Well no, that's not a trademark of SE amps. The SE output stage sees the PSU in series so if the PSU is not good enough you will listen to the sound of the SE and its PSU....that's why some people can hear the difference when swap tube rectifiers. Of course, this will be always true to some degree but there is a difference between clearly audible and minimal (or even inaudible is some cases).
For sure the SE amp PSU needs a lot more care than PP. As always it is a compromise among different aspects (regulation, recovery and noise). So when you start designing you should check you do not overdo in one area at large expense of the other. If that happens you modify it. That's where a simulator is very useful....That's my opinion, of course.
That's difficult to say. Who knows what they do....
It's the capacitor that discharges and recharges but the anode current of the tube is SE stage is constant as long as it works in a linear fashion. I don't have PSU designer because I don't use a PC but I guess that's what you see....
That's what the amplifier is supposed to do: take the power from the PSU and convert it into music.
https://dalmura.com.au/static/Power supply issues for tube amps.pdf
I have just found this article that covers several aspects and also uses PSUD for simulations. It might be helpful...
I have just found this article that covers several aspects and also uses PSUD for simulations. It might be helpful...
I am sure several people will disagree about certain statements, like the use of SiC diodes and vacuum damper diodes. But that's just a general recipe and of course SiC and damper diodes can be used with great results...if their specs are not exceeded.
SiC diodes have no reverse recovery spikes, but they have relatively high forward voltage and very abrupt cut-off and turn-on. These contribute to ringing in the tank of power transformer leakage inductance and winding capacitance.
SiC is like striking a bell with a hard steel hammer, and swinging the hammer wide. By contrast, using a slow recovery diode with low forward voltage is like striking the bell with a felt-tipped club and short swing.
Of course, snubbing takes care of prolonged ringing and should always be employed with any kind of SS rectifiers. But snubbing is ineffective against the initial bang. If a bell is damped by putting a hand on it, and striken, it will not remain silent.
This is the main reason why tube rectifiers are superior to SS.
BTW, psud2 doesn't have inputs for transformer parasitics, so it completely misses the ringing.
SiC is like striking a bell with a hard steel hammer, and swinging the hammer wide. By contrast, using a slow recovery diode with low forward voltage is like striking the bell with a felt-tipped club and short swing.
Of course, snubbing takes care of prolonged ringing and should always be employed with any kind of SS rectifiers. But snubbing is ineffective against the initial bang. If a bell is damped by putting a hand on it, and striken, it will not remain silent.
This is the main reason why tube rectifiers are superior to SS.
BTW, psud2 doesn't have inputs for transformer parasitics, so it completely misses the ringing.
Last edited:
Are SiC any faster (more abrupt) than regular fast Si diodes?
SIC have large Vdrop, but the 1.4-ish volts is far less than tube rect, so it depends where you are from in regards to SiC's Vdrop being good or bad. Vdrop is more constant with load tho, here a tube rectifier is like a series resistance and the Vdrop increases with load.
SiC have no reverse charge which regular Si diodes do have. What does that mean in practical terms? I am not sure I understand it fully, but I think when a diode HAS reverse charge, it will not start to conduct as soon as it should, but rather 'holds on' a bit, when it does conduct the current will come on a bit abrupt and aggressively. Someone correct me here please. This is a big deal in high speed switching supplies and class-D amplifiers. It can be a problem in regular AC rectification if the leakage inductance is such that the abrupt current produces spikes.
But SiC diodes are very capacitive. This I am not sure is good or bad...? It is bad when combined with leakage inductance since together they create parasitic LC circuit and we get ringing. But caps in parallel with fast diodes are often used to slow down the diode a bit, so to not allow for the abrupt and aggressive current. So the SiC diode's capacitance may be that parallel capacitance and help the diode switch calmly.
Or am I way off here? I can never fully decide for soft recovery Si diodes, SiC diodes, or tube rectifier... The latter is my top choice for the main amp, the others for less serious circuits, but perhaps I need to rethink?
SIC have large Vdrop, but the 1.4-ish volts is far less than tube rect, so it depends where you are from in regards to SiC's Vdrop being good or bad. Vdrop is more constant with load tho, here a tube rectifier is like a series resistance and the Vdrop increases with load.
SiC have no reverse charge which regular Si diodes do have. What does that mean in practical terms? I am not sure I understand it fully, but I think when a diode HAS reverse charge, it will not start to conduct as soon as it should, but rather 'holds on' a bit, when it does conduct the current will come on a bit abrupt and aggressively. Someone correct me here please. This is a big deal in high speed switching supplies and class-D amplifiers. It can be a problem in regular AC rectification if the leakage inductance is such that the abrupt current produces spikes.
But SiC diodes are very capacitive. This I am not sure is good or bad...? It is bad when combined with leakage inductance since together they create parasitic LC circuit and we get ringing. But caps in parallel with fast diodes are often used to slow down the diode a bit, so to not allow for the abrupt and aggressive current. So the SiC diode's capacitance may be that parallel capacitance and help the diode switch calmly.
Or am I way off here? I can never fully decide for soft recovery Si diodes, SiC diodes, or tube rectifier... The latter is my top choice for the main amp, the others for less serious circuits, but perhaps I need to rethink?
Forward voltage is not same thing as voltage drop. SS diodes do not conduct in forward direction before forward voltage exceeds certain value. It is about 0.2 V for Ge diodes, 0.4 - 0.6 V for Si diodes, and 1.5 V for SiC. Tubes have no forward voltage, their conductance starts at 0 V.
SiC are as fast as UF diodes. Fast recovery is critical for HF applications in SMPS, but does not have any advantage at mains frequency.
SiC are as fast as UF diodes. Fast recovery is critical for HF applications in SMPS, but does not have any advantage at mains frequency.
Well yes diodes and tube rects start conduct sooner than the standard forward voltage drop. I am lazy and just write Vdrop hoping most understand since it relates to the subject. The average Vdrop of a tube rect is surely more than Si diodes. Just b/c tibes start conduct at 0V everybody knows a typical 20-50V is dropped over the tube where this is roughly 0.7V for Si diodes.
Yes, exactly, but the issue here is abrupt current cut-off of SS diodes that causes transformer ringing. Tubes don't have such current cut-off, so they don't cause the ringing.
Well, one just needs to find the best recipe. With tube rectifiers, one can have a regulator downstream and the increased voltage drop with higher current demand can be addressed, if any. For true Class A operation I don't think it's a big deal and other factors are more important. Some damper diodes are good for rectification too and they have much less drop, in case.
With SiC I have no experience so cannot really comment.
With SiC I have no experience so cannot really comment.
You get the same ringing from smps dcm vs ccm modes between the switching and the inductors. Possibly could use some of the same technique to reduce the ringing?
Wow. resurrected SE thread from 2017 now mutates to discussions on OPT ringing...
Ever notice that TV damper diodes with the lowest internal resistance seem to require more heater current? 🙂
Ever notice that TV damper diodes with the lowest internal resistance seem to require more heater current? 🙂