1. Start with a 350-0-350VAC secondary, 2 solid state diodes in full wave configuration, use a cap input CLC filter, and 100mA DC load. Use 60Hz power mains (full wave rectification gives 120Hz).
Use a 10uF input cap in the CLC filter.
Xc = 1/(2 x pi x f x c).
c = 10 uF
Xc = 133 Ohms
In that case, the approximate peak to peak ripple across the first cap is 0.1A x 133 Ohms = 13.3V peak to peak.
The choke has to deal with 13.3V peak to peak. EASY Job, even for a cheap choke!
The power supply will give just a little less than (1.414 x 350VAC) - (13.3/2) = 488.25V
2. Start with a 350-0-350VAC secondary, 2 solid state diodes in full wave configuration,
a choke input LC filter, and 100mA DC load. Use 60Hz power mains (full wave rectification gives 120Hz).
Lcrit = 350/mA; 350/100mA = 3.5 Henry
Use a 5 Henry choke, and have a little design safety margin.
The peak voltage to the choke input is 1.414 x 350VAC = 494.9V
With a choke input filter, the DC output will be about 0.9 x 350VAC = 315VDC
But the choke has 494.9V peak on the input and 315VDC on the output.
The choke has to swing 494.9V - 315V = 179.9V peak to peak.
The choke has to deal with 179.9V peak to peak.
That is a HARD Job for a choke.
Do not us a cheap choke here!
Just open your wallet and "Swing" the purchase.
You also will find that to get the maximum ripple you will accept, you will probably have to use a LCRC filter.
That R will give some additional voltage drop, so figure how much larger of a primary voltage you will need to get the final DCV you need.
Use a 10uF input cap in the CLC filter.
Xc = 1/(2 x pi x f x c).
c = 10 uF
Xc = 133 Ohms
In that case, the approximate peak to peak ripple across the first cap is 0.1A x 133 Ohms = 13.3V peak to peak.
The choke has to deal with 13.3V peak to peak. EASY Job, even for a cheap choke!
The power supply will give just a little less than (1.414 x 350VAC) - (13.3/2) = 488.25V
2. Start with a 350-0-350VAC secondary, 2 solid state diodes in full wave configuration,
a choke input LC filter, and 100mA DC load. Use 60Hz power mains (full wave rectification gives 120Hz).
Lcrit = 350/mA; 350/100mA = 3.5 Henry
Use a 5 Henry choke, and have a little design safety margin.
The peak voltage to the choke input is 1.414 x 350VAC = 494.9V
With a choke input filter, the DC output will be about 0.9 x 350VAC = 315VDC
But the choke has 494.9V peak on the input and 315VDC on the output.
The choke has to swing 494.9V - 315V = 179.9V peak to peak.
The choke has to deal with 179.9V peak to peak.
That is a HARD Job for a choke.
Do not us a cheap choke here!
Just open your wallet and "Swing" the purchase.
You also will find that to get the maximum ripple you will accept, you will probably have to use a LCRC filter.
That R will give some additional voltage drop, so figure how much larger of a primary voltage you will need to get the final DCV you need.
Last edited:
You could also just use Sowter's custom part request form to see if they can make you a replacement choke that will work for what you're doing.
Hey, aren't swinging chokes the ones that are meant to saturate with current swinging, thus decreasing inductance?
Let's clarify this terminology first.
AFAIK, a swinging choke is preferable for a class AB amplifier. At idle/low currents, in the swinging choke input filter, the inductor has enough inductance to satisfy critical loading and DC voltage is roughly equal to 0.9*Vsecrms.
When current draw increases during signal peaks, the choke starts to saturate due to larger direct current draw, hence its inductance decreases and during these instants, the filter transacts from choke input to capacitor input and output DC voltage starts rising. It works like a passive voltage regulator.
For a class A amplifier choke input filter, one does not need a "swinging choke", but one with a LARGE ripple swing headroom and sturdily built, glued and potted to reduce mechanical hum, because it will have a high core magnetostriction due to large signal Bac.
Let's clarify this terminology first.
AFAIK, a swinging choke is preferable for a class AB amplifier. At idle/low currents, in the swinging choke input filter, the inductor has enough inductance to satisfy critical loading and DC voltage is roughly equal to 0.9*Vsecrms.
When current draw increases during signal peaks, the choke starts to saturate due to larger direct current draw, hence its inductance decreases and during these instants, the filter transacts from choke input to capacitor input and output DC voltage starts rising. It works like a passive voltage regulator.
For a class A amplifier choke input filter, one does not need a "swinging choke", but one with a LARGE ripple swing headroom and sturdily built, glued and potted to reduce mechanical hum, because it will have a high core magnetostriction due to large signal Bac.
Last edited:
A story of Cap Input B+, versus Choke Input B+ . . .
These amplifiers have the same power transformers, similar filament current draws, and similar B+ current draws. They have solid state rectifiers, no regulator circuits, and no switchers.
Cap Input B+
SE amp with a Beam Power tube in UL.
Line draw: 0.47A, 44.2W, 56.6VA. Power Factor: 0.76
435V B+
Choke Input B+
Push pull amp with a pair of Beam Power tubes in UL.
Line draw: 0.31A, 35.8W, 38.1VA. Power Factor: 0.94
285V B+
Consider the power factors of 0.76 versus 0.94.
That is a major difference between cap input and choke input.
Which amp passes some of the world’s more difficult Power Factor Regulations?
It is not the cap input model.
Just another choke input story.
These amplifiers have the same power transformers, similar filament current draws, and similar B+ current draws. They have solid state rectifiers, no regulator circuits, and no switchers.
Cap Input B+
SE amp with a Beam Power tube in UL.
Line draw: 0.47A, 44.2W, 56.6VA. Power Factor: 0.76
435V B+
Choke Input B+
Push pull amp with a pair of Beam Power tubes in UL.
Line draw: 0.31A, 35.8W, 38.1VA. Power Factor: 0.94
285V B+
Consider the power factors of 0.76 versus 0.94.
That is a major difference between cap input and choke input.
Which amp passes some of the world’s more difficult Power Factor Regulations?
It is not the cap input model.
Just another choke input story.
> Swinging chokes have no magical properties.
They are cheaper. (And lighter!)
You can get the same performance with an oversize choke. Big enough to give the required minimum-load inductance. Yet large enough not to saturate or sag at maximum load.
A swinging choke, correctly specified and built, may be half the cost/weight.
They are cheaper. (And lighter!)
You can get the same performance with an oversize choke. Big enough to give the required minimum-load inductance. Yet large enough not to saturate or sag at maximum load.
A swinging choke, correctly specified and built, may be half the cost/weight.
Maybe not so much saturate, as decrease gracefully. All iron core inductors "swing" but some swing more than others. A "swinging choke" has its copper to iron allowance weighted more towards the copper end than a "smoothing choke".
All good fortune,
Chris
Good point. This depends of the core as well. Amorphous cores have a very good "slow knee".