Hello all:
I have a few compactron tubes that I would like to try using. There are some @ 21v filament voltage. I am seeking advice on how to most effectively achieve this non standard voltage. Thanks for the help.
I have a few compactron tubes that I would like to try using. There are some @ 21v filament voltage. I am seeking advice on how to most effectively achieve this non standard voltage. Thanks for the help.
Plenty of 20VAC transformers at Digikey or Mouser. Or you could use a dual-secondary 10VAC tranny wired in series. Most of them have 115 primaries so you'll get a bit of over-voltage.
Hello folks:
thanks for contributing.
The first project may be a 21LU8 PP triode connected. The filament draw will be 4 @ 0,6A. Can be either AC or DC. Therefore with a bit of cushion to keep things cool I guess I am looking at 3A. I will explore the idea of using a 20v toroid under the top plate. My house mains is at 120VAC and I am the only one on the street transformer, so clean steady power coming into the house.
I am planning on using 6be3 compactron damper diodes for the B+. Filament power for these will easily come from the main transformer.
The next project I am interested in, is a 38he7PP. The filament draw is 0.45/tube so 1.8A with a bit of cushion, say 2 - 2.5A either AC or DC.
I guess I need to explore how exacting the filament voltage needs to be. What sort of operating range is going to work properly. ie a little starved, a little overfeed or just right.
I was also thinking of regulated DC, but the circuitry may produce a lot of under chassis heat.
Perhaps for 38V I can use a 36V toroid if I am in the right range for heater allowances.
Any experience comment or help is appreciated.
Thanks
thanks for contributing.
The first project may be a 21LU8 PP triode connected. The filament draw will be 4 @ 0,6A. Can be either AC or DC. Therefore with a bit of cushion to keep things cool I guess I am looking at 3A. I will explore the idea of using a 20v toroid under the top plate. My house mains is at 120VAC and I am the only one on the street transformer, so clean steady power coming into the house.
I am planning on using 6be3 compactron damper diodes for the B+. Filament power for these will easily come from the main transformer.
The next project I am interested in, is a 38he7PP. The filament draw is 0.45/tube so 1.8A with a bit of cushion, say 2 - 2.5A either AC or DC.
I guess I need to explore how exacting the filament voltage needs to be. What sort of operating range is going to work properly. ie a little starved, a little overfeed or just right.
I was also thinking of regulated DC, but the circuitry may produce a lot of under chassis heat.
Perhaps for 38V I can use a 36V toroid if I am in the right range for heater allowances.
Any experience comment or help is appreciated.
Thanks
A bit of quick research:
This may work for the 21lu8
https://www.antekinc.com/content/AS-1220.pdf
this is a bit overkill for the 38he7
https://www.antekinc.com/as-2238-200va-38v-transformer/
All good.
Clearly there are options available.
This may work for the 21lu8
https://www.antekinc.com/content/AS-1220.pdf
this is a bit overkill for the 38he7
https://www.antekinc.com/as-2238-200va-38v-transformer/
All good.
Clearly there are options available.
Weren't tubes with those odd voltages meant to be wired in series, so they'd add up to 110v? If you already got 4 of em, that's 84v...
I've been down this path recently. I'm not a fan of multiple power transformers in my builds (not sure why, it's just me. I like small amps). Since the odd voltage tubes often were meant for serial heaters, that's a potential option. Voltage multipliers can then be utilized to bring B+ to healthy levels. If the heater string turns out to be DC one can also implement a current regulator in series to act as a soft start.
I have a napkin design for a 25C5 PP amp. I use an Antek 40+40VAC toroidal transformer. Rectifying the 80VAC you get 113VDC which is perfect for the 100V needed for the heaters plus enough to add a current regulator that will double as a soft start. You could do something similar if you can work out a way to get the B+ you need.
I have a napkin design for a 25C5 PP amp. I use an Antek 40+40VAC toroidal transformer. Rectifying the 80VAC you get 113VDC which is perfect for the 100V needed for the heaters plus enough to add a current regulator that will double as a soft start. You could do something similar if you can work out a way to get the B+ you need.
120/120 to 12/12 transformers are common in control circuits,
The 24V runs the 21LU8s thru small dropping resistors & many plain Jane 12V voltage amp toobz on the 12V wdgs,
Check Hammond & many others. 👍
The 24V runs the 21LU8s thru small dropping resistors & many plain Jane 12V voltage amp toobz on the 12V wdgs,
Check Hammond & many others. 👍
Hello Mr: Dudadub:
An interesting thought. A series string, hmmm.
More to consider. Does the current draw play a role in the calculations and what about heat mgmt in dropping resistors?
Thanks
An interesting thought. A series string, hmmm.
More to consider. Does the current draw play a role in the calculations and what about heat mgmt in dropping resistors?
Thanks
Hi Scot,
Remember that if you go DC, you can regulate for either voltage or current. You don't have to make a full string to get to 120VAC, if you just put 2 in series for 42V and use an appropriate transformer for DC supply and then regulate for DC current you will have less heat to worry about.
I'm not sure a non-isolated AC heater string is safe, even if some here do you use it on new builds. I am thinking of the CSA custom tag certification procedure, and I would think that if you made a mains unisolated heater string it would fail at the tube socket on a Megger test, and thus not meet CSA.
Remember that if you go DC, you can regulate for either voltage or current. You don't have to make a full string to get to 120VAC, if you just put 2 in series for 42V and use an appropriate transformer for DC supply and then regulate for DC current you will have less heat to worry about.
I'm not sure a non-isolated AC heater string is safe, even if some here do you use it on new builds. I am thinking of the CSA custom tag certification procedure, and I would think that if you made a mains unisolated heater string it would fail at the tube socket on a Megger test, and thus not meet CSA.
Hello Tony:
Good thoughts.
I agree I would not use a non-islolated string connected to the mains. I would use a transformer for sure. The two I noted in post 6 are the first ones I found. Now I will carry on and find the perfect one for the 21LU8. A 50va 20vac toroid would be the minimum and not too expensive so that is like the way I will go.
I have never built a regulator so I would have to do a bunch of learning to get it correct, it is a possibility if there is an advantage over AC filaments. There is not a lot of information out there to reference against with the tubes so I would be trying something based on principle rather than experience.
I can say the 6LU8 is a very nice sounding tube. With the heaters referenced to 25% b+ the noise floor seems pretty low to me.
Lots think about.
Time to start designing. Pencil, Paper and Calculator for me, seems to work best.
Good thoughts.
I agree I would not use a non-islolated string connected to the mains. I would use a transformer for sure. The two I noted in post 6 are the first ones I found. Now I will carry on and find the perfect one for the 21LU8. A 50va 20vac toroid would be the minimum and not too expensive so that is like the way I will go.
I have never built a regulator so I would have to do a bunch of learning to get it correct, it is a possibility if there is an advantage over AC filaments. There is not a lot of information out there to reference against with the tubes so I would be trying something based on principle rather than experience.
I can say the 6LU8 is a very nice sounding tube. With the heaters referenced to 25% b+ the noise floor seems pretty low to me.
Lots think about.
Time to start designing. Pencil, Paper and Calculator for me, seems to work best.
For the heaters, the B+ or both? If you are relatively new at this, don't get too complicated,
otherwise you will do a lot of wheel spinning & waste a lot of time. Better to accomplish something simpler & get some experience.
And have some time to enjoy your music collection & a glass of wine. Or a beer if you prefer.👍
Hello Matt:
Thanks for commenting:
I am a little concerned about managing heat from the dropping resistor as I am pulling 2.6A. Perhaps that is a solvable concern.
I also realized that 16 x 1.4 is 22.2. With a little smoothing and a hot mains power supply I should have no problem getting 21vdc from a 16v, 50va transformer.
Hello Mike:
I will be exploring that option.
Good sage advice John.
I guess I am a little excited and reveling in recent success with a keen enthusiasm for my next project.
Thanks for all the help
Thanks for commenting:
I am a little concerned about managing heat from the dropping resistor as I am pulling 2.6A. Perhaps that is a solvable concern.
I also realized that 16 x 1.4 is 22.2. With a little smoothing and a hot mains power supply I should have no problem getting 21vdc from a 16v, 50va transformer.
Hello Mike:
I will be exploring that option.
Good sage advice John.
I guess I am a little excited and reveling in recent success with a keen enthusiasm for my next project.
Thanks for all the help
You will not get that at all. On a good day you will get about (1.25x16) - 2 = 18V
under load at the output of the rectifier bridge. Then you can smooth it. Or not.
Real fact you should carefully consider is simply why does this tube need DC on the heater filaments leads at all ???🙄
Would be considered very poor design. simply something else to fail.👎👎
What comes out of the rectifier stack depends on factors in the transformer.
DC resistance of the windings. And how tightly coupled the secondary is to the primary. Leakage Inductance.
Some of the small transformers are designed to sag under load. And the results for simple AC resistive loading
often vary a lot from a rectified load into large capacitors. All can result in a crap shoot.
Designer beware, don't build in a problem where there was none.
DC resistance of the windings. And how tightly coupled the secondary is to the primary. Leakage Inductance.
Some of the small transformers are designed to sag under load. And the results for simple AC resistive loading
often vary a lot from a rectified load into large capacitors. All can result in a crap shoot.
Designer beware, don't build in a problem where there was none.
All of the tubes used in any series heater string circuit MUST have the same heater CURRENT draw requirement, AND all of the series wired voltages must add up to the available supply voltage or a suitable dropping resistor must be chosen to drop the excess voltage.
I built a little series heater string guitar amp that used a pair of 32ET5's an 18FW6 and an 18FY6 to make 4 watts on 165 to 170 volts of B+. It used a 50 VA Triad N-68X 120 volt to 120 volt isolation transformer for power. The heaters get the pulsating DC from the rectified but unfiltered output from the bridge rectifier. The heater string consumes 100 mA and adds up to 100 volts. The output from the transformer is about 125 volts so I would need a resistor of about 250 ohms to drop 25 volts. The schematic shows a 220 ohm but the amp actually has a 240 ohm resistor which is close enough. The AC from the transformer goes directly into a full wave bridge rectifier, but there is no filter cap immediately after the bridge. Here the tubes are heated with pulsating DC that has exactly the same heating energy power as the pure AC did. Current flows through the bridge and the heaters 100% of the time, placing about 12.5 VA of purely resistive load on the transformer.
Diode D5 allows current to flow from the bridge into the main filter cap, C16 during the brief period at the peaks of the sinewave when the instantaneous voltage is higher than the cap's stored voltage. These brief current surges are not ideal for a small transformer and should not be the bulk of the load on the transformer. Fortunately, the load at full crank is only about 75 mA and idle is 50 mA so the transformer sees 8 to 12 VA of "ugly" load. Even at about 25 VA of load the little transformer gets noticeably warmer than it does unloaded when left on overnight.
I have a single channel breadboard design for a 50C5 PP amp using UNSET technology to deliver 20 watts at about 4% THD. The plan was for this to be a small guitar amp using series string heaters. The power amp was built and successfully tested. The preamp section was never built, but two dummy tubes were put in place that have the same heater requirements as the two 12AX7's that already exist on another PC board.
This setup is powered from a simple 120 volt to 120 volt isolation transformer. In this case I also used a Triad N-68X which is rated for 50 VA. The load seen by the transformer is mostly resistive so a 50 VA transformer works fine to supply about 18 VA of purely resistive load to the heaters, about 40 mA of 330 volts (13 VA) via a full wave voltage doubler and about 5 mA of 165 volts (1 VA) via a full wave bridge all at the same time. Here the transformer is conducting more than half the load all of the time.
If a filter capacitor was tied to the output of the bridge it would charge to somewhere near the peak value of the sinewave (160 to 170 volts). The current through the bridge flows only when the input sine wave is higher than the charge in the capacitor.
The diodes for the screen and plate supplies only pull current on the peaks of the sine wave.
If the heaters are powered from the rectified and filtered DC supply a 50 VA transformer all of the amp's current draw would be pulled from the transformer only in brief pulses near the peaks of the sine wave. This makes the transformer very unhappy leading to it buzzing and getting hot.
The breadboard was put together bit by bit from scratch with nothing to go by. I just built a stage, messed with it until it worked, then added another. There is no accurate schematic of what I have yet. I have been tracing it and modifying the schematic of an old design to match what's on the breadboard. Only the power supply and push pull output stage actually match the schematic.
