I noticed something for the first time using PSUD with a voltage doubler, the ripple sine wave is flatter on the positive swing and pointier at the bottom. Why is this?
First, why would you expect them to be the same or sinusoidal in shape. This is a fundamentally nonlinear circuit as demonstrated by the frequency doubling.
I will give this as an explanation.
The peaks of the waveform occur when a diode is forward biased and therefore represent a low impedance with a short time constant. The troughs occur when both diodes are reversed biased and represent a long time constant.
The longer time constant is determined by the load, not the diodes.
Homework, investigate how the waveform changes as the load is varied.
I will give this as an explanation.
The peaks of the waveform occur when a diode is forward biased and therefore represent a low impedance with a short time constant. The troughs occur when both diodes are reversed biased and represent a long time constant.
The longer time constant is determined by the load, not the diodes.
Homework, investigate how the waveform changes as the load is varied.
In the quest for a totally quiet high gain guitar amp I created what I call the "way too many diodes power supply." It has a symmetrical full wave voltage doubler that did prove quieter than the usual circuit in a guitar amp, though not the amp shown in rest of the schematic. I never tried to simulate it though.
This supply provides about 165 volts for the tube screens from an isolation transformer using a full wave bridge and about 330 volts from the symmetrical doubler. It also provides raw unfiltered pulsating DC for the tube heaters which carries the same heating power as if the tubes were wired directly to the transformer secondary. Diode D9 provides isolation for the heater supply. It can be omitted (shorted) if there is no heater requirement. D1, D2, D5, D6, D7, D8, C20, C21, and C22 are required for the doubler. D1, D2, D3, D4, C23, and C24 are required for the bridge. Other than sharing D1 and D2, they are independent of each other.
This supply provides about 165 volts for the tube screens from an isolation transformer using a full wave bridge and about 330 volts from the symmetrical doubler. It also provides raw unfiltered pulsating DC for the tube heaters which carries the same heating power as if the tubes were wired directly to the transformer secondary. Diode D9 provides isolation for the heater supply. It can be omitted (shorted) if there is no heater requirement. D1, D2, D5, D6, D7, D8, C20, C21, and C22 are required for the doubler. D1, D2, D3, D4, C23, and C24 are required for the bridge. Other than sharing D1 and D2, they are independent of each other.
I'll echo the comment "Why would you expect a sine wave?" A periodic (repeating) waveform contains a fundamental frequency and harmonics. Look at the waveform at C2 - a sawtooth. An R-C filter section attenuates the 120 Hz fundamental, but attenuates the harmonics even more. So the second filter section make the ripple waveform sort of like a sine wave. Adding another section would result in lower ripple that contains less of the harmonics and would look even more sine-like.
This different shape of positive and negative waveform is not just for the doubler ... probably the result of charging / discharging current of the filter cap being ( obviously ) not equal . If you modify the load and input resistance values , more or less symmetrical waveform would result
Last edited:
Full wave rectification; Full wave voltage doubler rectification:
Cut a sine wave in half; then invert the second half (it was 180 degrees to 360 degrees).
When you invert it, the shape of the second half is the same as the shape of the first half (now they both look like 0 degrees to 180 degrees).
That is the first determinant of the shape coming from the rectifiers.
That shape is further modified by the input cap, or the input choke.
I hope that simplifies the understanding of what you are seeing.
Cut a sine wave in half; then invert the second half (it was 180 degrees to 360 degrees).
When you invert it, the shape of the second half is the same as the shape of the first half (now they both look like 0 degrees to 180 degrees).
That is the first determinant of the shape coming from the rectifiers.
That shape is further modified by the input cap, or the input choke.
I hope that simplifies the understanding of what you are seeing.
Looks exactly like it should. If it were FW CT then you would see a difference for sure.
The FW CT transformers used in audio are normally serially wound, that results in a difference
of resistance between the two halves of the HV Wdg. The result can contain a large
line frequency component.🙂
That's what always happens when you RC-filter a sawtooth waveform. It doesn't matter that it's a voltage doubler.
The classic 5T rectifier setup should be renamed a Voltage Halver, to help explain Voltage Doublers. Would also go a long way towards explaining the advantages of a full bridge.
All good fortune,
Chris
All good fortune,
Chris
With a full wave bridge, and also with a full wave voltage doubler . . .
The complete secondary windings(*) have to supply current for each positive alternation and each negative alternation (100 times/second for 50 Hz mains; and 120 times/second for 60Hz mains).
With a center tapped secondary, and full wave rectification . . .
Each half of the center tapped secondary winding only conducts at a rate of 1 X the power mains frequency.
(50 times/second for 50 Hz mains; and 60 times/second for 60Hz mains).
Even though a center tapped secondary has to have 2 X the turns of the other method above (*) . . .
The center tapped secondary wire only has to have 1/2 the cross section area of the secondary above (*).
In terms of total copper, center tapped secondary does not have as much more volume of copper that many of you expect.
The cross section area is smaller; yes the total wire length is a little more than 2 x length, because the extra turns are wound over each other; the outer windings have more length per turn.
The complete secondary windings(*) have to supply current for each positive alternation and each negative alternation (100 times/second for 50 Hz mains; and 120 times/second for 60Hz mains).
With a center tapped secondary, and full wave rectification . . .
Each half of the center tapped secondary winding only conducts at a rate of 1 X the power mains frequency.
(50 times/second for 50 Hz mains; and 60 times/second for 60Hz mains).
Even though a center tapped secondary has to have 2 X the turns of the other method above (*) . . .
The center tapped secondary wire only has to have 1/2 the cross section area of the secondary above (*).
In terms of total copper, center tapped secondary does not have as much more volume of copper that many of you expect.
The cross section area is smaller; yes the total wire length is a little more than 2 x length, because the extra turns are wound over each other; the outer windings have more length per turn.
Center tapped has lower transformer utilisation than Bridge or FWDoubler. 1/2 of the secundary takes up space but is idle half of the time. The transformer utilisation disadvantage of CT can be outweighted by the possible advantage of only needing 2 rectifiers. CT, or Bridge, or Doubler, the best choice is case dependend. The usually higher winding resistance of the CT worsen regulation but with C loaded vacuum tubes this is often wanted. Unequal R, Lleakage (CT winding ), or C (split C doubler with unequal caps) or unequal rectfier forward voltage drops, all can produce main frequency ripple. This 50 or 60Hz ripple content on the of the dc output is usually unsignificant, because it mostly gets swamped by the forward voltage drop of the tube or/and the combined effects of the primary and secundary winding resistances and leakage inductance of the transformer.
Last edited:
Merlinb,
I did answer it in Post # 7, but it did not seem to be noticed.
My mistake was in not providing a graphical illustration.
Now the question of efficiency, etc. came up.
Efficient? Purist' delight . . . direct heated tube rectifier bridge:
4 tube rectifiers in a bridge (no cheating . . . not 2 tubes and 2 solid state)
How good of a utilization is 4 filament windings just to power the direct heated rectifiers?
I did answer it in Post # 7, but it did not seem to be noticed.
My mistake was in not providing a graphical illustration.
Now the question of efficiency, etc. came up.
Efficient? Purist' delight . . . direct heated tube rectifier bridge:
4 tube rectifiers in a bridge (no cheating . . . not 2 tubes and 2 solid state)
How good of a utilization is 4 filament windings just to power the direct heated rectifiers?
Last edited: