings.
> Does this mean that it can drive a subwoofer? Possibly better than a 3886 gainclone?
Nobody is born knowing this stuff.... but you do have some learning ahead of you.
Most of the things we call subwoofers need 100 watts or more to get high output. I happen to remember (took a while to learn all this) that the 25L6 was designed to make just 1 watt, enough for a kitchen radio. Colt45 can maybe get 4 watts out of one, but I hope he gets them cheap because I would expect them to die fast with that kind of abuse. (Yes, 25L6es are pretty cheap, a glut on the market, since many millions were used in the 1940s and 1950s for many kitchen radios.) You can get more than twice the power with two tubes, but to get subwoofer level power, you would need a whole bucket of 25L6 and it really makes more sense to start from a bigger tube. And any speaker amplifier needs a special output transformer, a rather large costly one for subwoofer duty.
> reading that tubes work best for bass and midrange.
??? You can read many things; not all of them are true. You can't even believe all that I tell you. Listen to everything, but only believe things that make sense all-together.
Tubes are good far past the audible range. If I had to, I could make a CB transmitter from your 25L6 (and a heck of a lot more parts), running at 27MHz (that's 27,000KHz). A lot depends on the output transformer. Maybe what they mean is: it is hard to make an output transformer for audio that is good both at deep-bass and at high-treble. It is a problem, but can be solved (money helps).
There is a $150 stereo tube-amp kit floating around. It isn't top-spec, but it will make a nice noise (and quite different from the gainclone). It would be a way to get started, without being stalled by design and part-selection hassles, and it does give great bang for the buck. See if you like working with and listening to tubes (allow for the fact that $140 does not buy great transformers or capacitors), then you can judge if you want to try something better (more expensive).
Oh: you want to put that 25L6 to "good use"? Poke the pins with an ohm meter: only two of them will show about 50 or 100 Ω conductivity. Wire those to a 24V 0.2A (AC or DC) power supply. The heater in the center of the tube will glow red. Looks cool in a dark room. You "should" use a socket, but if you didn't get one then I would just tack-solder wires to a 25L6. Tube are precious, you shouldn't blob-up the pins, but a junkbox 25L6 is about the least precious tube there is.
And was glad to get rid of "that junk".
I don't mean to laugh at you. You want to learn, and that's all good.
But I will offer a similar story. Pretend you never had a car. A friend gives you a carb from a 1953 Nash, a wheel from a 1967 Fiat, a seat from a 1987 Chevy, and some other car parts you don't know what they are. You want to know if you can build a car, and you probably mean "a good car", not just "something that moves".
You have a real odd-lot of tube-radio parts, some fairly specialized (the 25L6 is specialized for really cheap radios), all in unknown condition (loose parts from a tinkerer's bins are very often dead or half-dead parts left behind from repairs). You are missing some important parts to do anything useful. While a gainclone chip is nearly a complete amplifier in one blob, a tube is very much a "part", and you need a lot more parts before it does you any good.
> What kind of power supply is needed (voltage, amps, AC or DC, single or split)?
It Varies. Speaking very generally, you need a low-volt high-amp supply, like 6V 2A, to warm-up the "heaters". A tube won't do anything until its heater is red-hot. But the heater is just heat to loosen some electrons. You need a high-volt low-current, like 250V 0.050A, DC supply to suck those electrons through a load and do something useful. Yes, this is a real shock hazard and yes, tube gear runs very hot compared to most transistor/chip stuff. And then, because tubes are high-volt low-current and most loads are low-volt high-current, you need an audio transformer to make them work together. This transformer is often as big and more expensive than the power transformer.
I could sketch-up a 25L6 amplifier to make a watt or so. But you need lots more parts. While some of the parts you have might work, the general impression I get is that these are cheap parts and why go to a lot of trouble to make an amp that will sound cheap and give "tube amps" a bad name? By the time you buy good parts like power and output transformer and socket(s), there is no point in saving that $0.50 25L6 when a $15 tube will not only work better but possibly be easier to build. (The 26L6 has an unusual heater voltage, and you will need at least one more tube, which probably won't be 25V heater, so already your heater supply is getting complicated.)
Here is a simple one-tube amplifier, intended for headphones but will make a good sound in a loudspeaker. You probably won't understand the whole essay at first, and I don't say this is any great design, but it was intended to be about as simple and inexpensive as a tube audio power amplifier can be. It will give you an overall picture of what a tube amp project looks like. You can find far better projects on the web. Most of them assume you have some idea what you are doing already; you will need to read a bunch of them to get your bearings.
> Does this mean that it can drive a subwoofer? Possibly better than a 3886 gainclone?
Nobody is born knowing this stuff.... but you do have some learning ahead of you.
Most of the things we call subwoofers need 100 watts or more to get high output. I happen to remember (took a while to learn all this) that the 25L6 was designed to make just 1 watt, enough for a kitchen radio. Colt45 can maybe get 4 watts out of one, but I hope he gets them cheap because I would expect them to die fast with that kind of abuse. (Yes, 25L6es are pretty cheap, a glut on the market, since many millions were used in the 1940s and 1950s for many kitchen radios.) You can get more than twice the power with two tubes, but to get subwoofer level power, you would need a whole bucket of 25L6 and it really makes more sense to start from a bigger tube. And any speaker amplifier needs a special output transformer, a rather large costly one for subwoofer duty.
> reading that tubes work best for bass and midrange.
??? You can read many things; not all of them are true. You can't even believe all that I tell you. Listen to everything, but only believe things that make sense all-together.
Tubes are good far past the audible range. If I had to, I could make a CB transmitter from your 25L6 (and a heck of a lot more parts), running at 27MHz (that's 27,000KHz). A lot depends on the output transformer. Maybe what they mean is: it is hard to make an output transformer for audio that is good both at deep-bass and at high-treble. It is a problem, but can be solved (money helps).
There is a $150 stereo tube-amp kit floating around. It isn't top-spec, but it will make a nice noise (and quite different from the gainclone). It would be a way to get started, without being stalled by design and part-selection hassles, and it does give great bang for the buck. See if you like working with and listening to tubes (allow for the fact that $140 does not buy great transformers or capacitors), then you can judge if you want to try something better (more expensive).
Oh: you want to put that 25L6 to "good use"? Poke the pins with an ohm meter: only two of them will show about 50 or 100 Ω conductivity. Wire those to a 24V 0.2A (AC or DC) power supply. The heater in the center of the tube will glow red. Looks cool in a dark room. You "should" use a socket, but if you didn't get one then I would just tack-solder wires to a 25L6. Tube are precious, you shouldn't blob-up the pins, but a junkbox 25L6 is about the least precious tube there is.
Uh, I forgot my memory is failing with age....
To light-up the heater in a 25L6 you need 0.3 Amps at 24-25 Volts, not what I said.
And 25L6 can do 2 watts in a cheap radio, and is rated 3.8 watts (at a higher supply voltage than any cheap radio used). Clearly Colt45 can be getting 4-5 watts without gross abuse that would reduce tube life enough to care about.
We almost always saw them working more like 1 watt, because a few pennies saved in radio parts was worth more than another watt of power in the kitchen. But now that I check specs (as I should have done), I recall the 25L6 and its twin 50L6 working at several watts in living-room radios, and in pairs making even 7-10 watts in small movie projectors.
And for lurkers: no, a 25L6, 35L6, or 50L6 isn't a variant 6L6. They probably did pick "L" to resonate with the aura of the mighty 6L6, and because they do come out of the same research, but the 6L6 is 360V 19W, while the radio xxL6 tubes were 200V 10W.
To light-up the heater in a 25L6 you need 0.3 Amps at 24-25 Volts, not what I said.
And 25L6 can do 2 watts in a cheap radio, and is rated 3.8 watts (at a higher supply voltage than any cheap radio used). Clearly Colt45 can be getting 4-5 watts without gross abuse that would reduce tube life enough to care about.
We almost always saw them working more like 1 watt, because a few pennies saved in radio parts was worth more than another watt of power in the kitchen. But now that I check specs (as I should have done), I recall the 25L6 and its twin 50L6 working at several watts in living-room radios, and in pairs making even 7-10 watts in small movie projectors.
And for lurkers: no, a 25L6, 35L6, or 50L6 isn't a variant 6L6. They probably did pick "L" to resonate with the aura of the mighty 6L6, and because they do come out of the same research, but the 6L6 is 360V 19W, while the radio xxL6 tubes were 200V 10W.
Very well put!
Thanks PRR,
You filled in the many holes in my reply. I hope he has not given up hope about diy tube endeavors. I hope he understands that it's not about wattage, it's about music....and in my case, learning, tinkering, and being creative.
Tom
The more I learn, the more questions I have.
Thanks PRR,
You filled in the many holes in my reply. I hope he has not given up hope about diy tube endeavors. I hope he understands that it's not about wattage, it's about music....and in my case, learning, tinkering, and being creative.
Tom
The more I learn, the more questions I have.
also for the record, 12L6/25L6/50L6 are equivalents to 6W6/12W6 (take a look at the curves and reccomended op. points if you want to see what i mean).
35L6 is a slightly different animal, not quite the same as the others (due to lower perverance[sp?] from less total power on the cathode i guess? no idea - i dont remember 35L6 specs, just the fact that they arent the same)
6GC5 is also a 6W6/25L6 sub, but its a neonoval (T9 glass with 9pin mini base). slightly different rating though, but still same curves and reccomended operating points
Ask Tim about running 6W6's over their ratings
35L6 is a slightly different animal, not quite the same as the others (due to lower perverance[sp?] from less total power on the cathode i guess? no idea - i dont remember 35L6 specs, just the fact that they arent the same)
6GC5 is also a 6W6/25L6 sub, but its a neonoval (T9 glass with 9pin mini base). slightly different rating though, but still same curves and reccomended operating points
Ask Tim about running 6W6's over their ratings
> 12L6/25L6/50L6 are equivalents to 6W6/12W6
6W6GT is rated higher plate voltage than 25L6, but I do see what you mean. They may be the same stampings, but they didn't make the 6W6GT with the low-volt plate metal they used on many 25L6 to save a penny. (It may be true that a lot of late-production 25L6 did get the "better" plate stuff. As demand died down, it was simpler to have just a few grades of plate metal.)
The 6W6GT has a good set of triode curves. Peering into the bottom left corner, I get this:
That is: since we need 25V for the heater, try 25V on the Plate. It won't make "Power", but it might make a workable "Voltage Amplifier". You could put this between a phono preamp - volume control, and a power amp (even a Gainclone), to get a little more gain on under-recorded sources. And to get a slightly "bent" "tube" sound.
6W6 (or 25L6) strapped Triode
24V B+
5K plate resistor
3mA bias
about 800 ohm cathode resistor
output impedance about 800 ohms
voltage gain about 5
max output about 17V p-p, 6VRMS
Estimated 2%-3% THD at 1VRMS
6W6GT is rated higher plate voltage than 25L6, but I do see what you mean. They may be the same stampings, but they didn't make the 6W6GT with the low-volt plate metal they used on many 25L6 to save a penny. (It may be true that a lot of late-production 25L6 did get the "better" plate stuff. As demand died down, it was simpler to have just a few grades of plate metal.)
The 6W6GT has a good set of triode curves. Peering into the bottom left corner, I get this:
An externally hosted image should be here but it was not working when we last tested it.
That is: since we need 25V for the heater, try 25V on the Plate. It won't make "Power", but it might make a workable "Voltage Amplifier". You could put this between a phono preamp - volume control, and a power amp (even a Gainclone), to get a little more gain on under-recorded sources. And to get a slightly "bent" "tube" sound.
6W6 (or 25L6) strapped Triode
24V B+
5K plate resistor
3mA bias
about 800 ohm cathode resistor
output impedance about 800 ohms
voltage gain about 5
max output about 17V p-p, 6VRMS
Estimated 2%-3% THD at 1VRMS
25L6 as a line-amp: this is one crazy idea that just may work.
The idea is to run both heater and plate on one 24V DC supply, something you can find as a surplus wall-wart or excess industrial part. That greatly simplifies the power supply, and keeps all voltages down to "safe" levels so you can be relatively fearless working around it. (Yeah, 25L6 wants 25V on the filament, but it will be fine with 24V which is a standard supply voltage.)
My previous estimate, based on a squint at a corner of triode-6W6 curves, may be optimistic. Using a SPICE model for the 6L6 strapped triode, I get gain about 4 and output Z about 1.2KΩ. Oh, and of course my earlier estimate of plate current was just wrong: it is roughly 1mA.
My guess is that the 26L6 will have slightly lower voltage gain and output Z, but not much lower.
Using the 50L6 and a 48V DC supply would be "better" for gain, output, and distortion.
It isn't the world's beefiest line-amp, but it should be more than enough to interface hi-fi CD and tape decks to power amplifiers, even if the power amp is a little low on gain or if you want to boost a quiet passage to high levels. And of course the input impedance can be as high as you want. The book suggests 500K max grid resistance, but at these ultra-low voltages it is in no danger with a 5 MegΩ resistor, it will just want a little re-biasing to trim-out grid offset.
And it looks like the old rule of thumb, "peak output at 5% THD is about 20% of supply voltage", is roughly true for large triodes (or triode-strapped pentodes) working at low voltages. I get 5% THD at about 4.8 volts, without trying too hard to optimize bias and output. At 1.9V Peak, a typical power amp input, THD may be well below 1%, maybe 0.5%. At even lower levels I got SPICE-answers of 0.05%, though I don't trust numbers like that from SPICE, especially on tube models.
I did try various bias conditions. Because it is not so far below maximum output, a small shift of bias will cause a significant shift in the balance between 2nd and 3rd harmonic distortions. Mostly the 2nd harmonic can be trimmed from 0.15% to 1.5%, without much change of gain or 3rd harmonic.
_HARMONIC__FREQ____FOURIER_____NORMALIZED______PHASE(DEG)
_____1_____1000____1.914E+00____1.000E+00______000
_____2_____2000____2.861E-03____1.495E-03______445
_____3_____3000____4.240E-03____2.216E-03______360
_____4_____4000____1.290E-04____6.742E-05______650
_____5_____5000____3.120E-05____1.631E-05______835
___ TOTAL_HARMONIC_DISTORTION_=___0.2673718%
The lowest-distortion bias will have to be found by experiment: I don't trust curves or models at voltages and currents far below "normal use". I do think, from curves SPICE and experience, that the plate should sit about 2/3rds of B+, or 16V for a 24V supply. For the usual power tubes strapped triode, a plate resistance of 5K (or 4.7K) gives a good compromise between gain and impedance. Adjust the cathode bias resistor to get the plate to sit about 16V. 470Ω is a good starting trial value, but it may be higher or lower than my poor model says.
But lowest distortion is not everything. This wild idea begs for a front-panel bias pot. Different settings may give "clean", "sweet", "edgy", a "tone control" for distortion balance. A 220Ω fixed resistor and a 1K variable resistor should cover the range of usable bias settings.
I tried pentode-mode, gain is a wee bit higher but THD and output Z are significantly higher. Nothing wrong with that in many cases, so you might want a triode-pentode switch. You could even make the 5K plate resistor a pot, and connect G2 to the wiper, for continuous triode....pentode intermediate modes. Lowest distortion in pentode mode needs a larger cathode resistor than triode mode, 700Ω is a starting point.
The drawback to this plan is that we need about 8 times more power (mostly heater) than a 6SN7 or 12AU7 line-amp working up around 100V-200V, and the power tubes are more expensive than small triodes. Except on today's market, old radio power tubes are in good supply at non-fancy prices, and 24V 1A DC power bricks are commonplace. Compared to cost of construction and the value of good listening, the "waste" here is nothing.
The idea is to run both heater and plate on one 24V DC supply, something you can find as a surplus wall-wart or excess industrial part. That greatly simplifies the power supply, and keeps all voltages down to "safe" levels so you can be relatively fearless working around it. (Yeah, 25L6 wants 25V on the filament, but it will be fine with 24V which is a standard supply voltage.)
An externally hosted image should be here but it was not working when we last tested it.
My previous estimate, based on a squint at a corner of triode-6W6 curves, may be optimistic. Using a SPICE model for the 6L6 strapped triode, I get gain about 4 and output Z about 1.2KΩ. Oh, and of course my earlier estimate of plate current was just wrong: it is roughly 1mA.
My guess is that the 26L6 will have slightly lower voltage gain and output Z, but not much lower.
Using the 50L6 and a 48V DC supply would be "better" for gain, output, and distortion.
It isn't the world's beefiest line-amp, but it should be more than enough to interface hi-fi CD and tape decks to power amplifiers, even if the power amp is a little low on gain or if you want to boost a quiet passage to high levels. And of course the input impedance can be as high as you want. The book suggests 500K max grid resistance, but at these ultra-low voltages it is in no danger with a 5 MegΩ resistor, it will just want a little re-biasing to trim-out grid offset.
And it looks like the old rule of thumb, "peak output at 5% THD is about 20% of supply voltage", is roughly true for large triodes (or triode-strapped pentodes) working at low voltages. I get 5% THD at about 4.8 volts, without trying too hard to optimize bias and output. At 1.9V Peak, a typical power amp input, THD may be well below 1%, maybe 0.5%. At even lower levels I got SPICE-answers of 0.05%, though I don't trust numbers like that from SPICE, especially on tube models.
I did try various bias conditions. Because it is not so far below maximum output, a small shift of bias will cause a significant shift in the balance between 2nd and 3rd harmonic distortions. Mostly the 2nd harmonic can be trimmed from 0.15% to 1.5%, without much change of gain or 3rd harmonic.
_HARMONIC__FREQ____FOURIER_____NORMALIZED______PHASE(DEG)
_____1_____1000____1.914E+00____1.000E+00______000
_____2_____2000____2.861E-03____1.495E-03______445
_____3_____3000____4.240E-03____2.216E-03______360
_____4_____4000____1.290E-04____6.742E-05______650
_____5_____5000____3.120E-05____1.631E-05______835
___ TOTAL_HARMONIC_DISTORTION_=___0.2673718%
The lowest-distortion bias will have to be found by experiment: I don't trust curves or models at voltages and currents far below "normal use". I do think, from curves SPICE and experience, that the plate should sit about 2/3rds of B+, or 16V for a 24V supply. For the usual power tubes strapped triode, a plate resistance of 5K (or 4.7K) gives a good compromise between gain and impedance. Adjust the cathode bias resistor to get the plate to sit about 16V. 470Ω is a good starting trial value, but it may be higher or lower than my poor model says.
But lowest distortion is not everything. This wild idea begs for a front-panel bias pot. Different settings may give "clean", "sweet", "edgy", a "tone control" for distortion balance. A 220Ω fixed resistor and a 1K variable resistor should cover the range of usable bias settings.
I tried pentode-mode, gain is a wee bit higher but THD and output Z are significantly higher. Nothing wrong with that in many cases, so you might want a triode-pentode switch. You could even make the 5K plate resistor a pot, and connect G2 to the wiper, for continuous triode....pentode intermediate modes. Lowest distortion in pentode mode needs a larger cathode resistor than triode mode, 700Ω is a starting point.
The drawback to this plan is that we need about 8 times more power (mostly heater) than a 6SN7 or 12AU7 line-amp working up around 100V-200V, and the power tubes are more expensive than small triodes. Except on today's market, old radio power tubes are in good supply at non-fancy prices, and 24V 1A DC power bricks are commonplace. Compared to cost of construction and the value of good listening, the "waste" here is nothing.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.