Cappy said:
That's not my experience at all, which is why I like to use active regulation. But admittedly, I'm trying to have as little "sound" from my amps as possible. A supply which changes its output quickly with changes in current will certainly add something to the signal.
Allen Wright, in his little preamp book, wrote that one of his secrets is that anything in a circuit that should be pinned down should be pinned down as tightly as possible, and anything that should be free to move should be as free as possible. I think that's a nice way of summarizing it.
If you want to shorten the step response of this circuit, then get rid of the last RC section and reduce the transformer voltage. That will degrade regulation and decoupling, but that's what "faster response" means in this context. Ripple will still be far better than needed.
Sy,
Ok, thanks, I see that there are various schools of thought on this subject.
It's interesting that you suggest removing the last cell of Gary's supply to shorten the step response. That was also what I was going to suggest, based on my experiments over the last few days. This is the improvement I mentioned that I am going to write up today in my Aikido thread.
I am interested in comparing active regulation to the passive choke loaded supply, and hope to do that in the next few months. Swenson has moved to active regulation, your preference, after all was said and done.
I actually think there is a way for Gary to come up with a supply that satisfies both methodologies, to "have his cake and eat it too", if he wanted to do it. Then he could listen both ways and see what he prefers. If he reports back with his results we would all have another data point to learn by.
The way to do this is to make the first two LCLC cells abide by the quick response methodology. Then the last RC cell, the same as what he has now, would be in charge of damping. It would be simple to add or remove the last cell to see the differences.
Gary, what you have now is somewhat close to doing that. However, if you just take off the last RC cell, the supply won't work well, and I think most would agree with that. If you look in PSUD you will see mucho ringing.
By the way, my gut feeling here is that a ringing inductive supply with an RC network at the end to keep things under control is not optimum. It's like a person who doesn't operate well unless they are drinking alcohol to "damp them down". While it may look normal, under the surface it is still a chaotic system. That is what you currently have.
So I suggest you beef up the first inductor and reduce the second 47 uF capacitor, and change the second inductor to have more resistance to combat ringing. Then you can tune the impulse response by changing the first cap value.
Ok, thanks, I see that there are various schools of thought on this subject.
It's interesting that you suggest removing the last cell of Gary's supply to shorten the step response. That was also what I was going to suggest, based on my experiments over the last few days. This is the improvement I mentioned that I am going to write up today in my Aikido thread.
I am interested in comparing active regulation to the passive choke loaded supply, and hope to do that in the next few months. Swenson has moved to active regulation, your preference, after all was said and done.
I actually think there is a way for Gary to come up with a supply that satisfies both methodologies, to "have his cake and eat it too", if he wanted to do it. Then he could listen both ways and see what he prefers. If he reports back with his results we would all have another data point to learn by.
The way to do this is to make the first two LCLC cells abide by the quick response methodology. Then the last RC cell, the same as what he has now, would be in charge of damping. It would be simple to add or remove the last cell to see the differences.
Gary, what you have now is somewhat close to doing that. However, if you just take off the last RC cell, the supply won't work well, and I think most would agree with that. If you look in PSUD you will see mucho ringing.
By the way, my gut feeling here is that a ringing inductive supply with an RC network at the end to keep things under control is not optimum. It's like a person who doesn't operate well unless they are drinking alcohol to "damp them down". While it may look normal, under the surface it is still a chaotic system. That is what you currently have.
So I suggest you beef up the first inductor and reduce the second 47 uF capacitor, and change the second inductor to have more resistance to combat ringing. Then you can tune the impulse response by changing the first cap value.
Good stuff.
Ideally I would have lots of components around and just swap them out and reply back with what sounds better. But then that tells you nothing, right? Just another opinion.
Since PSUD2 is free, (not counting the risks associated with breaking and entering,) I'll model things according to Cappy's last recommendations. I'm not sure I am thoroughly convinced though; if the last RC cell cures the oscillations as well as reducing ripple and the only sacrifice is 5V, doesn't that justify it's use? If I may adapt your analogy, a drunk is a drunk and it might be like placing a cup of coffee and some aspirin in front of him after each nights booze-up. I have the impression from reading other choke filter threads that ringing is part of the bargain when using chokes. Mitigating it is the game and the RC helps. What do you think?
gary
Ideally I would have lots of components around and just swap them out and reply back with what sounds better. But then that tells you nothing, right? Just another opinion.
Since PSUD2 is free, (not counting the risks associated with breaking and entering,) I'll model things according to Cappy's last recommendations. I'm not sure I am thoroughly convinced though; if the last RC cell cures the oscillations as well as reducing ripple and the only sacrifice is 5V, doesn't that justify it's use? If I may adapt your analogy, a drunk is a drunk and it might be like placing a cup of coffee and some aspirin in front of him after each nights booze-up. I have the impression from reading other choke filter threads that ringing is part of the bargain when using chokes. Mitigating it is the game and the RC helps. What do you think?
gary
The RC helps with ringing, true. But what RC giveth, he also taketh away, or something like that.
Redesigning the LCLC part could cure the ringing, or improve it greatly, and then you could try with and without the RC part. I don't think ringing is a foregone conclusion with CLCLC power supplies. In any case, it just doesn't make sense to me to let the raving drunk take command of 2/3 of the house, even if the doors are locked. The AC current loops go through the ENTIRE house, including the parts the drunk is in. But that is just intuitive WAGing. If your intuition says something else go for it, it's all good fun. There are other people around who have been working on this for years and not just months; perhaps they will jump in.
I'll bet you 7 renminbi that if you build the LCLC part non-ringing per PSUD impulse response and also build the RC section as you designed it, you will prefer the "nude" CLCLC psu! I feel so strongly about this I'm even willing to put up the appreciating currency of our illustrious country's main creditor! But win or lose, it would be interesting to know what you thought of each type.
Redesigning the LCLC part could cure the ringing, or improve it greatly, and then you could try with and without the RC part. I don't think ringing is a foregone conclusion with CLCLC power supplies. In any case, it just doesn't make sense to me to let the raving drunk take command of 2/3 of the house, even if the doors are locked. The AC current loops go through the ENTIRE house, including the parts the drunk is in. But that is just intuitive WAGing. If your intuition says something else go for it, it's all good fun. There are other people around who have been working on this for years and not just months; perhaps they will jump in.
I'll bet you 7 renminbi that if you build the LCLC part non-ringing per PSUD impulse response and also build the RC section as you designed it, you will prefer the "nude" CLCLC psu! I feel so strongly about this I'm even willing to put up the appreciating currency of our illustrious country's main creditor! But win or lose, it would be interesting to know what you thought of each type.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.