I was given a box full of some Rifa 5600uf @ 420v capacitors today. These are electrolytic caps and I was planning to use them in series for the power supply of a PP 6550 amp I am working on. Anyone know of any drawbacks to using this much capacitance in a tube amp? The PS will be CLC with SS diodes and around 120uf for the first C; 6hy @ 400ma. choke and the two monsters in series for around 2800uf @ 800v. B+ is around 580v. Do such large caps normally have higher ESR/ESL?
Your PSU will have a subsonic resonance at about 1.2Hz. Not necessarily a problem, although this could be in the range of syllabic rate for speech or music. The characteristic impedance will be about 46 ohms, so difficult to damp effectively. Smaller C would give a higher frequency, but better damped.
You could change from CLC to CRC or flat out C and eliminate the resonance. With that amount of cap there should be plenty of ripple attenuation.
Beware that the caps will suck quite a bit of current at turn-on. Soft start (resistor in series with the primary of the transformer that's shorted with a relay after the supply is up) is definitely needed. Otherwise, you're almost guaranteed to pop fuses or blow the rectifier diodes.
~Tom
Beware that the caps will suck quite a bit of current at turn-on. Soft start (resistor in series with the primary of the transformer that's shorted with a relay after the supply is up) is definitely needed. Otherwise, you're almost guaranteed to pop fuses or blow the rectifier diodes.
~Tom
Resistance is good. It damps resonance. But maybe your choke has enough internal resistance and then you will need those voltage balancing resistors in parallel with the series caps. So it should work fine. Except the in-rush current at power on. You will blow a fuse as those caps will look like a short across the transformer. You might use a thermistor to reduce in-rush to something reasonable. Model the big caps as a short and add what it takes to keep the current limited to something less than infinity
I fould that large caps didn't work for me in a CLCRC circuit (excluding of course the first cap). The sound was very plain and boring, so I switched from the 330uF to 100uF and the music had better definition and timing. That might be due to the slower recharge/discharge times of the higher C values.
Anyway, give them a go, but please compare them with 100uF values as well.
Andrew
Anyway, give them a go, but please compare them with 100uF values as well.
Andrew
I had wondered about this having heard the difference before in a SE amp I built. I know that the PS in a SE amp has quite an impact on the sound however I was quite surprised to hear just how much. I was thinking that this application in a class AB push-pull amp with such a high amount of capacitance the caps would'nt be fully discharging, or is this incorrect? How much energy can a pair of 6550/KT88 really pull?
Yes the dynamic charge and discharge when in operation. If you use a dropper resistor in the PSU, or if the resistance of the choke is significant enough, then the RC time constant may become noticable.
However, thinking about it, if you are only using CLC (and not CLCRC) then you might be able to get away with higher values of C. It's when you get to using CRC, or similar, when cap size becomes very critical.
At the end of the day it is worth experimenting with values, hearing if there is a difference.
Andrew
The caps were the best price = free!! Now if only I can find a free pair of Lundahl OT'S I will dance a jig!!!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.