I just had to edit the results a little in the previous post for EL36, so check them again.
Still looking good for the tube.
Most tubes follow 3/2 power law quite well on g2. So they all look like 300Bs in screen grid drive mode (except for their high Rp). Grid 1 on the other hand is usually closer to 2 power law, due optimisation of it's gm, causing grid wire proximity effects. This also causes the g1/g2 Mu to increase with current (as 1/3 power) since the g1 gm is outpacing g2 gm. Low gm1 tubes like 6V6 however are closer to 3/2 power law on g1.
Still looking good for the tube.
Most tubes follow 3/2 power law quite well on g2. So they all look like 300Bs in screen grid drive mode (except for their high Rp). Grid 1 on the other hand is usually closer to 2 power law, due optimisation of it's gm, causing grid wire proximity effects. This also causes the g1/g2 Mu to increase with current (as 1/3 power) since the g1 gm is outpacing g2 gm. Low gm1 tubes like 6V6 however are closer to 3/2 power law on g1.
Last edited:
Another small (like 15 Watt) Beam pentode is the 38HE7. Don't know if you can get them in Oz readily. It looks to be half of a (the older twin pent. version) 42KN6. Its a Compactron Beam Pentode plus a damper rectifier in one envelope. The pentode is rated nominally at 10 Watts diss., but it is possible to operate the pentode alone, so it can be uprated some.
Pentode alone removes 7.5 Watts of filament power and 6.5 Watts of damper max plate rating. The filament for the pentode alone (on most, but not all tubes I've checked) is across pins 10 and 12 and requires 21 V to operate.
Should be able to get 15 Watt diss. for the pentode alone. (thats also 1/2 of the 30 W rating of the 42KN6)
The pentode has a knee current of 550 mA at 150 V so is quite good for screen grid drive. It also makes a great triode (at least on the curve tracer) when wired that way. Not much tilt-over in the triode curves.
(12HE7 tends to have the two filament sections wired in parallel, so no way to completely separate.)
And still available for $1 each: http://www.esrcvacuumtubes.com/dollar_days.html
Pentode alone removes 7.5 Watts of filament power and 6.5 Watts of damper max plate rating. The filament for the pentode alone (on most, but not all tubes I've checked) is across pins 10 and 12 and requires 21 V to operate.
Should be able to get 15 Watt diss. for the pentode alone. (thats also 1/2 of the 30 W rating of the 42KN6)
The pentode has a knee current of 550 mA at 150 V so is quite good for screen grid drive. It also makes a great triode (at least on the curve tracer) when wired that way. Not much tilt-over in the triode curves.
(12HE7 tends to have the two filament sections wired in parallel, so no way to completely separate.)
And still available for $1 each: http://www.esrcvacuumtubes.com/dollar_days.html
Last edited:
Another minor adjustment to the calculations for the 6CM5/EL34.
Since the g2/g1 Mu increases slowly with current some (approx. 1/3 power), the Mu in the 400 mA region would be more like 9 (datasheet rated 5.6 at 100 mA). So +9V gets added to the 150 V to compensate for the -1V on the g1 in the curves on page 5 of Philips datasheet. So the 159 V curve is wanted. A 160 V curve is available, giving 420 mA.
Or for 159V exactly: [420]*(159/160)^1.5 => 416 mA.
(alternatively: notice the curves on page 5 of the Philips data drop by around 50 mA between 170 V plate and 35 V plate in the knee region for the 140 V curve. So we can also just take the 468.5 mA figure (for 150V) derived in triode mode from page 4 (at 170 V plate) and subtract 50 mA to get the approx. knee. Giving 418.5 mA, close enough.)
So:
6CM5/EL34 416 mA 12W 34.66 *****************
Since the g2/g1 Mu increases slowly with current some (approx. 1/3 power), the Mu in the 400 mA region would be more like 9 (datasheet rated 5.6 at 100 mA). So +9V gets added to the 150 V to compensate for the -1V on the g1 in the curves on page 5 of Philips datasheet. So the 159 V curve is wanted. A 160 V curve is available, giving 420 mA.
Or for 159V exactly: [420]*(159/160)^1.5 => 416 mA.
(alternatively: notice the curves on page 5 of the Philips data drop by around 50 mA between 170 V plate and 35 V plate in the knee region for the 140 V curve. So we can also just take the 468.5 mA figure (for 150V) derived in triode mode from page 4 (at 170 V plate) and subtract 50 mA to get the approx. knee. Giving 418.5 mA, close enough.)
So:
6CM5/EL34 416 mA 12W 34.66 *****************
Last edited:
Hmmm, E235L does look like the same tube. And this Philips datasheet gives curves for 0V on g1 on page 6:
http://frank.pocnet.net/sheets/009/e/E235L.pdf
Using the 230 mA at 100V knee data:
for 150 V: [230]*(150/100)^1.5 => 422 mA
taking the average of all 3 approaches so far: (416+418+422)/3 => 418 mA
-----------------------------------------
Want to calculate where MH370 is?
http://frank.pocnet.net/sheets/009/e/E235L.pdf
Using the 230 mA at 100V knee data:
for 150 V: [230]*(150/100)^1.5 => 422 mA
taking the average of all 3 approaches so far: (416+418+422)/3 => 418 mA
-----------------------------------------
Want to calculate where MH370 is?
This is kind of off topic, but has anyone tried using screen driven sweep tubes in a guitar amp? I've read quite a bit about it's benefits and I have a boatload of sweep tubes just begging to be put to use, but I can't really find any reference to anyone using such a topology for guitar amplification. Granted, guitarists tend to be a conservative bunch when it comes to gear and this is a Hi-Fi oriented forum, but is there really any downside to using screen drive in a high-power guitar amp? I'll probably get around to trying it out for myself once I clear all my existing projects off my bench, but any insight into this would be helpful.
I've not tried screen drive but sweep tubes make excellent guitar amps.
This is the vintage Oz amp I'm using at the moment (with a few mods to improve its overdrive behaviour):
http://www.ozvalveamps.org/goldentone/1757bassmaster.jpg
60 Watts from a pair of 6DQ6B.
Before going screen drive I'd want to run some experiments to see what happens when overdriven in this mode.
Cheers,
Ian
This is the vintage Oz amp I'm using at the moment (with a few mods to improve its overdrive behaviour):
http://www.ozvalveamps.org/goldentone/1757bassmaster.jpg
60 Watts from a pair of 6DQ6B.
Before going screen drive I'd want to run some experiments to see what happens when overdriven in this mode.
Cheers,
Ian
RCA Victor 1952 television receiver
I just acquired a RCA Victor 1952 television receiver.
Have any of you tried a make a guitar amp from one of these?
CB
I just acquired a RCA Victor 1952 television receiver.
Have any of you tried a make a guitar amp from one of these?
CB
Yes, and if you play guitar anything like I do, the amp will blow up in a most spectacular and destructive manner. Screen drive has an inherent weakness. It can be anywhere from a non issue to a fatal flaw depending on the tube and how the amp is used.
Most screen driven amps operate the control grid near, or at zero volts and all drive is applied to the screen grid. This affords a higher efficiency, lower idle current, and often lower distortion in exchange for higher drive requirements and higher output impedance.
The higher efficiency is achieved because the tube can be driven further into saturation reducing the power lost due to voltage drop across the tube. I have indeed seen the plate pulled all the way down to zero volts before settling at 35 volts when the amp is driven to clipping. The screen grid required 250 volts of positive drive to reach this point. At the instant the plate voltage is in the 35 volt and below region, the screen grid is at +250 volts. Where do you think most of the cathode current is going? That's right.....it's going directly to the screen grid and warming it up, QUICKLY!
If you have built a nice HiFi amp and never drive it to clipping, you won't have a problem. The non issue.
If you built a 150 WPC screen driven amp and play LOUD rock music with peaks clearly clipped off, you probably won't have a problem, but watch the screen grids carefully for a while when you lean on it hard. If there is ANY glow appearing on the screen, turn it down!
Let's say you (or I) built a screen driven guitar amp. You decide to do your best Jimi imitation and set all the controls on 11, and engaged every pedal you own, and cranked them to 11. You are greeted with an awesome full metal racket for about 30 seconds, then the fire gods appear quickly. Before you can hit the kill switch, the driver circuitry, both output tubes, and the OPT are burnt! WHY???? I had to blow the amp up a few times to figure it all out.
The screen grids are being overloaded by being driven highly positive at the same moment the plate is near zero. When the amp is driven to square wave operation the duty cycle is about 50%. One of the output tubes is saturated while the other is cut off. The screen grids are glowing white hot at this moment.
What happens when the screen grid is white hot, when the tube switches from fully saturated to cut off.....it cools down, right? WRONG. The screen grid is driven to zero, or negative to cut off the tube. Mean while the plate is pulled up to twice the B+ voltage by the OPT. OK the glowing grid is at -100 volts, and the plate is at 1200 volts. Does current flow? BIG TIME!!!! the screen grid can emit electrons and will begin to act like a cathode. Reverse screen current will flow causing further heating of the screen grid. Within a few seconds there will be a large lightning bolt through the tube leading to an arc from plate to screen, blowing the tube the screen circuitry, and possibly the OPT.
Some tubes are much more susceptible than others, but all sweep tubes can be blown up in this manner. I have been working with several method to prevent this, but so far the best compromise applies drive to both grids simultaneously. Grounded grid (cathode drive) also works, since it's really the same thing. Current limiting the screen circuitry helps, but costs some output power, and it's possible to blow up the limiter too.
See post 104 in this thread for a picture. It was 2008 when I first blew up a screen drive amp....I still haven't made one that can eat my guitar playing.
http://www.diyaudio.com/forums/tubes-valves/128533-tube-sale-aes-11.html
Hah! - No need to reinvent the wheel, Tubelab has done the overdrive experiments for us.
When using sweep tubes for guitar in normal pentode mode, however, this is not a problem.
The one thing those 6DQ6 Goldentones were known for, was their reliability. While Fenders and Marshalls were emitting smoke all around them, they just "kept on truckin'" and eventually (at the end of the tube life) just faded out over a few gigs rather than expiring instantly and dramatically like their EL34 and 6L6 equiped cousins.
Cheers,
Ian
When using sweep tubes for guitar in normal pentode mode, however, this is not a problem.
The one thing those 6DQ6 Goldentones were known for, was their reliability. While Fenders and Marshalls were emitting smoke all around them, they just "kept on truckin'" and eventually (at the end of the tube life) just faded out over a few gigs rather than expiring instantly and dramatically like their EL34 and 6L6 equiped cousins.
Cheers,
Ian
Well then, I guess screen drive is out for a monster guitar amp. Those Goldentone amps are interesting though, and I've got a few 6DQ6s and 12DQ6s (as well as some 12GC6s, which appear to be the same tube with a different pinout), so there's plenty of promise there. However, I also have over 20 *6LW6 tubes of various constructions, as well as 15 or so 26HU5s that are just begging to be put into something big. I've also got a 625-0-625V .5A plate transformer and a hefty 5K OPT, although I'll probably put a 4 ohm load on the 8 ohm tap to get 2.5K. Around 150V on the screens seems to be fairly normal practice for most of the high powered sweep tube amp designs I've seen, is there any benefit to going higher or is it just an unnecessary risk? The screens on a 6LW6 are rated at 275V max, but is there any real advantage to running them that hot or is that just shortening their lifespan?
As for drive, they'll likely need a ton of negative bias on g1, which means a ton of drive voltage. Do big sweeps typically draw much grid current? I've thought about going the SVT route and using DC coupled cathode followers to drive the grids directly (or using a MOSFET source follower, but even though it probably wouldn't make a difference tonally, it seems like heresy), but I'd rather not trouble with building a negative rail capable of sourcing significant current if it's unnecessary. A pentode LTP sort of similar to that found in Pete Millett's Engineer's Amplifier looks like an interesting proposition, albeit minus the CCS in the tail. Not sure if there's any precedent for it in the guitar world or how it'll sound when overdriven, though, but I'm doubtful that the standard LTP found in most guitar amps can swing anywhere close to enough voltage.
As for drive, they'll likely need a ton of negative bias on g1, which means a ton of drive voltage. Do big sweeps typically draw much grid current? I've thought about going the SVT route and using DC coupled cathode followers to drive the grids directly (or using a MOSFET source follower, but even though it probably wouldn't make a difference tonally, it seems like heresy), but I'd rather not trouble with building a negative rail capable of sourcing significant current if it's unnecessary. A pentode LTP sort of similar to that found in Pete Millett's Engineer's Amplifier looks like an interesting proposition, albeit minus the CCS in the tail. Not sure if there's any precedent for it in the guitar world or how it'll sound when overdriven, though, but I'm doubtful that the standard LTP found in most guitar amps can swing anywhere close to enough voltage.
The LW6 comes in several sizes from big to huge. All work great. The 26HU5 is also another good tube.
Pete's amp is an excellent guitar amp as is with no modifications other than a serious power increase and whatever wiring modifications needed to fit your tube of choice. I took a fairly stock board, hacked the heater wiring to run 35LR6's in it and plugged 6GU5's in for the LTP (despite being a hexode, it's a drop in). On the bench it made 250 WPC with clean sine wave testing at 650 volts plate supply and a 2500 ohm load. There was a pale red plate glow in a dark room after several minutes of continuous testing. Paralleling both channels through a 1250 ohm OPT yielded 525 Watts, but I didn't let that combo run long.
I connected my old ADA MP1 guitar preamp up to both channels wired a speaker cabinet up to one channel and a load resistor to the other channel. I dialed the power supply back to just under 600 volts and attempted to blow it up. No dice.....I annoyed my neighbors for 3 or 4 days a few years ago and the whole thing just worked. Power output was about 200 WPC.
All of my testing was done with 150 to 175 volts on the screen. Most true sweep tubes work best in this range. More screen voltage requires more negative grid bias, which allows (and requires) more drive voltage, but after 150 or 160 volts, doesn't produce much more power. The exception seems to be when trying to extract maximum power at low plate voltages and low load impedances, but this drives up the screen and cathode current shortening tube life and reliability. I still never got past 200 volts (13GB5 on 275 plate volts).
The details are buried throughout the red board thread.
Pete's amp is an excellent guitar amp as is with no modifications other than a serious power increase and whatever wiring modifications needed to fit your tube of choice. I took a fairly stock board, hacked the heater wiring to run 35LR6's in it and plugged 6GU5's in for the LTP (despite being a hexode, it's a drop in). On the bench it made 250 WPC with clean sine wave testing at 650 volts plate supply and a 2500 ohm load. There was a pale red plate glow in a dark room after several minutes of continuous testing. Paralleling both channels through a 1250 ohm OPT yielded 525 Watts, but I didn't let that combo run long.
I connected my old ADA MP1 guitar preamp up to both channels wired a speaker cabinet up to one channel and a load resistor to the other channel. I dialed the power supply back to just under 600 volts and attempted to blow it up. No dice.....I annoyed my neighbors for 3 or 4 days a few years ago and the whole thing just worked. Power output was about 200 WPC.
All of my testing was done with 150 to 175 volts on the screen. Most true sweep tubes work best in this range. More screen voltage requires more negative grid bias, which allows (and requires) more drive voltage, but after 150 or 160 volts, doesn't produce much more power. The exception seems to be when trying to extract maximum power at low plate voltages and low load impedances, but this drives up the screen and cathode current shortening tube life and reliability. I still never got past 200 volts (13GB5 on 275 plate volts).
The details are buried throughout the red board thread.
You need to be more specific about the TV you have. The schematic, or even the tube lineup would help. There is a high likelihood that many of the old parts are no longer usable. I wouldn't reuse any old resistors or capacitors.
There were lots of TV's from the 50's in the trash when I started making guitar amps in the early 60's. That was all I had for raw material when I started building tube amps as a kid, but there was plenty of it, and failures were simply carted back to the trash dump where they came from. Whether this is even economical today, or practical, I don't know, but if you want to do it "just because", here is the info:
Most of the circuitry in a TV is useless for a guitar amp, but there are 3 exceptions.
Every TV has an audio amp to drive the speaker. It is usually a single ended amp capable of making a couple of watts. The common audio output tubes in the 50's were 6V6, 6K6 or 6AQ5. Follow the speaker wires to a small transformer. It will have 4 wires, usually red, blue, green and black. The blue wire will go to the plate pin of the audio output tube. All of the associated circuitry is an audio amplifier. It can directly be copied as the output stage of a guitar amp.
Every TV also has a vertical deflection circuit. It drives the deflection yoke (the coil of wire and ferrite mounted on the neck of the picture tube). It is a single ended amp capable of a few watts. In many early TV sets it is an exact copy of the audio amplifier, with the exception of the output transformer, which is usually bigger since it runs at 60 Hz. The transformer will have 3 or 4 wires. The 3 wire versions are useless for guitar amps (it is an autotransformer which will put B+ on the speaker). Common vertical output tubes in the early 50's were the 6V6, 6K6 and the 6AQ5, but some small TV's still used the 6SN7. Again the vertical output stage is an audio amp and can be used as the output stage of a guitar amp. This is what I used in my early guitar amps. I fed this with a preamp circuit copied from a Fender Champ circuit. I used the 6SJ7 circuit since I had plenty of these tubes from old radios. The same circuit could be built with the 6au6 or 6CB6 tubes found in 1950's TV's. 12AX7's weren't commonly available for free back then, but that's what I would use today.
Either of these circuits will need a power supply, which could be lifted from the TV, but it is total overkill and will be heavy. I used the transformer and power supply circuitry from old radios since they were common in the trash dump.
The horizontal sweep and high voltage supply circuit of a TV contains a power tube capable of making a powerful guitar amp, but you will need more parts. The common tubes from the 1950's are the 6BQ6 and 6DQ6. One of these in a single ended circuit will make about 10 watts. The 6BQ6 is the easiest to deal with in SE and it is what I used for my "Champ" circuits. I usually mixed and matched tubes, transformers and parts from various TV's radios, and HiFi sets that were available on any given day that I visited the trash dump (or the dumpster behind the TV repair shop), but this is no longer an option. I used the biggest vertical output transformer that I could fine, a 6BQ6GT for output, a 6SJ7 for the input tube, and a power supply liberated from an old HiFi or radio.
The horizontal output tubes could be used for a push pull guitar amp. Again I used mix and match parts. I often used a power transformer from a TV radio, or HiFi for an output transformer. Transformers with multiple or higher voltage heater windings are best. Use the high voltage winding for the plate circuit, and wire all of the heater windings in series for the speaker connection. Do not use the line voltage winding. Some don't work so well, but some sounded reasonably good. The turns ratio favors a low impedance speaker load, which is why my practice then of wiring every speaker I brought home together in parallel worked so well back then. I didn't quite understand that impedance matching stuff, or those load line things until high school, but that didn't stop me from melting a few tubes in the name of making sound.
The horizontal sweep tubes commonly found in TV sets are recently being "discovered" as excellent audio amp tubes, but as the post about the Goldentone states they were used as guitar amps several decades ago. The 6DQ6 used in the Goldentone was common in the 60's but isn't in huge supply today. The venerable 6BQ6 was used from the 40's through the 60's and millions of them were made. It is considered "too small" to be sucked up into the CB/ham radio linear amplifier vortex, so there are still plenty of them around.
"I dialed the power supply back to just under 600 volts and attempted to blow it up."
You know George, if I were your neighbour, I would be annoyed too after 3 or 4 days. I would be annoyed that I wasn't invited over to be in your garage with this experiment going on.
The anode dissipation ratings of TV valves are a bit confusing: the EL36 is rated at 12W, while the E235L is rated at 20W. Given the size of the structure that looks more like it, with the usual 'audio' outputs in mind.
I think I read somewhere that the 6V6 for TV sweep duty was rated at 9W in contrast to the 12-14W for audio.
In the datasheets for the PL519 there are two 'Limiting Values': Design center rating system (35W) and Design Max rating system (45W).
So how should we interpret the ratings?
In the case of the PL519 35W idle is ok and at max power output 45W would be still ok?
Or do we go the Tubelab way: turn it up untill it glows and back off a little?
I see Philips rates the E235L at 15 Watts, Telefunken rates it at 20 W but has 12 W dotted curves drawn in on all the pentode curve graphs. And Siemans rates it at 12 W. Maybe they just built the tubes differently from each other, or put in some extra margin for marketing.
I think the design center rules were allowing for typical line voltage variation and component variation safety margin. And absolute max. would be if you have regulated voltages and control of other variables. The RCA manual should have a write up on the different rating rules. Audio is different because of the signal peak to avg. ratio.
I think the design center rules were allowing for typical line voltage variation and component variation safety margin. And absolute max. would be if you have regulated voltages and control of other variables. The RCA manual should have a write up on the different rating rules. Audio is different because of the signal peak to avg. ratio.
Last edited: