2
\$\begingroup\$

This circuit is from Fig. 20.3 in Douglas Self, Small Signal Audio Design, Third Edition, Routlege, 2020. Headphone Amp, Fig. 20.3 in D. Self, Small Signal Audio Design, Routlege, 2020

This is a headphone amplifier, and as such, both input and output signals are alternating current. Because the TL072 can't drive large loads, the design includes a very simple Class-AB amplifier.

My question concerns the capacitors C1 and C2. I have very carefully copied them from the figure in the book, and they are very clearly drawn as polarised capacitors. Now I read, in books, and also as answers to questions on this very site, that you should not put AC across polarised capacitors. On the other hand, Mr Self is not some random dude with a YouTube channel. He really knows what he's doing and so I doubt that he made a simple mistake. So my question is, what is going on here? Can I use polarised caps in this design safely? And if yes, does the polarity really matter, given that the voltages across them will not be polarised in one particular direction for very long?

And if I can not use polarised caps safely, what should I use instead, given the rather large value especially of C2?

\$\endgroup\$
3
  • \$\begingroup\$ See this question: electronics.stackexchange.com/questions/66893/… \$\endgroup\$
    – Bart
    Commented Jul 30, 2021 at 10:46
  • \$\begingroup\$ I have seen it, and it says that this type of design will probably destroy the cap eventually. So apparently we have in Mr Self a person who has been designing audio circuits for decades, yet who would design an unsafe circuit that will ultimately self (as it were) destruct. This is what I can't wrap my head around and which ich why I asked the question. \$\endgroup\$
    – Stephan Neuhaus
    Commented Jul 30, 2021 at 10:57
  • 1
    \$\begingroup\$ it says that this type of design will probably destroy the cap eventually. And I'm saying that there is no issue, the DC biasing of both capacitors is 0 V so there is no issue. Actually the circuit doesn't really need C2 as C2 is intended to block the DC voltage but there will only be a very small DC voltage present due to the DC offset of the opamp, which is small. So C2 isn't actually needed. It is still safe to add it though. And this circuit will last a very long time, there is no issue with the capacitors. \$\endgroup\$
    – Bimpelrekkie
    Commented Jul 30, 2021 at 11:06

1 Answer 1

Reset to default
3
\$\begingroup\$

Can I use polarised caps in this design safely?

Yes you can.

...does the polarity really matter

It does but polarized capacitors / electrolytic capacitors can always withstand a small voltage of the "wrong" polarity. A small voltage could be 1 V DC for example.

Also it is the DC voltage that matters. In this circuit both sides of the capacitors are biased at a DC voltage of 0 V. When there is no (audio) signal , the DC voltages across the capacitors is 0 V.

This can easily be done in this circuit as the amplifier is powered from a symmetrical power supply of +/- 17 V. For comparison, have a look at an amplifier which is powered from a single supply and note how the input and output capacitors will have a DC voltage acros them so then the polarity is much more important compared to this design.

Only when a signal is applied will there be some (small) AC signal across the capacitors, remember that the capacitors short the AC signal so there will not be a lot of signal across the capacitors. If there is, then a larger capacitor value needs to be used.

..given that the polarity of the voltages across it will not be polarised in one particular direction for very long?

Yes and that is why the AC sigal is OK even if that means that the voltage across the capacitor will be negative half of the time. What matters is the average or DC value. A proper AC signal has a DC value of 0 V so it does not really have any polarity.

\$\endgroup\$
4
  • \$\begingroup\$ Ah, the remark that "there will not be a lot of signal across the capacitor" makes sense, and "If there is, then a larger capacitor value needs to be used" would explain why C2 is much larger than C1. \$\endgroup\$
    – Stephan Neuhaus
    Commented Jul 30, 2021 at 11:00
  • \$\begingroup\$ The comment about why the type of power supply is important also makes sense. When you only have a positive power supply, you will need to bias your signal, which is just a fancy term for adding a large DC component. \$\endgroup\$
    – Stephan Neuhaus
    Commented Jul 30, 2021 at 11:06
  • \$\begingroup\$ ...need to bias your signal, which is just a fancy term for adding a large DC component Correct, then you need to "superimpose" (combine, add) a (large) DC voltage to the AC signal so that the voltage is always in the range in which the amplifier can operate properly. I would not call "biasing" a fancy term, it is essential for proper operation. When you design an amplifier or any signal processing circuit, an experienced designer will first check that the biasing is correct. Because if it is not, the circuit will have issues. (and beginners often forget to check the biasing!). \$\endgroup\$
    – Bimpelrekkie
    Commented Jul 30, 2021 at 11:10
  • \$\begingroup\$ Sorry, I was a bit flippant there. I didn't want to imply that biasing itself was fancy, I do understand that proper biasing is important. I just remembered the moment when I understood what biasing was and going "oooh, it's just adding DC!" :-) \$\endgroup\$
    – Stephan Neuhaus
    Commented Jul 30, 2021 at 11:18

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