Hi Paolo,
Of the Hammond catalog, for a single piece, the 193C best fits your application. 20H 100ma. Input chokes are more demanding than others as significant AC current is present, sometimes exceeding the DC load. Add DC and AC currents, and then leave some margin past that to size choke. Also, soft mounts might help mechanical noise. You can ground the core with a wire instead if need be.
Peak AC current= VRMS/1115L@50hz
VRMS/1386L@60hz
VRMS is your Power Transformer secondary voltage
Of the Hammond catalog, for a single piece, the 193C best fits your application. 20H 100ma. Input chokes are more demanding than others as significant AC current is present, sometimes exceeding the DC load. Add DC and AC currents, and then leave some margin past that to size choke. Also, soft mounts might help mechanical noise. You can ground the core with a wire instead if need be.
Peak AC current= VRMS/1115L@50hz
VRMS/1386L@60hz
VRMS is your Power Transformer secondary voltage
Thanks Tweeker, very helpful 🙂
Thus you are saying I have not to be worried about the series resistance ( 180 ohm each) in my CL ( or CLCL) application, right?
I know I can decoupling the chocke , but if it is noising a lot, it is not sufficent IMO.
Cheers,
Paolo
Thus you are saying I have not to be worried about the series resistance ( 180 ohm each) in my CL ( or CLCL) application, right?
I know I can decoupling the chocke , but if it is noising a lot, it is not sufficent IMO.
Cheers,
Paolo
Oh thanks, very clear explanation.
I have tried to use the Duncan's but I am so incompetent that
it was a total failure
Maybe I have to re-try again 🙄
Cheers,
Paolo
I have tried to use the Duncan's but I am so incompetent that
it was a total failure
Maybe I have to re-try again 🙄
Cheers,
Paolo
I was planning to use a couple of monster size UTC S-38 swinging chokes, 4H-20H @ 550mA in a power supply for a pair of mono amps drawing 300V @ 280mA:
(HUGE) 1000VCT - 5AR4-GY - S38 - 400uf - S38 - 400uf > 300V/280mA class A amp
Of course the power supply is going to weigh in at about 50 kilos...but i like iron...
Would these swinging chokes be suitable for such use? - Using PSUD2 i assumed a value of 8H for the UTC's and get a nice flat result...thoughts?
(HUGE) 1000VCT - 5AR4-GY - S38 - 400uf - S38 - 400uf > 300V/280mA class A amp
Of course the power supply is going to weigh in at about 50 kilos...but i like iron...
Would these swinging chokes be suitable for such use? - Using PSUD2 i assumed a value of 8H for the UTC's and get a nice flat result...thoughts?
Continuing ...
I have recently experimented further with choke-input filters.
I managed to cure the magnetic radiation problem (post #13) by mounting the choke in a steel enclosure.
Then it might be interesting to note that one can resonate the choke impedance with a capacitor at 100 Hz (or 120 Hz) to decrease ripple by a factor of some 3 - 5. This needs to be done in situ as the inductance at the specific load current is in play. As this changes with load current, best is to do this at zero output where signal/noise is worst.
The disadvantage is that the L.C impedance will decrease at higher frequencies, thus allowing more mains noise and also rectified voltage harmonics through (the rectified voltage ripple is not a pure 100 (120) Hz sine wave. This can be cured by suitable filtering at the mains input to the amplifier and high value filter output capacitance. (A choke-input filter is usually followed by a second LC or regulator.)
Some experimentation (observing the necessary safety measures!) might yield good results here. Also note that the voltage across such a resonant LC might be high - if a scope is available, do check. Using 630V capacitors is wise.
I have recently experimented further with choke-input filters.
I managed to cure the magnetic radiation problem (post #13) by mounting the choke in a steel enclosure.
Then it might be interesting to note that one can resonate the choke impedance with a capacitor at 100 Hz (or 120 Hz) to decrease ripple by a factor of some 3 - 5. This needs to be done in situ as the inductance at the specific load current is in play. As this changes with load current, best is to do this at zero output where signal/noise is worst.
The disadvantage is that the L.C impedance will decrease at higher frequencies, thus allowing more mains noise and also rectified voltage harmonics through (the rectified voltage ripple is not a pure 100 (120) Hz sine wave. This can be cured by suitable filtering at the mains input to the amplifier and high value filter output capacitance. (A choke-input filter is usually followed by a second LC or regulator.)
Some experimentation (observing the necessary safety measures!) might yield good results here. Also note that the voltage across such a resonant LC might be high - if a scope is available, do check. Using 630V capacitors is wise.
In short, swinging choke has to be designed for higher voltage swing without saturation than chokes for CLC filters.
Thanks for the replies.
I will, of course, 'scope it on the bench before committing anything to a chassis. But after reading the other threads regarding swinging chokes I'm feeling less confidant....
I will, of course, 'scope it on the bench before committing anything to a chassis. But after reading the other threads regarding swinging chokes I'm feeling less confidant....
Johan Potgieter said:
Hi Johan,
Have you jet listened at it? Despite the various issues I think you will found the sound absolutely fantastic !😉
Cheers,
Paolo
Paolo,
Sowter in the UK offers three models of choke-input filter chokes as stock items (I believe they can also make to order). Sowter web site
I have no commercial interest in Sowter
Sowter in the UK offers three models of choke-input filter chokes as stock items (I believe they can also make to order). Sowter web site
I have no commercial interest in Sowter
Wavebourn said:
I think you meant to say that chokes designed for choke input duty need to be designed for higher voltage swing than chokes for CLC filters.
While Swinging choke reside in the same position as the first element after the rectifier, the nature of their behavior relies on saturation to "swing". Granted at low DC currents, the design needs to handle the AC voltage without saturation providing maximum inductance (and reach critical inductance). As the DC current is increased the AC headroom is diminished and the choke begins to saturate. This causes a reduction of inductance making the filter gradually shift to cap input which increases the output voltage. If the choke is designed so the increase in voltage from increased current offsets the voltage drop across the associated resistances regulation is improved.
Simply put, the reason swinging chokes work is because they are designed to take advantage of the predictable behavior of saturation.
dave
"I have tried to use two different "standard" chokes but they was so noisy , unfortunately!"
I've worked in two transformer factories. The best way to make quiet inductors is to wedge the cores tight and then vacuum varnish or vacuum epoxy them Anything else will tend to be noisy in time.
While semi off topic, I knew an old school audio transformer designer, that had worked for ADC, that swore by air gapped core chokes for speaker crossovers to reduce the saturation in the core. He claimed that the main objection that people have to iron core inductors is that they saturate. He claimed that using a properly designed air gapped inductor for a speaker crossover network has sonic advantages over air core inductors.
Scott Novak
I've worked in two transformer factories. The best way to make quiet inductors is to wedge the cores tight and then vacuum varnish or vacuum epoxy them Anything else will tend to be noisy in time.
While semi off topic, I knew an old school audio transformer designer, that had worked for ADC, that swore by air gapped core chokes for speaker crossovers to reduce the saturation in the core. He claimed that the main objection that people have to iron core inductors is that they saturate. He claimed that using a properly designed air gapped inductor for a speaker crossover network has sonic advantages over air core inductors.
Scott Novak
- Status
- Not open for further replies.