1
\$\begingroup\$

Hello there i am building a variable power supply as a summer project. I have ordered some parts, and want to begin building it. I start with the transformer power supply:

I've ordered a transformer (230 V - 36 V) and i expected it to give a output voltage of 36 V peak (silly me), and then found out the output will be in RMS, i. e. an peak outout of 50,92 V. I can work with this, but the capacitors i've orderes is only rated at 50 V, so this will give me some smoking problems i guess. The rectifying diodes will give a voltage drop of minimum 0,6 V, which is not quite enough to go under 50 V.

Is it possible to just use two diodes extra? As seen in the simulation:

enter image description here

The diodes added is of course D5 and D6. Their only purpose is to lower the voltage, to under 50 V. This will theoretical be enough.

But will i need to go lower in voltage, as the overhead on the WV of the capacitor is too small?

And is the "extra diodes" way even a viable way of doing this?

\$\endgroup\$
  • \$\begingroup\$ I am using 1N5402 diodes. \$\endgroup\$ – keffe Jul 10 '17 at 10:40
  • \$\begingroup\$ As a rule of thumb it is always desirable to select capacitors which are 1.5 times (2times is even better) the desired voltage. So, in short, the answer is NO. Theoretically(ideally) it should, but practically it will be a bad design. \$\endgroup\$ – Mayank Jul 10 '17 at 10:43
  • \$\begingroup\$ By 50,92 V you mean 50.92V ? I would suggest ordering higher rated capacitor or new transformer. \$\endgroup\$ – Mayank Jul 10 '17 at 10:45
  • \$\begingroup\$ I will order new capacitors then. And yes i mean 50.92 V. The , is the symbol to use in Denmark, and i sometimes forget that . is the way to go in almost any other place. \$\endgroup\$ – keffe Jul 10 '17 at 10:47
1
\$\begingroup\$

As a rule of thumb, it is always desirable to select capacitors which are 1.5 times (2times is even better) the desired voltage. So, in short, the answer is NO. Theoretically(ideally) it should, but practically it will be a bad design. An Even slight increase in input voltage or any small variation in the circuit than ideal and you might see the circuit smoked.

\$\endgroup\$
  • \$\begingroup\$ If it's a European 230V, then that's 230V +/- 10%. So a maximum input of 253V. \$\endgroup\$ – Simon B Jul 10 '17 at 11:31
1
\$\begingroup\$

What is important for the capacitor selection is the rms current it can accept and at what temperature. This is the major cause of failures in power supplies as manufacturers often select cheap capacitors which quickly overheat due to poor characteristics (high equivalent series resistance or ESR). Current ratings are usually given for a 120-Hz ripple so you should find the adequate information in the data-sheet if you operate from a 60-Hz network. So from your simulation results, extract the rms current in worst case (max output current and min input voltage for instance) and base your capacitor selection including some derating factor (safety margin). Then, voltage is of course important but not as sensitive as reported by some of the answers. A capacitor can operate close to its max voltage without problems. For instance, you can find 420-V bulk capacitors in hi-volume PFC-based adapters in which the nominal output voltage is 390 V (7% derating). However, using diodes to reduce the peak by a few volts is not reasonable especially considering a voltage drop depending on the absorbed current plus variations in the input voltage (+/- 15% in Europe). If you are stuck with the caps you have and providing they can accept the current, the best is to stack them in series and add equilibrium resistances so that they equally share the voltage. This is often done in hi-volume power supplies.

enter image description here

\$\endgroup\$
0
\$\begingroup\$

this will give me some smoking problems i guess.

No point in guessing. Capacitor ratings are magic and it wouldn't explode just because the voltage is a tiny bit off when unloaded

Yes you can use more diodes to further reduce the voltage on the capacitor but it isn't the best solution

\$\endgroup\$
0
\$\begingroup\$

No, adding a few diodes to drop the voltage is not a good idea.

At best, you'd be running the electrolytic capacitors right at their maximum voltage. That is not a good strategy. It will decrease their lifetime, and it will still leave you vulnerable to voltage spikes or a little higher than usual line voltage.

You should aim for running electrolytic capacitors at around 75% of their rated voltage most of the time. Then it's OK for spikes to occasionally go up to the rated voltage, but never above it.

Do it right. Get the right transformer, or get the right capacitors.

\$\endgroup\$

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.