Over the last few years I have been collecting transformers from Sony TC-500a tape decks. I recently saw a very simple schematic for a transformer input power amplifier and wondered if there was a reason I would not want to create something like this using TC500a OPT for both the input and output transformers. With a 5k primary and 8 Ohm secondary, the voltage gain of 25 seemed to be approximately right, especially if driven using a balanced input. I have a matched pair of 45 tubes for the outputs. My knowledge of tube amp design is very close to zero, but it appears this design does not include NFB. Would there be a way to incorporate it? Or at least provide for the option. Or would it not be needed?
I would appreciate any suggestions.
Thanks.
I would appreciate any suggestions.
Thanks.
1. The 8 Ohm winding has a very low impedance, and a very low inductance (it is 1/625 of the 5k winding inductance).
Most CD players, preamps, etc. will not like to drive the 8 Ohm secondary of that transformer.
No free lunch here.
2. A single ended 45 tube does not really need negative feedback.
I built a simple 2 stage stereo 45 amp for use at work. A 45 can put out up to about 2 watts; my amplifier only did 1.5 watts per channel, partly because of the low gain of the 12AU7 that could not put out as much peak voltage as the 45 bias voltage, and the slightly low B+ voltage I had to work with. But there was plenty of power for the work environment. I used Usher S-520 two way speakers.
One night after 5:00 when almost everyone went home, I turned it up. A lady came from several aisles over to tell me how good the music sounded (there were really high sound absorber walls between the aisles, the sound had to travel up and over.
You had a nice idea in post # 1, but I recommend you try another 45 circuit than the one you posted above.
Most CD players, preamps, etc. will not like to drive the 8 Ohm secondary of that transformer.
No free lunch here.
2. A single ended 45 tube does not really need negative feedback.
I built a simple 2 stage stereo 45 amp for use at work. A 45 can put out up to about 2 watts; my amplifier only did 1.5 watts per channel, partly because of the low gain of the 12AU7 that could not put out as much peak voltage as the 45 bias voltage, and the slightly low B+ voltage I had to work with. But there was plenty of power for the work environment. I used Usher S-520 two way speakers.
One night after 5:00 when almost everyone went home, I turned it up. A lady came from several aisles over to tell me how good the music sounded (there were really high sound absorber walls between the aisles, the sound had to travel up and over.
You had a nice idea in post # 1, but I recommend you try another 45 circuit than the one you posted above.
Last edited:
Assuming your preamplifier can drive 8 Ohm, there is an additional toll. Using the Sony transformer as input will multiply the Miller capacitance provided by the 45 by 625 times (25x25). The 45 will have about 30 pF capacitance and so that means the "preamplifier" or DAC will need to be able to drive almost 19 nF up to 20KHz. It's a very demanding load. Most headphone amps cannot do that.
Besides, the TC-500a transformer has a 600 Ohm tap too. Driving 600R it's a whole different story and the capacitance will be reduced to 2 nF. For 1.5W output, running the 45 at 250V/34mA, about 6Vpeak (4.25 Vrms) drive will be needed. Still, it's not a load for an ordinary tube preamp. Maybe a White Cathode follower but it would not be made with tiny tubes....
The AD797 OP-amp, I remember is able to drive capacitive loads up to 5 nF with good closed-loop gain. Feedback around the 45 is not necessary (most 45 amps do not use it) and would be complicated with 2 transformers in the loop and minimal gain at disposal.
It seems that most preamps and CD players can not drive that transformer. So you need another tube stage between the signal source, and that transformer.
Instead, just use a tube to drive the 45, and do not use the transformer.
Lots of tubes can provide a gain of 25, and can be used instead of the transformer.
Save the other 2 transformers for another project.
"You should keep things as simple as possible, but no simpler" - Albert Einstein
Instead, just use a tube to drive the 45, and do not use the transformer.
Lots of tubes can provide a gain of 25, and can be used instead of the transformer.
Save the other 2 transformers for another project.
"You should keep things as simple as possible, but no simpler" - Albert Einstein
Thank you very much for the very useful comments. Clearly I had not realized a number of aspects. Clearly, the low impedance is an issue. I guess I was wrong in thinking it would be possible to terminate the 5k windings into something more like 250k and raise the impedance on the 8 ohm windings to more like 400ohms. That would not address the capacitance issues, of course.
45, if I understand you correctly, if I instead used the 600 0hm windings to drive the tube I would have a gain of about 3 and an impedance ratio of more like 9. Using a balanced input would provide adequate voltage ( with the gain of 3) to drive the 45?
45, if I understand you correctly, if I instead used the 600 0hm windings to drive the tube I would have a gain of about 3 and an impedance ratio of more like 9. Using a balanced input would provide adequate voltage ( with the gain of 3) to drive the 45?
The 600 Ohm winding inductance would be 1/8.3 of the 5k winding inductance.
If the 5k inductance is 20 Henry, the 600 Ohm winding would only be 2.4 Henry.
2.4 Henry is only 1,500 Ohms at 100Hz, and even less at low bass frequencies.
A lot of CD players are rated to work with loads of 10k Ohms or higher.
Many CD players maximum output is 2.1Vrms, that is 3V peak.
The transformer voltage gain from 600 Ohms to 5k Ohms is about 2.9.
3V peak from the CD player x transformer gain of 2.9 = 8.7V peak, far less than any practical 45 bias voltage.
Why did you mention a balanced input?
Do you have a CD player or preamp that has a balanced output?
Or, are you afraid of interference when using an unbalanced CD player or unbalanced preamp output?
Often, the reasons behind decisions are almost as important as the decisions themselves.
If the 5k inductance is 20 Henry, the 600 Ohm winding would only be 2.4 Henry.
2.4 Henry is only 1,500 Ohms at 100Hz, and even less at low bass frequencies.
A lot of CD players are rated to work with loads of 10k Ohms or higher.
Many CD players maximum output is 2.1Vrms, that is 3V peak.
The transformer voltage gain from 600 Ohms to 5k Ohms is about 2.9.
3V peak from the CD player x transformer gain of 2.9 = 8.7V peak, far less than any practical 45 bias voltage.
Why did you mention a balanced input?
Do you have a CD player or preamp that has a balanced output?
Or, are you afraid of interference when using an unbalanced CD player or unbalanced preamp output?
Often, the reasons behind decisions are almost as important as the decisions themselves.
Yes, balanced CD player, balanced dac, balanced preamp. I always try and use balanced connections if possible.
I am clearly confused about how transformers work. I always envision them as a combination of inductance and the load resistance. I guess I am completely unaware of effects beyond the ratio of windings. When a 3:1 transformer is attached to a 100k load, the impedance on the other winding is 11k or 900k.
I am clearly confused about how transformers work. I always envision them as a combination of inductance and the load resistance. I guess I am completely unaware of effects beyond the ratio of windings. When a 3:1 transformer is attached to a 100k load, the impedance on the other winding is 11k or 900k.
And sorry, I am not trying to disagree, just understand. It seems the original idea was a bad one.
I guess I missed the actual answer to your question about balanced connections. The output of the dac I was thinking about using is 8Vrms from the balanced output. With a gain of 3 and transconductance values of about 2000 micromhos, I thought that about 67V, peak to peak, would be more than enough for 34mA of cathode current.
Does your DAC output use one RCA phono connector per channel? (1 RCA left and 1 right). That is not balanced.
Or, does your DAC output use one XLR connector per channel? (1 XLR left and 1 XLR right). That usually is balanced.
Or, does you DAC use a different kind of output connector per channel than the above connectors? What kind of connector?
Generally, an output transformer, like from the Sony TC-500a was not meant to be used in a balanced application.
I think an EL84 or similar was used in pentode mode to drive the 5k primary, with the B+ end at AC ground; and that there was global negative feedback from the 8 Ohm secondary.
That application requires one end of the secondary to be grounded. That is not a balanced application.
If your DAC has a balanced output, has lots of voltage output, and has low impedance output, it might be able to drive a specialized balanced input to grid transformer.
Or, does your DAC output use one XLR connector per channel? (1 XLR left and 1 XLR right). That usually is balanced.
Or, does you DAC use a different kind of output connector per channel than the above connectors? What kind of connector?
Generally, an output transformer, like from the Sony TC-500a was not meant to be used in a balanced application.
I think an EL84 or similar was used in pentode mode to drive the 5k primary, with the B+ end at AC ground; and that there was global negative feedback from the 8 Ohm secondary.
That application requires one end of the secondary to be grounded. That is not a balanced application.
If your DAC has a balanced output, has lots of voltage output, and has low impedance output, it might be able to drive a specialized balanced input to grid transformer.
Last edited:
As already pointed out, you have 8.33 voltage gain with the input and somewhere between 2-2.5 H inductance for the 600R tap. The input capacitance will be about 2 nF. This is the basic info you need to know to build the driver.
The 45 running at 250V/34mA will need about 48.5V peak for full output of 1.5W. Because you have 8.33 step up the input voltage becomes 48.5/8.33=5.82V peak or 4.12V rms.
At the low end you need to provide 4.12V into 2H (worst case) down to 20Hz, in theory. I don't know the exact specs of the Sony output transformer, in particular if it is capable of 1.5W at 20Hz without significant distortion. 1.5W seems stupidly low but that transformer is tiny too.. Maybe 30Hz is more like the lower limit which is very good already for a SE amplifier without feedback to reproduce music rather than sounds.. 30Hz is also the limit of excellent and expensive transformers too. The main reason for this is that on the same core one will need less turns and will get better efficiency. Lower series resistance is more beneficial to the sound than full power down to 20Hz. Of course you can get the same efficiency and go down to 20Hz but size and cost will go up....full power at 30Hz does not mean that it won't go lower, it will only be capable of less power.
To know how much current the driver stage will need to deliver into 2H for 4.12Vrms:
Irms= Vrms/(4.44*L*f)
The coefficient 4.44 is relative to sinusoidal regime. In your case you will need 23.2 mA rms for 20Hz and 15.46 mA for 30Hz.
The AD797 I was suggesting before can do 13V peak into 600R (with +/-15V supply). This means 21.67 mA peak or 15.32 mA rms and that's what you need in the worst case to go down to 30Hz without distortion from the driver.
The circuit to drive the capacitive load with good closed loop gain, say between 5 and 10, is shown in the datasheet. You have a SS front-end using OP-amps and other "sand" I don't see why you should no use this "easy" and effective solution.
The circuit for capacitive loads is at pag.14 of the datasheet
https://www.analog.com/media/en/technical-documentation/data-sheets/AD797.pdf
Yes, XLR connections. The outputs of the dac are proper balanced signals.
Though, certainly, in their original application one side of the 600 ohm windings was attached to ground, because of galvanic isolation between the windings I do not think it matters if the winding is attached to the +/- signals of the balanced input.
In 45's message it was indicated that only 6V peak, 12V peak to peak, would be necessary to drive the 45. Is that consistent with your experience?
But I have come to realize why this simple approach is not used. I think I neglected to think about the power the source would need to supply, not just the voltage. I really appreciate your feedback.
Though, certainly, in their original application one side of the 600 ohm windings was attached to ground, because of galvanic isolation between the windings I do not think it matters if the winding is attached to the +/- signals of the balanced input.
In 45's message it was indicated that only 6V peak, 12V peak to peak, would be necessary to drive the 45. Is that consistent with your experience?
But I have come to realize why this simple approach is not used. I think I neglected to think about the power the source would need to supply, not just the voltage. I really appreciate your feedback.