1
\$\begingroup\$

I'm building the following power supply (credits to Audio design guide): enter image description here

I would like to replace C5, C8 and C9 with 1000uF capacitors (with the same voltage rate). From my understanding these capacitors are meant to limit the ripple and to provide a clean and stable voltage, so I think it should be ok to have more capacitance here. Am I right?

\$\endgroup\$
  • 1
    \$\begingroup\$ You will have a much higher inrush current, so make sure your diodes are up to the job. also, 3000 uF at 310 V is a lot of stored energy, and potentially life threatening. Be careful. \$\endgroup\$ – Colin Mar 5 '18 at 8:37
  • 1
    \$\begingroup\$ so I think it should be ok to have more capacitance here. Yes, you can use higher value capacitors. R6-C8 and R7-C9 pairs will bring better filtering. But please note that higher capacitance brings lower ESR thus the inrush will be higher. \$\endgroup\$ – Rohat Kılıç Mar 5 '18 at 8:45
  • 1
    \$\begingroup\$ You will add bigger capacitors, possibly with higher ESL to the circuit (because they are bigger). The inrush current will be larger and the danger of them damaging things upon a failure is also higher. And since they are bigger, the cost is also higher. But there is a form of regulation after the capacitors for both supply rails, so it is unlikely that you will notice any benefit in terms of ripple due to the increase in capacitance. \$\endgroup\$ – Joren Vaes Mar 5 '18 at 10:23
  • \$\begingroup\$ Thank you for your comments. If I understand well the main problem would be the inrush current, and I will check if the diodes for this. @JorenVaes I understand your points, and I also believe the benefits don't justify the replacement of the capacitors. I want to use them only because I already have plenty of them available at home. \$\endgroup\$ – Enrico Mar 5 '18 at 10:45
  • 2
    \$\begingroup\$ @Enrico In that case it seems like you should be fine. If you really want to be on the safe side you can always consider some form of NTC inrush current limiter. That way you don't need to over-spec your diodes as much just so they survive that inrush. \$\endgroup\$ – Joren Vaes Mar 5 '18 at 10:47
4
\$\begingroup\$

Due to R7 and R8, you can increase the size of C8 and C9 without an issue. But I would not increase the size of C3 that much. As you increase the size of C3, the diodes will be conducting for a shorter and shorter time per cycle, yet the power being supplied to the output will not vary -- so the current through the diodes (while the are conducting) must be higher. Also, as the commenters pointed out, there is no need at the first stage of filtering to stress the diodes that much. For the sake of reliability, I would stick with the 100 uFd specified for C3

\$\endgroup\$
  • 1
    \$\begingroup\$ Note the bigger the caps you have for C8 and C9, the slower the rise time of the B+ supply will be, this may or may not be an issue. \$\endgroup\$ – Colin Mar 5 '18 at 12:33
  • 1
    \$\begingroup\$ @Colin__s This is true. The thumbrule is 5 x the RC timeconstant to full voltage. This won't be an issue in a tube circuit. I would be more concerned with the long time it takes to discharge after you shut the power off. Depending on what's downstream of the power supply, C8 and C9 could have a high voltage shock potential for a long time after power off. \$\endgroup\$ – Michael Gorsich Mar 6 '18 at 10:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.