2
\$\begingroup\$

enter image description here

This is in the right hand power amp in a Nakamichi AV-10 receiver. Whilst watching the Olympics opening ceremony at considerable volume, the unit overheated (hot pcb smell) and cut out.

Having cooled it now works up to a point.

The main problem is that now cuts out at moderate volume. Using an Infrared Thermometer I have identified this capacitor running at 63C. It's mate on the other amp is at 30C. Cooling it with a fan allows the amp to operate at higher volume.

The other slight problem is that there is a faint crackle on the right speaker.

So, my question.

Do you think that the capacitor is the problem, or is it merely the indicator of a different failure?

\$\endgroup\$
9
  • \$\begingroup\$ The link is not readable. \$\endgroup\$
    – Dave Tweed
    Commented Dec 13, 2012 at 15:34
  • 2
    \$\begingroup\$ I'm not going to create a Google Account just to follow this link! \$\endgroup\$
    – user16324
    Commented Dec 13, 2012 at 15:44
  • 2
    \$\begingroup\$ It's not readable even if you have a google account. \$\endgroup\$
    – pjc50
    Commented Dec 13, 2012 at 15:57
  • 1
    \$\begingroup\$ I get error 403 even with Google account. \$\endgroup\$
    – Turbo J
    Commented Dec 13, 2012 at 16:25
  • \$\begingroup\$ Well I'm definitely not going to get a Google account just for a 403! :-) We're all working on the assumption that it's a big fat electrolytic here ... but what if it isn't? In any case, 0.15pf sounds a rather unlikely value... \$\endgroup\$
    – user16324
    Commented Dec 13, 2012 at 16:32

2 Answers 2

Reset to default
7
\$\begingroup\$

This question gives some of the background here. The capacitor may or may not have developed excessive leakage but its electrolyte is drying out - its ESR (equivalent series resistance) is increasing, so any ripple voltage dissipates more power in it, heating it up and drying it out faster. If you can measure its capacitance, you'll probably find it's lost most of that too... It's not totally unexpected, they are rated for a few thousand hours at a certain temperature - say, 85C.

Dave is right - replace it, paying attention to its value, voltage, and ripple current rating. And expect to replace its twin in a couple of months... better to do both at once.

********* EDIT *******

Now that the photo is visible, the above is wrong! (do I have to downvote it? )

This is not an electrolytic capacitor but some plastic film type, at a lower value (0.15 uF), and without the electrolytic's problems! The most likely reason for its temperature is that it is resting against a (probably rather hot) driver transistor. Is its mate on the other channel clear of the transistor? It should be safe to (gently) straighten it, but I think you have to look elsewhere for the real problem. The equivalent of these capacitors would be suspect in a 1940s valve amp, but not in a modern amp...

Check temperature of the output transistors first (one suspect is the bias current, if there is an adjustment for it) and check the power supply voltages if you can. Oh, and if you can measure capacitance, check the big electrolytics :-)

\$\endgroup\$
2
  • \$\begingroup\$ The picture shows a non-polarized capacitor, most likely a ceramic or a film type. There's no electrolyte to dry out IMHO. \$\endgroup\$
    – Adam Lawrence
    Commented Dec 13, 2012 at 18:34
  • \$\begingroup\$ @mmgm : you beat me to the edit! \$\endgroup\$
    – user16324
    Commented Dec 13, 2012 at 18:41
2
\$\begingroup\$

Yes, it's very likely that the capacitor in question has developed excessive leakage and/or series resistance. It would explain both the heat and the noise. Replace it.

You might consider replacing both capacitors; if one has reached its end of life, the other one is probably close, too.

\$\endgroup\$
9
  • \$\begingroup\$ Dave, you have made my day! I bought this lovely thing for over £1k nine years ago, and at a time when I could afford to. Things could not be more different today. the amps are attached to a monster heats sink and fan assembly, but it's all plugged and socketed, and the thermal paste looks very much re-usable. What type would be the very best (I know that it's a 0.15pf) but now the type. \$\endgroup\$
    – John
    Commented Dec 13, 2012 at 15:48
  • 1
    \$\begingroup\$ Can you fix the picture? Then we can give a more definitive answer. \$\endgroup\$
    – Dave Tweed
    Commented Dec 13, 2012 at 16:54
  • \$\begingroup\$ I would suggest that you even increase the rating on the capacitor so that the heat will not cause it to fail as quickly (rule of thumb is life is halfed for every 10 degree increase). \$\endgroup\$
    – Gustavo Litovsky
    Commented Dec 13, 2012 at 17:16
  • \$\begingroup\$ Done. Didn't realise that Picasa web albums were Google only. How silly. AND TYPO I meant mF not pF - sorry. \$\endgroup\$
    – John
    Commented Dec 13, 2012 at 18:16
  • \$\begingroup\$ You're saying it's the little blue capacitor that's getting hot? Very unusual. BTW, the value is 0.15 uF (150000 pF). This is NOT an electrolytic, like we've all been assuming. In any case, if it's the only thing getting hot, replace it. \$\endgroup\$
    – Dave Tweed
    Commented Dec 13, 2012 at 18:31

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