I'm currently making a power supply with a 32 volt AC out, so the rectified voltage would be 32*1.414-diode drop. The value comes around 44 volts. My question is, is it okay to hook up a 50 V rated capacitor for filtering in this case? As the voltage would be 44 V and a very low headroom.
-
1\$\begingroup\$ How well regulated is your 32 VAC? +-10%? \$\endgroup\$– winnyCommented Dec 15, 2018 at 18:18
-
2\$\begingroup\$ Depending on your country, line voltage can be allowed to vary by usually between 5% and 10% so you need at least that margin, plus there is a gain in expected life from having a higher margin. \$\endgroup\$– Phil GCommented Dec 15, 2018 at 18:18
-
\$\begingroup\$ How do you expect to get 44VDC rectified from a 32VAC Supply? Is it DC Biased? \$\endgroup\$– crasicCommented Dec 15, 2018 at 20:43
3 Answers
Probably not, unless your load is fairly constant and heavy (in relation to the transformer rating). And even then, it's a bit close. The line voltage can vary as well, but it's not unusual for a lightly loaded transformer to output 15% or more higher voltage than the rated voltage. The rectifier voltage loss will also be less at relatively low current, so more volts across the capacitor.
Assuming a bridge rectifier you can get about 60% of the current rating of the transformer secondary in DC amperes.
Sometimes electrolytic capacitors have a "surge" rating that can handle brief overvoltage above the "WV" = Working Voltage, but aside from the cheapest consumer devices, it's better to keep well within the voltage rating marked on the capacitor. This is especially true when the ripple current is high in relation to the ripple current rating. Also, I see no such rating on your part.
TL;DR: A 63V part would be much better.
One can operate an electrolytic cap. up to its maximum voltage as stated on the cap. It really is not a good idea to operate the cap. at its limit. higher cap. voltage above your requirements improves the power factor and visa versa. Very demanding current required using a single cap? Use multiple units (adding to cap. required) Such as a 100 Uf required, use 2 X 50 Uf, (and the highest voltage cap's you can afford.) this increases cap. life. Use the largest area (via AWG stat's) of wire to work best. Heat is the enemy of cap's! Good hunting!
There is no single straight forward answer. For a homebrew/lab/hobby device it will probably work , but I would recommend using a 100V capacitor and as much capacitance you can afford. Assuming you actually do have 44V Rectified Output
- From the point of view of safety, one can operate an electrolytic capacitor up to its rated voltage level in most cases.
Electrolytic capacitors do not tolerate much, if any overvoltage and thus must be overrated for use in power supply filtering circuits where ripple or spikes may exceed the voltage. 50V cap for a 44V Half Rectified supply is too close for my comfort
Large ripple, like that of a Half rectified supply decrease the lifetime of the device, much more capacitance is required than expected to keep the effective ripple on the output caps tolerable .
- Electrolytic capacitors age and effective capacitance drops as it ages, further increasing the ripple
- Electrolytic capacitors can fail implosively from overvoltage or heat failure.
- Ceramic Capacitors can take 100% overvoltage and do not age, in most cases and work well for filtering applications
- Ceramic Capacitors are more expensive and harder to find in equivalent capacitance ratings
- Ceramic Capacitors have a significant change in capacitance with DC Bias so a=even with the 100% tolerance, a much larger voltage or even larger capacitance must be used to compensate.
- Capacitors can be installed in series/parallel configurations (with biasing resistors) to tolerate capacitance failure and to increase the effective voltage rating (series) and effective output capacitance (parallel)
-
1\$\begingroup\$ I would truly love to see a capacitor implode. Vacuum tubes can implode. Capacitors generally explode. \$\endgroup\$– JRECommented Dec 15, 2018 at 20:58
-
\$\begingroup\$ And, I would like to see a ceramic capacitor that could be used in place of a 10000uF electrolytic capacitor. I'm thinking like a shoe box full of ceramics might get enough capacitance together. \$\endgroup\$– JRECommented Dec 15, 2018 at 21:00
-
\$\begingroup\$ @JRE 10000uF was not explicitly in the op question and I didn't even notice, certainly overkill. Ceramic cap info is just that, informational for comparison . Btw my friendly cap rep will fight you over saying they explode it's bad marketing ;) \$\endgroup\$– crasicCommented Dec 15, 2018 at 21:17
-
\$\begingroup\$ @JRE This one might be capable of imploding. \$\endgroup\$ Commented Dec 16, 2018 at 0:29