Given a power supply of a specific B+ Voltage, and B+ Average Load Current (single ended output stage), and for Push Pull - both the average, and the peak load current:
A CLC power supply filter is less demanding on the Choke, True.
But a CLC power supply filter has Poorer Load Voltage Regulation.
And a CLC power supply filter has more Secondary Heating (and Primary Heating too).
The secondary voltage requirements for a capacitor input filter are less than for a choke input filter.
Requires cheap capacitors and a cheap choke.
An LCRC power supply filter is less demanding on the Power Transformer, in terms of Secondary Heating (and Primary Heating too).
An LCRC power supply has Better Load Voltage Regulation.
The secondary voltage requirements for a choke input filter are more than for a capacitor input filter.
Requires a big, space hogging, heavy, expensive choke.
You make your design tradeoffs, and you live with them.
A CLC power supply filter is less demanding on the Choke, True.
But a CLC power supply filter has Poorer Load Voltage Regulation.
And a CLC power supply filter has more Secondary Heating (and Primary Heating too).
The secondary voltage requirements for a capacitor input filter are less than for a choke input filter.
Requires cheap capacitors and a cheap choke.
An LCRC power supply filter is less demanding on the Power Transformer, in terms of Secondary Heating (and Primary Heating too).
An LCRC power supply has Better Load Voltage Regulation.
The secondary voltage requirements for a choke input filter are more than for a capacitor input filter.
Requires a big, space hogging, heavy, expensive choke.
You make your design tradeoffs, and you live with them.
Amazing what can be achieved with just the right string of passive components; VR!.An LCRC power supply has Better Load Voltage Regulation.
I knew this 'swinging' - whatever that actually means with the physics going on - choke would be relatable to the single ended, versus p-p OPTs. Maybe it means "swinging" about a quiescent DC current, without touching saturation.
If the ripple is the same regulation is not very important ... the swinging choke was surely not invented for audio amplifiers but stuff that would really need regulation .
“Swinging” just means the actual inductance varies — smoothly. At low currents the required critical inductance for a full 180 degree conduction angle is greater. As current increases the amount of inductance required reduces. The “swinging” choke just maintains enough inductance over the full range of load current expected. If it’s sized correctly for its job.
Any iron cored choke will have its inductance vary with load current - but many are designed to give as high as possible an inductance till the current gets too high and they just saturate. The change in inductance is rather abrupt and not as useful for filtering a wide range of current with the minimum of iron. Very useful if the choke is required to be linear, as plate loads, transformer primaries, or audio-band filtering such as in speaker crossovers.
Any iron cored choke will have its inductance vary with load current - but many are designed to give as high as possible an inductance till the current gets too high and they just saturate. The change in inductance is rather abrupt and not as useful for filtering a wide range of current with the minimum of iron. Very useful if the choke is required to be linear, as plate loads, transformer primaries, or audio-band filtering such as in speaker crossovers.
B+ Voltage regulation is never important . . . Unless:
You want your output stage to perform as you planned, and as you wish, and at optimum.
LCRC and CLC can easily be designed to have the same ripple voltage.
It is more difficult to design them to have the same voltage regulation.
I bet if anybody fully researches the Early Western Electric push pull audio amplifiers, they will find an LCRC B+ power supply in some of them.
Then check Ampex's 807 push pull (or parallel 807s push pull), and a few other audio amplifier manufacturers B+ filter configuration.
I see some amplifiers that address the ripple and the regulation issue together . . .
They use CLC, and then they follow that with a Regulator circuit.
I used to work on Collins Commercial RF Linear Amplifiers, 10kW rms, and 22.5kW rms (45kW PEP).
Just my opinions and my observations.
You want your output stage to perform as you planned, and as you wish, and at optimum.
LCRC and CLC can easily be designed to have the same ripple voltage.
It is more difficult to design them to have the same voltage regulation.
I bet if anybody fully researches the Early Western Electric push pull audio amplifiers, they will find an LCRC B+ power supply in some of them.
Then check Ampex's 807 push pull (or parallel 807s push pull), and a few other audio amplifier manufacturers B+ filter configuration.
I see some amplifiers that address the ripple and the regulation issue together . . .
They use CLC, and then they follow that with a Regulator circuit.
I used to work on Collins Commercial RF Linear Amplifiers, 10kW rms, and 22.5kW rms (45kW PEP).
Just my opinions and my observations.
Last edited:
I know those "Sola" and other AC line regulators have ferroresonant transformers, where they have an extra winding with a capacitor to provide the resonance. If one made a ferroresonant inductor, does that "go" in the direction of swinging - or does it do the opposite?
Hi jjasniew,
Ferroresonant transformers are noisy as heck, and they run hot. You can hear them buzzing across the room without any difficulty. They do have excellent regulation but you trade off efficiency and noise levels (plus heat!).
Hi 6A3sUMMER,
At high power levels you are correct. But when we are talking a normal type of home amplifier you do not normally see swinging chokes as they are more electrically noisy. Yep, the transition to saturated core is kind of sudden and that is reflected in the power supply. However a push-pull stage isn't that sensitive to B+ and you design for full power operation. I normally won't touch a class A single ended design, those were proved inferior in the 1920's for ever reason you can imagine. Some people like the sound - so okay, that's fine. Those would, I think, be very sensitive to the supply noise from a swinging choke even with ahum-bucking winding in the output transformer. Don't forget that we saw parallel field coils with some major manufacturers as well. That doesn't mean they were the right solution.
Engineering always progressed back then, and those methods were left behind in normal low power applications for very good reasons. Only new engineers refry those old ideas for market share and spin a story about their choices.
The 807 is a high voltage 6L6GC, a 7581A is my favorite, or a 6CA7 in PP circuits.
Ferroresonant transformers are noisy as heck, and they run hot. You can hear them buzzing across the room without any difficulty. They do have excellent regulation but you trade off efficiency and noise levels (plus heat!).
Hi 6A3sUMMER,
At high power levels you are correct. But when we are talking a normal type of home amplifier you do not normally see swinging chokes as they are more electrically noisy. Yep, the transition to saturated core is kind of sudden and that is reflected in the power supply. However a push-pull stage isn't that sensitive to B+ and you design for full power operation. I normally won't touch a class A single ended design, those were proved inferior in the 1920's for ever reason you can imagine. Some people like the sound - so okay, that's fine. Those would, I think, be very sensitive to the supply noise from a swinging choke even with ahum-bucking winding in the output transformer. Don't forget that we saw parallel field coils with some major manufacturers as well. That doesn't mean they were the right solution.
Engineering always progressed back then, and those methods were left behind in normal low power applications for very good reasons. Only new engineers refry those old ideas for market share and spin a story about their choices.
The 807 is a high voltage 6L6GC, a 7581A is my favorite, or a 6CA7 in PP circuits.
Anatech, why do you proffer that " But when we are talking a normal type of home amplifier you do not normally see swinging chokes as they are more electrically noisy. Yep, the transition to saturated core is kind of sudden and that is reflected in the power supply. "
Do you relate 'electrically noisy' to the effective inductance being lower when the dynamic choke current swings higher during each portion of a mains cycle, and that is the reason for swinging chokes not being in common use in home amplifiers, or is it just the operation of a high ac voltage presented to the input of a choke input filter?
Do you also characterise 'the transition to saturated core is kind of sudden' from some form of measurement, or perhaps an interpretation of one of the gapped curves in the log plot in post #49 ?
Do you relate 'electrically noisy' to the effective inductance being lower when the dynamic choke current swings higher during each portion of a mains cycle, and that is the reason for swinging chokes not being in common use in home amplifiers, or is it just the operation of a high ac voltage presented to the input of a choke input filter?
Do you also characterise 'the transition to saturated core is kind of sudden' from some form of measurement, or perhaps an interpretation of one of the gapped curves in the log plot in post #49 ?
Hi trobbins,
Seen them, worked with them. What more do you want from me?
If you want a quiet DC supply, you do want the choke current to remain within limits so the choke does not saturate. It is the change in current in the inductor that shows up on the supply.
I did not reference that post, I'm going off practical experience here.
Seen them, worked with them. What more do you want from me?
If you want a quiet DC supply, you do want the choke current to remain within limits so the choke does not saturate. It is the change in current in the inductor that shows up on the supply.
I did not reference that post, I'm going off practical experience here.
anatech,
I agree: Experience
Try using a Hammond 193H 5H 200mA in a choke input B+ filter to run an 80mA load (I never claimed the 193H is a swinging choke).
I bet it is Not a swinging choke, but it works very well for me. Quiet, efficient, and cool.
Then, using the 193H, check out your amplifiers power factor, it will be about 0.9. Try and get a power factor higher than 0.65 with a CLC B+ filter
(especially if you are using solid state rectifiers, to save power versus tube rectifiers).
Then, measure the cool power transformer that is loaded with the average current that is the characteristic of a choke input filter.
To compare, check out the hot power transformer that is loaded with a CLC filter (the consequential peak I-squared R secondary heat losses).
And get a load of that power factor that can not pass as a commercial product in the EU, and many other countries.
Oh . . . perhaps the 'power police' are not watching the power factors of audio amplifiers.
I agree: Experience
Try using a Hammond 193H 5H 200mA in a choke input B+ filter to run an 80mA load (I never claimed the 193H is a swinging choke).
I bet it is Not a swinging choke, but it works very well for me. Quiet, efficient, and cool.
Then, using the 193H, check out your amplifiers power factor, it will be about 0.9. Try and get a power factor higher than 0.65 with a CLC B+ filter
(especially if you are using solid state rectifiers, to save power versus tube rectifiers).
Then, measure the cool power transformer that is loaded with the average current that is the characteristic of a choke input filter.
To compare, check out the hot power transformer that is loaded with a CLC filter (the consequential peak I-squared R secondary heat losses).
And get a load of that power factor that can not pass as a commercial product in the EU, and many other countries.
Oh . . . perhaps the 'power police' are not watching the power factors of audio amplifiers.
Anatech, are you saying that operating a swinging choke within its limits will provide a quiet dc supply, and perhaps implying that the majority of others weren't operating swinging chokes within limits, and that then is why swinging chokes are not normally seen in home amplifiers ?
I've been looking for manufacturer specs for swinging chokes, and they are pretty much non-existent or don't seem to provide relevant test conditions. Very few magazines or transformer maker catalogs even mention them for choke input power supply applications (given they can be used for other applications). Stancor have a group of 11 swinging choke parts in their 1954 catalog 'for input section of dc power supplies', and test inductance at 100% and 10% of rated dc current, but testing only applies 10Vac 60Hz. So far I have only found one other datasheet from a manufacturer with that level of spec detail, and it too only tests at 10Vac. If anyone comes across any data/specs then I'd welcome seeing it.
As Suncalc pointed out in post #49, a swinging choke requires design that specifically sets the gap to suit the span of dc load current, and the inductance range to suit the choke-input application. To date, the only reference to go through that has been in the QST mag back in 1932. This would all suggest that competent swinging choke use for diy would be beyond many, and relegated to the 'too hard' basket (especially nowadays when a capacitor input filter and ss post regulator meet most needs).
https://dalmura.com.au/static/Choke QST-1932-02.pdf
https://dalmura.com.au/static/Choke QST-1932-03.pdf
https://dalmura.com.au/static/Choke QST-1932-04.pdf
I've been looking for manufacturer specs for swinging chokes, and they are pretty much non-existent or don't seem to provide relevant test conditions. Very few magazines or transformer maker catalogs even mention them for choke input power supply applications (given they can be used for other applications). Stancor have a group of 11 swinging choke parts in their 1954 catalog 'for input section of dc power supplies', and test inductance at 100% and 10% of rated dc current, but testing only applies 10Vac 60Hz. So far I have only found one other datasheet from a manufacturer with that level of spec detail, and it too only tests at 10Vac. If anyone comes across any data/specs then I'd welcome seeing it.
As Suncalc pointed out in post #49, a swinging choke requires design that specifically sets the gap to suit the span of dc load current, and the inductance range to suit the choke-input application. To date, the only reference to go through that has been in the QST mag back in 1932. This would all suggest that competent swinging choke use for diy would be beyond many, and relegated to the 'too hard' basket (especially nowadays when a capacitor input filter and ss post regulator meet most needs).
https://dalmura.com.au/static/Choke QST-1932-02.pdf
https://dalmura.com.au/static/Choke QST-1932-03.pdf
https://dalmura.com.au/static/Choke QST-1932-04.pdf
Boy, these chokes, and swinging chokes, keep reminding me how a single ended output transformer has to work,
Especially when someone insists on turning the amplifier volume up to both hard positive clipping, and hard negative clipping.
Example:
A single ended 300B that has 60mA quiescent curent through the air-gapped output transformer primary.
The user turns up the volume until the 300B current goes into cut-off (0 mA), and then goes to 120mA in the other direction.
Sounds terrible, Right?
Just as it should when you abuse the amplifier.
Now you know why I say to turn the volume down.
Not loud enough for you, then get an amplifier that has at least 2x the power (+3dB).
Especially when someone insists on turning the amplifier volume up to both hard positive clipping, and hard negative clipping.
Example:
A single ended 300B that has 60mA quiescent curent through the air-gapped output transformer primary.
The user turns up the volume until the 300B current goes into cut-off (0 mA), and then goes to 120mA in the other direction.
Sounds terrible, Right?
Just as it should when you abuse the amplifier.
Now you know why I say to turn the volume down.
Not loud enough for you, then get an amplifier that has at least 2x the power (+3dB).
Hi trobbins,
I am not here to defend or explain what successful engineers the world over have settled on as best design practices. Engineering is all about balance and balancing cost and performance, and simply finding a better way to do something. I am not experienced in the world 6A3sUMMER worked in, so he would be a good authority on swinging chokes. They are not generally used in my neck of the engineering woods because of what is important at lower power levels with low signal levels.
I'll give you an example. We went from using several pounds of chokes to smooth ripple, to using C-R-C-R-C-R-C networks to reduce supply ripple to low levels. Then we developed electronic regulators which went from tubes to solid state in construction. We have further refined these to further reduce noise and increase stability. Today we use a mix depending on situation and what delivers the required performance at the desired cost with the desired reliability. Engineering.
Early in time, capacitors were not that robust and tough to make (= expensive). They were not as reliable as something like a coil of wire. So early design for power supplies went with a mix of cost, performance and reliability. That meant low value main filter capacitors and swinging chokes. In high power/voltage electronics I would not be terribly surprised to find out early issues with filter capacitors remain. High voltage capacitors are not easy to make, and they are large and very expensive as a result. Also, ambient temperature plays a huge factor. Capacitors do not like high temperatures.
Now while I do not design anything with swinging chokes, I am familiar with some of the engineering behind them. But that is it, I would never pretend I could advise anyone on how to design with them. There are others here far better suited to that job. All I can do is suggest a general path to follow with applications I normally work with. However make no mistake. I am not going to design anything for anyone here, I am a working technician who gets paid for engineering time and I am busy. All I can do is try to help. What I do not have any time for are games when I could be helping someone else.
I believe I have been extremely clear in what I have said and right now you're trying to put words in my mouth. I do not appreciate the games. It doesn't take a lot of intellect to poke away at what people say. Why not try something different? Let's see you state what you believe is true, backed up with direct experience or accepted industry or engineering practice.Anatech, are you saying that operating a swinging choke within its limits will provide a quiet dc supply, and perhaps implying that the majority of others weren't operating swinging chokes within limits, and that then is why swinging chokes are not normally seen in home amplifiers ?
I am not here to defend or explain what successful engineers the world over have settled on as best design practices. Engineering is all about balance and balancing cost and performance, and simply finding a better way to do something. I am not experienced in the world 6A3sUMMER worked in, so he would be a good authority on swinging chokes. They are not generally used in my neck of the engineering woods because of what is important at lower power levels with low signal levels.
I'll give you an example. We went from using several pounds of chokes to smooth ripple, to using C-R-C-R-C-R-C networks to reduce supply ripple to low levels. Then we developed electronic regulators which went from tubes to solid state in construction. We have further refined these to further reduce noise and increase stability. Today we use a mix depending on situation and what delivers the required performance at the desired cost with the desired reliability. Engineering.
Early in time, capacitors were not that robust and tough to make (= expensive). They were not as reliable as something like a coil of wire. So early design for power supplies went with a mix of cost, performance and reliability. That meant low value main filter capacitors and swinging chokes. In high power/voltage electronics I would not be terribly surprised to find out early issues with filter capacitors remain. High voltage capacitors are not easy to make, and they are large and very expensive as a result. Also, ambient temperature plays a huge factor. Capacitors do not like high temperatures.
Now while I do not design anything with swinging chokes, I am familiar with some of the engineering behind them. But that is it, I would never pretend I could advise anyone on how to design with them. There are others here far better suited to that job. All I can do is suggest a general path to follow with applications I normally work with. However make no mistake. I am not going to design anything for anyone here, I am a working technician who gets paid for engineering time and I am busy. All I can do is try to help. What I do not have any time for are games when I could be helping someone else.
Anatech, to be clear, it seems that no poster here has knowingly used, or tested, or even tried to purchase a swinging choke. Your post #68 indicated you had experience with them, but it appears not. Many of us come to this forum with a range of choke experience, and I also have plenty of experience with chokes, whether that is water-cooled for thousands of amps, to switchmode applications, to ..., but I've never knowingly used or had the opportunity to assess a swinging choke even though I had seen passing references to them over the decades.
So I don't think it is valid to tag swinging chokes as 'electrically noisy'. In post #54, Suncalc summarises operation that can cause problems when chokes are misapplied for choke-input filter applications, and those operational boundaries apply for any choke (swinging or not).
So I don't think it is valid to tag swinging chokes as 'electrically noisy'. In post #54, Suncalc summarises operation that can cause problems when chokes are misapplied for choke-input filter applications, and those operational boundaries apply for any choke (swinging or not).
Okay, enough playing.
I have dealt with them enough to have a little experience. Not enough to suggest how to design with them. I was clear about this.
Whatever your conclusions are are your own. I will not be dragged into a debate or useless conversation with someone who isn't advancing any knowledge. If you have something you actually know about, please feel free to elucidate us. Otherwise, you're only making noise. Stop trying to tie what I said to other posts, or suggesting I said something I did not.
Now, the swinging choke itself is not noisy and not once did I say it was. You really need to read what is posted more carefully. However when the inductance value changes it will change how much current flows through it rapidly. I am reasonably sure you can figure out on your own what this may mean. Right?
I have dealt with them enough to have a little experience. Not enough to suggest how to design with them. I was clear about this.
Whatever your conclusions are are your own. I will not be dragged into a debate or useless conversation with someone who isn't advancing any knowledge. If you have something you actually know about, please feel free to elucidate us. Otherwise, you're only making noise. Stop trying to tie what I said to other posts, or suggesting I said something I did not.
Now, the swinging choke itself is not noisy and not once did I say it was. You really need to read what is posted more carefully. However when the inductance value changes it will change how much current flows through it rapidly. I am reasonably sure you can figure out on your own what this may mean. Right?
anatech,
By any chance were the swinging chokes you have experienced, driven by Mercury Rectifiers.
Swinging chokes plus Mercury Rectifiers ARE Noisy!
I gave up using (playing with) Mercury as a kid.
It has not affected, affected, me, me, yet, yet. (so far as I know or that I can remember).
Everybody, let's just have fun designing, building, and listening to our tube amplifiers.
By any chance were the swinging chokes you have experienced, driven by Mercury Rectifiers.
Swinging chokes plus Mercury Rectifiers ARE Noisy!
I gave up using (playing with) Mercury as a kid.
It has not affected, affected, me, me, yet, yet. (so far as I know or that I can remember).
Everybody, let's just have fun designing, building, and listening to our tube amplifiers.
lol!
Nope! That would be a combination all right! I am very familiar with Mercury rectifiers, the 83 especially!
The equipment I have worked on used early rectifiers, single element each in a pair usually. I rebuild radios and equipment from the 1920's on up.
Yup, share knowledge and assist where we can ...
Nope! That would be a combination all right! I am very familiar with Mercury rectifiers, the 83 especially!
The equipment I have worked on used early rectifiers, single element each in a pair usually. I rebuild radios and equipment from the 1920's on up.
Yup, share knowledge and assist where we can ...
- Home
- Amplifiers
- Tubes / Valves
- Rectifier tube glows, sparks, then behaves. What gives? (Video inside)