How bad is this power supply ringing?

Tubeglowpio

Active Member
So I'm wiring up this amp this weekend, finally had the time. I've never used a clcl filter before and know very little about ringing in a power supply, basically just know I want as little as possible. I ran an emulation in psud2 and believe I set it up right and these are some of the results I got. I'm wondering how bad this ringing is and if there is anything I should do to improve it, thank you!Screenshot (3412).png Screenshot (3413).png
 
Try playing with the values of the c2 and c3. I noticed that if you double c1 for c2 then double or triple c2 for c3 usually gets rid of the ringing.
 
If the major load is a triode output stage, then it looks much more resistive than a CCS load, so the ringing should be substantially reduced. Try a resistive load in the simulation.
 
If the major load is a triode output stage, then it looks much more resistive than a CCS load, so the ringing should be substantially reduced. Try a resistive load in the simulation.
its just a pair of el84 in pentode then 12at7 driver mike, since its not triode that makes it less of an issue correct? I'll change to resistive load, what resistance value should I try?
 
To see the ringing under transient load, I usually stepped my current load, to 2uA for the first few seconds (5) then apply the full current peak value with the second parameter of the stepped current load.

Then you'll see how your PS is reacting and if there is oscillation until it reaches equilibrium. You don't want oscillation and the game is to get rid of them :)
 
To see the ringing under transient load, I usually stepped my current load, to 2uA for the first few seconds (5) then apply the full current peak value with the second parameter of the stepped current load.

Then you'll see how your PS is reacting and if there is oscillation until it reaches equilibrium. You don't want oscillation and the game is to get rid of them :)
Just to clarify, did you mean milli amps or micro amps? So enter 5 seconds in the stepped screen then just 2 uA, and for second part of stepped current apply full expected current. This is basically a rh84 btw.
 
To see the ringing under transient load, I usually stepped my current load, to 2uA for the first few seconds (5) then apply the full current peak value with the second parameter of the stepped current load.

Then you'll see how your PS is reacting and if there is oscillation until it reaches equilibrium. You don't want oscillation and the game is to get rid of them :)
Screenshot (3418).png
 
Yes You set it at nothing for the first period, let's say 2uA (as you can't set it to 0), then after the period, apply the full current peak you need.
You wil see how your LCL filter reacts to sudden current demand and does not ring.
 
Yes You set it at nothing for the first period, let's say 2uA (as you can't set it to 0), then after the period, apply the full current peak you need.
You wil see how your LCL filter reacts to sudden current demand and does not ring.
Did it, this is what I got Screenshot (3419).png
 
Yes You set it at nothing for the first period, let's say 2uA (as you can't set it to 0), then after the period, apply the full current peak you need.
You wil see how your LCL filter reacts to sudden current demand and does not ring.
I also messed with the cap values and even tried adding some resistance in line with second filter choke. Tough to get it just right so I have lowest ripple and no ringing.
 
Screenshot (3420).png This may be as good as its going to get, I went back to my original design and adding a little resistance in line with second filter choke got me this. Shouldn't cause any problems correct?
 
Sorry for all the posts, this is another one that I have the parts for. Between the two which one would work best?Screenshot (3423).png
 
I'm trying to document these steps because things can change a lot, I think I'm starting to get the hang of using this to check ringing. So I think having that second choke is really screwing up my supply, I think putting the chokes in series looks a lot better. Here is a before and after of that.Screenshot (3432).png Screenshot (3431).png
 
Why do you need two chokes? What ripple voltage are you aiming for? Here's a simpler idea that may achieve your goals just as well. There is zero ring at the current transition at 10 seconds into the simulation. The red trace shown is V(C2).

upload_2018-5-28_16-55-51.png

This has ripple of 73 mv p-p at V(C2) and 5 mV p-p at V(C3). Changing C2 to 220 uF produces ripple of 35 mV at V(C2) and 7 mV at V(C3), according to simulation.
 
Last edited:
Why do you need two chokes? What ripple voltage are you aiming for? Here's a simpler idea that may achieve your goals just as well. There is zero ring at the current transition at 10 seconds into the simulation. The red trace shown is V(C2).

View attachment 1198854

This has ripple of 73 mv p-p at V(C2) and 5 mV p-p at V(C3). Changing C2 to 220 uF produces ripple of 35 mV at V(C2) and 7 mV at V(C3), according to simulation.
For SE amps there seems to be a benefit in sound quality with a dual choke. The second choke can also help to reduce voltages without a resistor. Since class A is current dependant and chokes are designed for that dual chokes are a viable consideration.
 
Back
Top Bottom