Hey
@A2029 , here is the question I mentioned I had for you. I'll explain why these design considerations are important after all the technical chat afterward for others!
I've made some specific design changes to my power supply for this amp, wanted to explain why and get your $0.02 on them. Here is a chicken-scatch schematic of the HT circuit:
The first change I made was adding 150ohm resistance between the mains transformer secondary and the rectifier. The U18/20 datasheet states a minimum R-source value of 180ohms for a maximum input capacitor value of 16uF. The secondary resistance of the Lundahl LL1650 mains transformer I plan to use is ~20ohm. I've added this resistor to make up the difference to protect the rectifier and prevent arcing. If I keep it, would not be a 150ohm resistor, more likely two 75ohm, three 50ohm etc. with the appropriate wattage. Probably doesn't need to be quite this high since I am only using a 6.8uF input cap, but just for the sake of the example.
The second change I made was adding some series resistance across both LC filters. I have been using Morgan Jones
Valve Amplifiers book as a resource. Following his method (and because I am still working on figuring out some of this functionality in LTSpice), I modeled my power supply in PSUD2 without this resistance added. What I found is there is some minor ringing present. Here is a plot from PSUD2 without any of the added resistance. It is a stepped current load, increasing from 10mA to 160mA at 300ms to elucidate the ringing. May be hard to appreciate here (doing this on my laptop so the plot is a bit squished), but there is ringing on the output in the form of capacitor voltage variations. Note that this is without any "soft-start" enabled. Green is the 65uF cap and blue is the 175uF cap.
From reading
Valve Amplifiers, I understand that to provide adequate damping of this LC filter ringing, we are targeting a Q factor of 0.5. The formula for Q factor is:
Q = 1/R * sqrt(L/C)
So to provide additional damping, we can 1) increase C 2) increase R or 3) decrease L. Well increasing C is not preferable as these Clarity Caps are already enormous! Likewise, decreasing L is not a great option since I am then loosing valuable ripple filtering. Increasing R seems best and gives me the most bang for my buck. So, by adding 150ohms in series with each LC filter, we get this:
Again, might be hard to see, but at least visually, the voltage variation is gone on the output and the ringing appears to be well-damped. Obviously there will be a significant voltage drop from adding these resistors, but this is actually preferable for me as I have a bit too much B+ to work with, getting it down to 240-270V is a good place for where I want to bias the power tubes. The resistance could also be added by getting custom chokes made with a higher DC resistance and max current rating.
So, by adding these resistors before the rectifier and in series with the inductors, I am trying to accomplish the following:
1) protect the rectifier from current spikes, arcing
2) decrease the Q factor of the filter section and damp any ringing
3) drop the B+ to my preferred bias point
Okay, with all of that being said, other than the obvious heat dissipation that will take place in the power supply, do you see any negative consequences of adding this resistance?
)
