Power transformer...
Well 2 really.
One for filament supply, one for main/HT supply.
Its for a tubed line stage project.
Requirements are:
12V @ 3A for the filaments...
Dual-rail 30V for the main.. (+ / - 30V) @ 2A should be good.
I'm looking at these:
Hammond Mfg. - Toroid Power Transformers - (1182 Series)
The model 1182M12 for filaments, 12V @6A
and 1182N15 for the main dual rail supply, 30V @ 4A... Center tap.
I realize these are both a bit overkill.... but I prefer it that way...
Cool running and plenty of "headroom"... and only cost about ten bucks more each than the minimum spec'd models..
So I've doubled the current needs...
Or do I have this backwards??????
The center tap with one side of each winding connected to ground? No?
IE: The series connection with the center leads to case ground, for the dual rail supply..
The parallel connection for the straight 12v to the (seperate) filament supply?
Connection sheet here FWIW:
http://www.hammondmfg.com/pdf/1182_Insert.pdf
I'd be extremely greatful for some help, before I make an expensive mistake!!!
TIA!!!
Greg
Well 2 really.
One for filament supply, one for main/HT supply.
Its for a tubed line stage project.
Requirements are:
12V @ 3A for the filaments...
Dual-rail 30V for the main.. (+ / - 30V) @ 2A should be good.
I'm looking at these:
Hammond Mfg. - Toroid Power Transformers - (1182 Series)
The model 1182M12 for filaments, 12V @6A
and 1182N15 for the main dual rail supply, 30V @ 4A... Center tap.
I realize these are both a bit overkill.... but I prefer it that way...
Cool running and plenty of "headroom"... and only cost about ten bucks more each than the minimum spec'd models..
So I've doubled the current needs...
Or do I have this backwards??????
The center tap with one side of each winding connected to ground? No?
IE: The series connection with the center leads to case ground, for the dual rail supply..
The parallel connection for the straight 12v to the (seperate) filament supply?
Connection sheet here FWIW:
http://www.hammondmfg.com/pdf/1182_Insert.pdf
I'd be extremely greatful for some help, before I make an expensive mistake!!!
TIA!!!
Greg
I'm quite confused.
36 Watts of heaters, but a +/-30V DC power supply? At 2A?
Twenty LOW-volt 12A_7 feeding a 2*50W transistor amp?
And you do not get +/-30V DC from 30V CT AC. More like +/-20V or +/41V(ignore CT).
Maybe you better make a sketch of what this is.
From here, it BEGS for unification. "12V 3A" could very-likely be re-rigged as 36V 1A string, which could run from +20V to -20V, and be DC heat.
36 Watts of heaters, but a +/-30V DC power supply? At 2A?
Twenty LOW-volt 12A_7 feeding a 2*50W transistor amp?
And you do not get +/-30V DC from 30V CT AC. More like +/-20V or +/41V(ignore CT).
Maybe you better make a sketch of what this is.
From here, it BEGS for unification. "12V 3A" could very-likely be re-rigged as 36V 1A string, which could run from +20V to -20V, and be DC heat.
Hi..
Ok.. no.... its a tube buffer board.. 2 6DJ8/6922 tubes per board.
Each board calls for +/- 12 to 30VDC and 12V for the filaments..
I'm using 2 of those boards...
I have a dual rail regulator board for each... as well as a regulator filament supply board for each as well.
Ok.. no.... its a tube buffer board.. 2 6DJ8/6922 tubes per board.
Each board calls for +/- 12 to 30VDC and 12V for the filaments..
I'm using 2 of those boards...
I have a dual rail regulator board for each... as well as a regulator filament supply board for each as well.
I should also mention:
The filament supply boards max out at 1.5A each.
I'd like to be able to plug in the russian tubes directly... meaning without any circuit mods.
They draw something like 350ma each for the filaments... (if memory serves)
I realize the rest won't draw anywhere near the 4A I mentioned..
Though the main regulator boards can do 1A continuous, so might as well supply then with that I figured...
Sorry... I never know how much detail to go into initially..
The filament supply boards max out at 1.5A each.
I'd like to be able to plug in the russian tubes directly... meaning without any circuit mods.
They draw something like 350ma each for the filaments... (if memory serves)
I realize the rest won't draw anywhere near the 4A I mentioned..
Though the main regulator boards can do 1A continuous, so might as well supply then with that I figured...
Sorry... I never know how much detail to go into initially..
Last edited:
Over-engineering often turns out to be actually under-engineering. If a circuit uses, say, 60mA then there is no advantage in designing a 600mA supply for it. There may be a disadvantage, though: bigger transformer leaking more magnetic field, bigger caps needing longer wires (so more magnetic field).
Heaters usually run on AC direct from the heater winding.
if you use a dedicated heater transformer and use an EI style then you will get a high transformer regulation due to using an EI and due to using a small transformer. Expect a 20VA to 30VA EI to have a regulation anywhere from 10% to 20%.
Take your normal 6.3Vac heater that needs 350mAac.
Attach a 30VA transformer that is rated for (30/6.3) 4.76Aac and draw only 3times 350mAac for your three valves (total = 1050mAac = 22% of rated load) and you will get somewhere around 7.2Vac at your heaters. What life does one get from the heaters at this elevated voltage?
That may be why most valve tube amplifiers have a combined HT plus heater from a single transformer. The bigger rating gives a lower regulation and the more constant load from all the ClassA valve stages holds the winding voltage closer to rated voltage. = longer life from the heaters.
I have only built one pair of valve amplifiers and that was from a kit of bits with a proper build guide, so I know very little about valve transformers.
if you use a dedicated heater transformer and use an EI style then you will get a high transformer regulation due to using an EI and due to using a small transformer. Expect a 20VA to 30VA EI to have a regulation anywhere from 10% to 20%.
Take your normal 6.3Vac heater that needs 350mAac.
Attach a 30VA transformer that is rated for (30/6.3) 4.76Aac and draw only 3times 350mAac for your three valves (total = 1050mAac = 22% of rated load) and you will get somewhere around 7.2Vac at your heaters. What life does one get from the heaters at this elevated voltage?
That may be why most valve tube amplifiers have a combined HT plus heater from a single transformer. The bigger rating gives a lower regulation and the more constant load from all the ClassA valve stages holds the winding voltage closer to rated voltage. = longer life from the heaters.
I have only built one pair of valve amplifiers and that was from a kit of bits with a proper build guide, so I know very little about valve transformers.
Last edited:
> Dual-rail 30V for the main.. (+ / - 30V) @ 2A should be good.
> Each board calls for +/- 12 to 30VDC
12V, 30V, even 60V is not much for tubes.
The current can NOT be anything like 2A! I can hardly squint "24K" and "10K" on the schematic. So even if the tubes were dead-short (not a happy operating point), each channel is worst-case a 7K resistor, 3.5K for two channels. 60V in 3,500r is 0.017 Amps. 1/100th what you propose.
I also have real doubt the supply needs to be dual-rail. I agree that is what the plan shows. I strongly suspect though that "Vdd" (CMOS lingo!) could be strapped to ground, +60V to "Vcc", and work like 99% of all tube amps ever built. The split-rail could allow lower cap ratings, but at these voltages you can't buy low-V caps of good quality.
60V DC (maybe CT) at (round up) 0.05A 50mA DC. This can be a 42VAC 0.1A (not 4A!) AC winding. Unless you actually find sweet 63V caps, it can be 48VAC 0.1A.
> 12V @ 3A for the filaments...
Likely suspects are 0.3A or 0.35A. If you must fool with odd tubes, you could find 0.45A@12V bottles. So you need just shy of 1A of DC, which can be 2A of AC rating. 2A or 3A is fine. _IF_ you can really rectify 12VAC to a high enough DC voltage to cover regulator drop so you do not have ripple-dips in the "DC".
A hasty sim of 12V 2A AC into fat diodes and 2,200uFd gives ripple-dips to 13.8V. This will dip-out with the usual regulators. The LDO types may give 12.6V through ripple dips, but that leaves near-no margin for line-voltage dips. 10,000uFd gives 15.5V ripple-dips, which may allow a non-LDO reg to get by.
> Each board calls for +/- 12 to 30VDC
12V, 30V, even 60V is not much for tubes.
The current can NOT be anything like 2A! I can hardly squint "24K" and "10K" on the schematic. So even if the tubes were dead-short (not a happy operating point), each channel is worst-case a 7K resistor, 3.5K for two channels. 60V in 3,500r is 0.017 Amps. 1/100th what you propose.
I also have real doubt the supply needs to be dual-rail. I agree that is what the plan shows. I strongly suspect though that "Vdd" (CMOS lingo!) could be strapped to ground, +60V to "Vcc", and work like 99% of all tube amps ever built. The split-rail could allow lower cap ratings, but at these voltages you can't buy low-V caps of good quality.
60V DC (maybe CT) at (round up) 0.05A 50mA DC. This can be a 42VAC 0.1A (not 4A!) AC winding. Unless you actually find sweet 63V caps, it can be 48VAC 0.1A.
> 12V @ 3A for the filaments...
Likely suspects are 0.3A or 0.35A. If you must fool with odd tubes, you could find 0.45A@12V bottles. So you need just shy of 1A of DC, which can be 2A of AC rating. 2A or 3A is fine. _IF_ you can really rectify 12VAC to a high enough DC voltage to cover regulator drop so you do not have ripple-dips in the "DC".
A hasty sim of 12V 2A AC into fat diodes and 2,200uFd gives ripple-dips to 13.8V. This will dip-out with the usual regulators. The LDO types may give 12.6V through ripple dips, but that leaves near-no margin for line-voltage dips. 10,000uFd gives 15.5V ripple-dips, which may allow a non-LDO reg to get by.
Hello Gents..
Thanks for the replies!!
I think I might have committed the error of "solid state thinking".
DF..
Thank you for teaching me the upper limit of overengineering (as a purposeful approach).
That most of the equipment I've heard and liked best used overbuilt PS is probably what led me there...
Hello Andrew.... thank you as usual for your input!
Incidentally, this is for the project mentioned in that other thread..
A valve buffer before and after attenuator, as "linestage".
2 x 6dj8/6922 tubes per board x 2 boards... 4 tubes total.
(2 stereo channel, of course)
They call for 12v for the filaments, per board.
I should probably add, at this point....
I have boards for both the main supply and the fillament supply. (regulated dc output).
This is my first "scratch build" (obviously lol) and there's nothing like the learning curve!!!
Its not JUST a learning project though... there's also a need for a certain piece of equipment in my life...
In fact, its probably a 60/40 or maybe even a 70/30 proposition at this point..
IE: the need to get it done!
(lest a box of parts lay unfinished in back of a closet for years due to indecision etc).
I'd rather proceed with what I have so far, than start from scratch at this point...
Part of the learning for me, will be to re-do/improve certain portions.. after determining what worked/didn't/could be improved upon.
It'll take long enough/be hard enough just to get something/anything up and running..
There's all the other aspects to contend with too.. physical assembly, soldering, layout, etc. etc.
I'd rather do that, then think... re-think certain aspects afterwards (possibly while listening to it).
A different/better valve buffer configuration for instance, possibly even p2p wired (as that other fellow mentioned)....
A more ideal supply for it...
All of those things take time... which I'd rather spend later and go back to re-do/improve aspects...
I'd rather not have to replace the transformer(s) later on though!
The PS boards I have, both main and filament, spec like this:
Filament supply:
Input = 7.5 - 12VAC
Output = 1.2 - 12.6 VDC @ up to 1.5A
Maker says use trafo secondary 12V 2A
That's per board... I have 2.
Maybe I should use just 1 board? (to supply both buffer boards)
The main supply board (I keep saying main rather than HT since its not very high... lol):
"Low noise LT1083CP dual rail power supply"
Output: +/- 3 - 30VDC up to 1A continuous.
Input: 12-30VAC x2
BTW, I thought that torroids had better inherent regulation??
Hello PRR...
But as mentioned.... I'd rather stay with what I have for now... though please DO help me to understand, so as to know what to change later on, and WHY!!
(how to re-design the project in a more nearly ideal way)
Makes sense when you don't need a lot of gain IMHO.
Its why I want to use 30/30VDC though.... brings you nearer to optimum operating point...
Interesting discussion on 6922 operating points here incidentally:
Augustica.com • View topic - 6922 Operating Point Limitations
Augustica.com • View topic - 6922 Operating Point Limitations
RE: Odd tubes...
Well, I've never wanted to.... never wanted to use other than the normal western varients...
BUT... those are getting expensive.. even the current production.. "NOS" classics impossible now, and word is that certain of the russian tubes sound very good for cheap. Going forward, gotta plan for the future/tube changes.
This becomes particularly important as everyone seems to agree that tube type becomes important in these low voltage applications; that the sound of a particular tube type "comes across" more, whereas much less when they're "run hot".
Just a plug-in replacement with the russians and no component/board changes (for around double filiament draw) becomes a big convenience.
I'd at least like to hear for myself how they sound, compared to current JJ/EH/Sovtek etc...
If I had to make circuit changes, I'd probably never bother... as a skeptic myself.
So, given what I already have... I'm still wondering... what transformer??
Or at least, what set of specs to look for...
(Hopefully my clarifications are helpful in narrowing it down?)
Have a great weekend guys!
Thanks again!
Thanks for the replies!!
I think I might have committed the error of "solid state thinking".
DF..
Thank you for teaching me the upper limit of overengineering (as a purposeful approach).
That most of the equipment I've heard and liked best used overbuilt PS is probably what led me there...
Hello Andrew.... thank you as usual for your input!
Incidentally, this is for the project mentioned in that other thread..
A valve buffer before and after attenuator, as "linestage".
2 x 6dj8/6922 tubes per board x 2 boards... 4 tubes total.
(2 stereo channel, of course)
They call for 12v for the filaments, per board.
I should probably add, at this point....
I have boards for both the main supply and the fillament supply. (regulated dc output).
This is my first "scratch build" (obviously lol) and there's nothing like the learning curve!!!
Its not JUST a learning project though... there's also a need for a certain piece of equipment in my life...
In fact, its probably a 60/40 or maybe even a 70/30 proposition at this point..
IE: the need to get it done!
(lest a box of parts lay unfinished in back of a closet for years due to indecision etc).
I'd rather proceed with what I have so far, than start from scratch at this point...
Part of the learning for me, will be to re-do/improve certain portions.. after determining what worked/didn't/could be improved upon.
It'll take long enough/be hard enough just to get something/anything up and running..
There's all the other aspects to contend with too.. physical assembly, soldering, layout, etc. etc.
I'd rather do that, then think... re-think certain aspects afterwards (possibly while listening to it).
A different/better valve buffer configuration for instance, possibly even p2p wired (as that other fellow mentioned)....
A more ideal supply for it...
All of those things take time... which I'd rather spend later and go back to re-do/improve aspects...
I'd rather not have to replace the transformer(s) later on though!
The PS boards I have, both main and filament, spec like this:
Filament supply:
Input = 7.5 - 12VAC
Output = 1.2 - 12.6 VDC @ up to 1.5A
Maker says use trafo secondary 12V 2A
That's per board... I have 2.
Maybe I should use just 1 board? (to supply both buffer boards)
The main supply board (I keep saying main rather than HT since its not very high... lol):
"Low noise LT1083CP dual rail power supply"
Output: +/- 3 - 30VDC up to 1A continuous.
Input: 12-30VAC x2
BTW, I thought that torroids had better inherent regulation??
Hello PRR...
LOL!
Ah! I'd wondered...
Wondered about that too!
But as mentioned.... I'd rather stay with what I have for now... though please DO help me to understand, so as to know what to change later on, and WHY!!
(how to re-design the project in a more nearly ideal way)
I kinda like the low voltage (plate starved do they call it?) approach..
Makes sense when you don't need a lot of gain IMHO.
Its why I want to use 30/30VDC though.... brings you nearer to optimum operating point...
Interesting discussion on 6922 operating points here incidentally:
Augustica.com • View topic - 6922 Operating Point Limitations
Augustica.com • View topic - 6922 Operating Point Limitations
RE: Odd tubes...
Well, I've never wanted to.... never wanted to use other than the normal western varients...
BUT... those are getting expensive.. even the current production.. "NOS" classics impossible now, and word is that certain of the russian tubes sound very good for cheap. Going forward, gotta plan for the future/tube changes.
This becomes particularly important as everyone seems to agree that tube type becomes important in these low voltage applications; that the sound of a particular tube type "comes across" more, whereas much less when they're "run hot".
Just a plug-in replacement with the russians and no component/board changes (for around double filiament draw) becomes a big convenience.
I'd at least like to hear for myself how they sound, compared to current JJ/EH/Sovtek etc...
If I had to make circuit changes, I'd probably never bother... as a skeptic myself.
So, given what I already have... I'm still wondering... what transformer??
Or at least, what set of specs to look for...
(Hopefully my clarifications are helpful in narrowing it down?)
Have a great weekend guys!
Thanks again!
Last edited:
- Status
- Not open for further replies.