Hello,
the chokes are 5H
B+ = 430V
the output and driver stages draw about 95ma current, class A
barely meet the critical inductance threshold for the PS to work as choke input.
I use IDH rectifiers so no problem with over-voltage at start up.
Putting aside the case of valve failure causing the voltage to go up and supply caps exploding, is there any benefit to put in a big hot shunt resistor, and if yes, then how much current should I shunt away? My main transformers are rated 3 times the idle current so I can be generous with the shunt.
Also, there will be no time during which the current through the first choke reduce when the class A amp is fed with music signal, is my understanding correct? the reason I ask is because when playing with PSDU and step-current, even a small negative step (reduce the current say from 95ma to 90ma) produce some very strange result to B+, which I think because PSDU see that the supply is no longer a choke input one due to critical inductance / minimum current condition is no longer met. The amp would always draws more current when playing music than when idling, never less, right?
Thanks.
Vi
the chokes are 5H
B+ = 430V
the output and driver stages draw about 95ma current, class A
barely meet the critical inductance threshold for the PS to work as choke input.
I use IDH rectifiers so no problem with over-voltage at start up.
Putting aside the case of valve failure causing the voltage to go up and supply caps exploding, is there any benefit to put in a big hot shunt resistor, and if yes, then how much current should I shunt away? My main transformers are rated 3 times the idle current so I can be generous with the shunt.
Also, there will be no time during which the current through the first choke reduce when the class A amp is fed with music signal, is my understanding correct? the reason I ask is because when playing with PSDU and step-current, even a small negative step (reduce the current say from 95ma to 90ma) produce some very strange result to B+, which I think because PSDU see that the supply is no longer a choke input one due to critical inductance / minimum current condition is no longer met. The amp would always draws more current when playing music than when idling, never less, right?
Thanks.
Vi
Hi!
You want to stay well above the critical current. As you pointed out already, if a tube fails, the voltage rises and if your capacitors are not rated high enough, they might blow up.
The Radiotron Designers Handbook, recommends to have a current draw twice the critical current. In order to reach that you would need a low enough bleeder resistor which would dissipate a lot of power. A better way would be to exchange the first choke for one with twice the inductance
Best regards
Thomas
You want to stay well above the critical current. As you pointed out already, if a tube fails, the voltage rises and if your capacitors are not rated high enough, they might blow up.
The Radiotron Designers Handbook, recommends to have a current draw twice the critical current. In order to reach that you would need a low enough bleeder resistor which would dissipate a lot of power. A better way would be to exchange the first choke for one with twice the inductance
Best regards
Thomas
You could add on a "protective" shunt regulator.
The reg is set for the maximum voltage you expect for all "normal working conditions"
It will pull zero current when choke output is less than reg set point.
It will pull progressively more current as the choke starts to enter the critical region.
You could add a current limit detector into the shunt regulator that latches the power off, if the system starts to go into overload.
The reg is set for the maximum voltage you expect for all "normal working conditions"
It will pull zero current when choke output is less than reg set point.
It will pull progressively more current as the choke starts to enter the critical region.
You could add a current limit detector into the shunt regulator that latches the power off, if the system starts to go into overload.
Yes, a fat zener diode (or string of diodes) with a voltage just above the normal running voltage will pull current only when necessary.
Apparently strange results from PSUD2 when changing load current may be a sign of ringing. Put Ls and Cs together with little resistive damping and you get tuned circuits. Try making most of the load a resistor, with just a current load to make up the difference.
Apparently strange results from PSUD2 when changing load current may be a sign of ringing. Put Ls and Cs together with little resistive damping and you get tuned circuits. Try making most of the load a resistor, with just a current load to make up the difference.
DF96 Sir, can you please explain what you meant by "try making most of the load a resistor, with just a current load to make up the difference".
the chokes I have are pretty big ones, they are rated 5H @ 200ma 100hz with a DCR 20ohm. By running only 95ma-100ma through them would the inductance be more than 5H? I asked the maker of the chokes (he is quite reputable) who said at that level of current the inductance is around 8-10H, does this sound right? then I should not be worry too much about not meeting critical current?
I need to worry now about LC ringing, at 5H choke inductance the ringing is not an issue on PSDU, but at 8-10H the supply output swings wildly in response to a step current change.
the chokes I have are pretty big ones, they are rated 5H @ 200ma 100hz with a DCR 20ohm. By running only 95ma-100ma through them would the inductance be more than 5H? I asked the maker of the chokes (he is quite reputable) who said at that level of current the inductance is around 8-10H, does this sound right? then I should not be worry too much about not meeting critical current?
I need to worry now about LC ringing, at 5H choke inductance the ringing is not an issue on PSDU, but at 8-10H the supply output swings wildly in response to a step current change.
A resistive load damps down any LC resonance. A constant current load does not. A real amplifier is likely to be more like a resistive load than a current load. PSU2 provides two types of load. I am not sure if it allows you to use both together, though.
High quality chokes are more likely to suffer from ringing, as they have lower DC resistance.
There may also be a worse problem with a choke input supply running above the critical current. Below the critical current the DC voltage varies with load, as the choke current cuts off for part of the AC cycle. This simulates a resistor at subsonic frequencies, which dampens resonances. The same effect can be seen with a capacitor input filter. Once above the critical current the choke (and hence the rectifiers) conducts all the time - this means that at subsonic ringing frequencies the transformer end of the choke is simply grounded with no resistive-like damping. I suspect this is what you are seeing in your simulation.
High quality chokes are more likely to suffer from ringing, as they have lower DC resistance.
There may also be a worse problem with a choke input supply running above the critical current. Below the critical current the DC voltage varies with load, as the choke current cuts off for part of the AC cycle. This simulates a resistor at subsonic frequencies, which dampens resonances. The same effect can be seen with a capacitor input filter. Once above the critical current the choke (and hence the rectifiers) conducts all the time - this means that at subsonic ringing frequencies the transformer end of the choke is simply grounded with no resistive-like damping. I suspect this is what you are seeing in your simulation.
I've tested 8H/100mA chokes and found a 20% rise in inductance when used at half current. Those had 100 ohm copper resistance.
What effect do you think that data is showing?
The test plots are for an increasing applied test voltage, which is also increasing the choke current - two parameters changing, so can't state that inductance falls when operating current falls.
Vi, that's not always the situation for class A PP - the current draw could move around a bit as the output stage moves from no signal to approaching clipping from what I've observed with triode PP (Williamson).
- Status
- Not open for further replies.