That may be correct, but it appears that your application is very different to a valve type amp so all the part values and how to approach any design aspect will be quite different.
Choke Input filters require at least Critical Inductance, in order to work as a choke input filter:
For 60Hz mains, full wave rectification (120Hz), the Critical Inductance is:
350 / DC Load mA
Example: 350 / 100 mA load = 3.5 Henry, the minimum inductance.
A common inductance value is 5 Henry, that will work if the current rating is also high enough.
For 50Hz mains, full wave rectification (100Hz), the Critical Inductance is:
420 / DC Load mA
Example: 420 / 100 mA load = 4.2 Henry, the minimum inductance.
A common inductance value is 5 Henry, that will work if the current rating is also high enough.
The voltage from a choke input filter is 0.9 x ACVrms voltage of the secondary; But it will be lower than that due to the voltage drops of the power transformer Primary DCR and secondary DCR, the rectifier (solid state - low V drop) or vacuum tube rectifier (significant V drop), and the choke DCR.
. . . That is much different than the capacitor input filter, which starts out as 1.414 x ACVrms, and has the same causes of additional drops, as listed above. It should be noted that the drops of a capacitor input filter is according to the peak current.
The peak current is several times larger than the load current. That larger current of a capacitor input filter, causes the DC Voltage drops of the power transformer primary and secondary DCRs, the vacuum tube rectifier, and the choke DCR to much larger than for the choke input filter.
The choke input filter current is the average current, and that average current is the same as the load current. So the voltage drops are lower.
Power transformers with choke input filters run cooler for the same DC load current, versus power transformers with capacitor input filters that have the same DC load current.
The peak transient currents of a capacitor input filter has many upper harmonics, and they fall into more sensitive frequencies of the human ear.
The average current of a choke input filter has much lower harmonic energy, so are not into the ear's sensitive frequency range.
If there are ground loops, you may hear hum and the harmonics. Better a choke input filter, versus a capacitor input filter.
One thing to pay attention to, if you use a choke input filter:
Space the choke away from interstage transformers, and output transformers; and orient the angular rotation of the choke by 90 degrees, versus the interstage and output transformers
Conclusion: Run choke input filters (Run Cool)
Welcome to the world of Choke Input Filters.
Just my experience and my opinions.
Happy designing, happy building, and happy listening!
For 60Hz mains, full wave rectification (120Hz), the Critical Inductance is:
350 / DC Load mA
Example: 350 / 100 mA load = 3.5 Henry, the minimum inductance.
A common inductance value is 5 Henry, that will work if the current rating is also high enough.
For 50Hz mains, full wave rectification (100Hz), the Critical Inductance is:
420 / DC Load mA
Example: 420 / 100 mA load = 4.2 Henry, the minimum inductance.
A common inductance value is 5 Henry, that will work if the current rating is also high enough.
The voltage from a choke input filter is 0.9 x ACVrms voltage of the secondary; But it will be lower than that due to the voltage drops of the power transformer Primary DCR and secondary DCR, the rectifier (solid state - low V drop) or vacuum tube rectifier (significant V drop), and the choke DCR.
. . . That is much different than the capacitor input filter, which starts out as 1.414 x ACVrms, and has the same causes of additional drops, as listed above. It should be noted that the drops of a capacitor input filter is according to the peak current.
The peak current is several times larger than the load current. That larger current of a capacitor input filter, causes the DC Voltage drops of the power transformer primary and secondary DCRs, the vacuum tube rectifier, and the choke DCR to much larger than for the choke input filter.
The choke input filter current is the average current, and that average current is the same as the load current. So the voltage drops are lower.
Power transformers with choke input filters run cooler for the same DC load current, versus power transformers with capacitor input filters that have the same DC load current.
The peak transient currents of a capacitor input filter has many upper harmonics, and they fall into more sensitive frequencies of the human ear.
The average current of a choke input filter has much lower harmonic energy, so are not into the ear's sensitive frequency range.
If there are ground loops, you may hear hum and the harmonics. Better a choke input filter, versus a capacitor input filter.
One thing to pay attention to, if you use a choke input filter:
Space the choke away from interstage transformers, and output transformers; and orient the angular rotation of the choke by 90 degrees, versus the interstage and output transformers
Conclusion: Run choke input filters (Run Cool)
Welcome to the world of Choke Input Filters.
Just my experience and my opinions.
Happy designing, happy building, and happy listening!
Last edited:
My vote as well for choke input supplies.they also sound way better than cap input supplies. Just need to follow the rules closely.
Thx so much for the information
My mains is 50hz
Hence 420/1000mA= 0.42H. Would it be too big of a choke?
Is this seems correct?
My choke is only 5mh. Too small right?
My application need 1-1.5a
My main is 220v 50hz
What happen is putting too small inductance for choke input?
This is my experience only. If your choke inductance is to small or you saturate the choke my ps diodes pooped. Once you saturate your choke it is no longer a choke input power supply. It simply becomes a cap input supply with dc at peak volts in.
Always consider the Un-loaded secondary voltage of the power transformer.
I mean the Un-loaded secondary's peak voltage with the Power Mains at its highest voltage (power mains varies day to day, and hour to hour).
Example:
The 230VAC power mains is higher than normal/usual, and is at 240VAC
The Power transformer Has a 230VAC primary; it has a 350V-0V-350V center tapped secondary.
With 230VAC and an Un-loaded secondary, it probably has 360V-0V-360V; but with 240VAC, it has . . .
240/230 x 360-0-360 = 375.7VAC-0V-375.7VAC.
375.7VAC x 1.414 = 531V Peak, yes . . . 531V Peak!
Suppose you use a Vacuum Tube Rectifier, with no cathode (direct heated filament) or solid state diodes, then the B+ will come up almost instantly.
Suppose the output tubes have a cathode, and take 11 seconds to warm up;
In that case the B+ comes up rapidly, but it is Un-loaded for 11 seconds.
The Unloaded B+ is now + 531VDC, and the secondary tap is -531VDC; the peak inverse voltage across the rectifier (tube or solid state) is 2X 531VDC = 1062VDC inverse voltage.
You should use a rectifier that has a Peak Inverse Voltage rating of 1500V.
By the way, a choke input supply becomes a cap input supply during the time the B+ is Un-loaded. That is because with an
Un-loaded choke, the critical inductance has to be Infinite Henrys (an Un-obtanium part).
If you use any rectifier that has a peak inverse voltage rating of less than 1500VDC,
then Your Mileage Will Vary, in fact the mileage will be cut short (Short, get it?).
Just my opinion (do Not ask me how I know).
PS: The Q factor is normally not an issue with center tapped full wave rectification, or a full wave bridge rectifier; But it can be an issue if the input choke and capacitor that follows the choke are resonant at 2X the power mains frequency; calculate the resonance 1/(2 x pi x (Root (L x C))). If needed, change the L or the C to move the resonance. Changing either L or C by 2X will move the resonant frequency by 1.4X.
If the rectifier is a half wave rectifier (like a bias rectifier for example), then make sure the LC resonance is not the same as 1X the power mains frequency.
Have fun designing and building a reliable power supply.
I mean the Un-loaded secondary's peak voltage with the Power Mains at its highest voltage (power mains varies day to day, and hour to hour).
Example:
The 230VAC power mains is higher than normal/usual, and is at 240VAC
The Power transformer Has a 230VAC primary; it has a 350V-0V-350V center tapped secondary.
With 230VAC and an Un-loaded secondary, it probably has 360V-0V-360V; but with 240VAC, it has . . .
240/230 x 360-0-360 = 375.7VAC-0V-375.7VAC.
375.7VAC x 1.414 = 531V Peak, yes . . . 531V Peak!
Suppose you use a Vacuum Tube Rectifier, with no cathode (direct heated filament) or solid state diodes, then the B+ will come up almost instantly.
Suppose the output tubes have a cathode, and take 11 seconds to warm up;
In that case the B+ comes up rapidly, but it is Un-loaded for 11 seconds.
The Unloaded B+ is now + 531VDC, and the secondary tap is -531VDC; the peak inverse voltage across the rectifier (tube or solid state) is 2X 531VDC = 1062VDC inverse voltage.
You should use a rectifier that has a Peak Inverse Voltage rating of 1500V.
By the way, a choke input supply becomes a cap input supply during the time the B+ is Un-loaded. That is because with an
Un-loaded choke, the critical inductance has to be Infinite Henrys (an Un-obtanium part).
If you use any rectifier that has a peak inverse voltage rating of less than 1500VDC,
then Your Mileage Will Vary, in fact the mileage will be cut short (Short, get it?).
Just my opinion (do Not ask me how I know).
PS: The Q factor is normally not an issue with center tapped full wave rectification, or a full wave bridge rectifier; But it can be an issue if the input choke and capacitor that follows the choke are resonant at 2X the power mains frequency; calculate the resonance 1/(2 x pi x (Root (L x C))). If needed, change the L or the C to move the resonance. Changing either L or C by 2X will move the resonant frequency by 1.4X.
If the rectifier is a half wave rectifier (like a bias rectifier for example), then make sure the LC resonance is not the same as 1X the power mains frequency.
Have fun designing and building a reliable power supply.
Last edited:
Critical inductance for a choke input filter is not a simple reciprocal function of DC current, but is more like a linear function of load "resistance". A good rule of thumb is output volts divided by output current (in milliAmps) gives critical inductance in Henries. Remember it as load "resistance" in kOhms equals critical inductance in Henries.
All good fortune,
Chris
All good fortune,
Chris
Chris Hornmeck,
Please explain, you said:
"Remember it as load "resistance" in kOhms equals critical inductance in Henries."
By that rule, 1 kOhm load resistance needs 1 Henry (or 1k Henry)?
Please explain, you said:
"Remember it as load "resistance" in kOhms equals critical inductance in Henries."
By that rule, 1 kOhm load resistance needs 1 Henry (or 1k Henry)?
1 Henry. We should also remember that "critical inductance" isn't a red line in the sand; it's more a bend in the curve. We need to allow very considerable headroom, at least a factor of two. So actually use at least 2 Henries.
All good fortune,
Chris
All good fortune,
Chris
Why? The amp hopefully is isolated from the filter by its LF cutoff. Hardly needed for a Class A Amp, if it is working properly the plate current varies little.
But choke input was often used in the past for Class AB & Class B Amps where the plate current varies a lot. 🙂
In extreme cases & to save weight swinging chokes were common. Used for Public Address Systems & Transmitter Modulators.