Greetings, friends. I've been interested in building an amp choke-input power supply for some time. The pros and cons are a bit nebulous, the components required seem to be subject to mythical requirements (what's a swinging choke?), and neither Merlin nor Morgan devote much ink to them. So I come to you.
The RH84 webpage offers 2 designs for a power supply: the first, a traditional CLC
straightforward. He's nice enough to include voltage ranges for 5R4 and 5AR4 rectifiers.
It's the second PS design that intrigues me, a Choke-input LCLC supply:
And here's where I'm seeking the wisdom. The image above lists the first criteria for the first Choke: 10H, 100mA, Enclosed or Potted, something like the Hammond 193G. The second choke can be a cheaper, open bracket Hammond 158M or 159P, ok.
In the text below the image, the author states that, "Particular attention should be given to the choice of the input choke – it will be prone to vibrations, and thus needs to be a potted or enclosed unit, as stiff as possible. Care should be given as well to the positioning of this choke, since it will have an important magnetic field due to the difficult task."
What is the ideal position for this Choke? Many tube amplifiers I've seen employ a layout where the power supply components occupy one end of the chassis, keeping any AC voltages as far from the audio components as possible:
sometimes the Choke is turned 90 to the PT:
would either of these layouts prove suitable for a Choke-input power supply? Is it worth the extra $$ in components to arrive at the same b+ ?
advice, wisdom welcome.
thanks!
w
The RH84 webpage offers 2 designs for a power supply: the first, a traditional CLC
straightforward. He's nice enough to include voltage ranges for 5R4 and 5AR4 rectifiers.
It's the second PS design that intrigues me, a Choke-input LCLC supply:
And here's where I'm seeking the wisdom. The image above lists the first criteria for the first Choke: 10H, 100mA, Enclosed or Potted, something like the Hammond 193G. The second choke can be a cheaper, open bracket Hammond 158M or 159P, ok.
In the text below the image, the author states that, "Particular attention should be given to the choice of the input choke – it will be prone to vibrations, and thus needs to be a potted or enclosed unit, as stiff as possible. Care should be given as well to the positioning of this choke, since it will have an important magnetic field due to the difficult task."
What is the ideal position for this Choke? Many tube amplifiers I've seen employ a layout where the power supply components occupy one end of the chassis, keeping any AC voltages as far from the audio components as possible:
sometimes the Choke is turned 90 to the PT:
would either of these layouts prove suitable for a Choke-input power supply? Is it worth the extra $$ in components to arrive at the same b+ ?
advice, wisdom welcome.
thanks!
w
1. Forget about the Choke to Power Transformer orientation.
Worry about the Choke to Output Transformer orientation, and the distance too.
2. Use an aluminum chassis, not a magnetic steel chassis.
3. I am not sure what the DC current of the 4 output tubes is. But I bet a 5H choke will have at least the required minimum critical inductance.
All my Hammond 5H 200mA chokes (193H) work very well for choke input B+ filters, no vibration.
But I purchased them long ago. I hope Hammond is still building them to the same quality level. The DCR is 65 Ohms.
4. I use at least 50uF right after the 5H choke.
I do not use a second choke. Instead, I use either a 100 Ohm or 180 Ohm resistor to the next capacitor, at least 100uF. That capacitor runs the output stage B+.
Then, I use a series resistor of 1k from the output stage B+, to another capacitor of at least 50uF, that Becomes the 2nd B+, it runs the driver tubes.
You might have to change the values slightly to get the B+ and low ripple you need. Typically, doubling the capacitance is what I do, to keep the ripple low.
Choke input power supplies do the following:
Require higher secondary voltage
Create a lower frequency and lower amplitude local hum ground loop (cap input filters have fast rise, high frequency, high transient current hum ground loops).
The power transformer runs cooler with choke input filters.
The rectifier tube has an easier time to drive the B+ filter.
Choke input filters have better load regulation.
Choke input filters help to make my amplifiers "sound" good
Other than that, I can not remember why I use a choke input filter.
Have fun designing, building, and listening!
Worry about the Choke to Output Transformer orientation, and the distance too.
2. Use an aluminum chassis, not a magnetic steel chassis.
3. I am not sure what the DC current of the 4 output tubes is. But I bet a 5H choke will have at least the required minimum critical inductance.
All my Hammond 5H 200mA chokes (193H) work very well for choke input B+ filters, no vibration.
But I purchased them long ago. I hope Hammond is still building them to the same quality level. The DCR is 65 Ohms.
4. I use at least 50uF right after the 5H choke.
I do not use a second choke. Instead, I use either a 100 Ohm or 180 Ohm resistor to the next capacitor, at least 100uF. That capacitor runs the output stage B+.
Then, I use a series resistor of 1k from the output stage B+, to another capacitor of at least 50uF, that Becomes the 2nd B+, it runs the driver tubes.
You might have to change the values slightly to get the B+ and low ripple you need. Typically, doubling the capacitance is what I do, to keep the ripple low.
Choke input power supplies do the following:
Require higher secondary voltage
Create a lower frequency and lower amplitude local hum ground loop (cap input filters have fast rise, high frequency, high transient current hum ground loops).
The power transformer runs cooler with choke input filters.
The rectifier tube has an easier time to drive the B+ filter.
Choke input filters have better load regulation.
Choke input filters help to make my amplifiers "sound" good
Other than that, I can not remember why I use a choke input filter.
Have fun designing, building, and listening!
Last edited:
Simulate your power supply with PSUD. For simplicity, assume that the power supply ripple at the output tube tap is stepped down by the output transformer ratio (i.e 5K:8 is 25:1). Look for less than a millivolt RMS or 3 mVp-p. The RH84 is pentode mode, but the screen connection is not filtered, so it's a triode as far as power supply rejection. Using a resistor and capacitor to feed the screen can give additional hum rejection if your power supply doesn't meet the suggested hum level. Doubling choke inductance or filter cap (not the first one!) can each give 2x (6 dB) further hum reduction. Obviously caps do it for less.
Last year I dug up an old Hammond 28mH 3A choke and tested it for choke input duty in a 10V 1A PSU (intended to feed a Coleman filament regulator into a 6B4G). That choke was made in Canada and worked just fine so I ordered four more for my 6B4G PP project. The new ones were made in China and buzzed quite loudly at the exact same AC load.
No experience, but I read here and elsewhere about a quasi-choke input (= small value cap before choke; 0.25 to 1uF). That would give you the regulation benefits of a choke input filter, reduce strain (= buzz) on choke and allow you to finetune the B+ voltage. The best of both worlds?
Klimon,
The tradeoffs of using a small capacitor in front of a choke input filter:
At 120Hz full wave rectification (60Hz power mains) . . .
A 0.1 uF capacitor has 13,300 Ohms of capacitive reactance.
Suppose the amplifier draws 50mA from B+.
50mA (0.050A).
0.050A x 13,300 Ohms = ripple of 665V peak to peak [will not actually slew that far] (665Vpp is more than a 350V-0-350V secondary which is 500V peak to ground at each secondary lead).
As you can imagine, 0.1uF will not stop a choke from vibrating at 120Hz.
A quality choke will not have very much distributed capacitance across the windings. That means it will attenuate high frequencies and transients.
But, if you need to reduce those transients for a less than ideal choke, that 0.1uF will reduce any high frequency transients, and shunt them to ground (Caution, this forms a very high frequency ground loop, with frequencies that are right in the ears most sensitive range . . . so be sure to keep the ground loop very short, and make it local, do not connect it directly to the central ground.
Instead, tie it to the negative of the filter cap that follows the choke. Then, connect a separate wire from there to the central ground.
I have used from 0.5uF and up to 4 uF to adjust the B+ voltage. I call that Pseudo Choke Input Filter. Less load regulation is the result.
0.5uF is 2660 Ohms capacitive reactance at 120Hz. At 0.050A x 2660, that will have 133V peak to peak ripple. It will actually do something to reduce the 120Hz vibration of the choke.
For 50Hz power mains, 100Hz full wave, the numbers above are 20% differnt, but that is usually not significant.
The tradeoffs of using a small capacitor in front of a choke input filter:
At 120Hz full wave rectification (60Hz power mains) . . .
A 0.1 uF capacitor has 13,300 Ohms of capacitive reactance.
Suppose the amplifier draws 50mA from B+.
50mA (0.050A).
0.050A x 13,300 Ohms = ripple of 665V peak to peak [will not actually slew that far] (665Vpp is more than a 350V-0-350V secondary which is 500V peak to ground at each secondary lead).
As you can imagine, 0.1uF will not stop a choke from vibrating at 120Hz.
A quality choke will not have very much distributed capacitance across the windings. That means it will attenuate high frequencies and transients.
But, if you need to reduce those transients for a less than ideal choke, that 0.1uF will reduce any high frequency transients, and shunt them to ground (Caution, this forms a very high frequency ground loop, with frequencies that are right in the ears most sensitive range . . . so be sure to keep the ground loop very short, and make it local, do not connect it directly to the central ground.
Instead, tie it to the negative of the filter cap that follows the choke. Then, connect a separate wire from there to the central ground.
I have used from 0.5uF and up to 4 uF to adjust the B+ voltage. I call that Pseudo Choke Input Filter. Less load regulation is the result.
0.5uF is 2660 Ohms capacitive reactance at 120Hz. At 0.050A x 2660, that will have 133V peak to peak ripple. It will actually do something to reduce the 120Hz vibration of the choke.
For 50Hz power mains, 100Hz full wave, the numbers above are 20% differnt, but that is usually not significant.
Last edited:
Done right choke input sounds awsome. Once you here there sound cap input sounds shutin,slow,muddy. Be warned they are tricky to use in that the choke must be rated for choke input duty.this includes the load current and the ripple current across the choke so they will be big and expensive.well worth it.think hashimoto LC 15 200. Beautifull sound. Looks like a audio jewel.