I have a dead audio mixer power supply (Soundcraft MTK 22) which I am trying to bring back to life and learn more about flyback converters in the process. This particular supply runs on 230VAC and outputs +5V, -15V, +15V and +48V. The user guide states that the power consumption is <120VA. The problem lies in the switching circuit based on the UC3842A switching controller which I've sketched out here: enter image description here

(Connections on the right are connected to the primary side of the transformer)

The power mosfet was short-circuited so during my previous repair attempts I replaced the mosfet, the switching controller, C103, C107, C105, C102, R102 and R103 but the mosfet shorted again. Thus, I decided to also replace all the SMD components around the UC3842A. However, as I don't have a component tester at hand I can't determine the values of the C106, C108 and C104 capacitors. Also, I don't know the switching frequency, which is in turn determined by C106 and R106 (whose value I do know).

Is there some other way to work out the switching frequency or choose those capacitor values?


1 Answer 1


Is there some other way to work out the switching frequency or choose those capacitor values?

The UC3842 data sheet shows this graph: -

enter image description here

RT and CT in your diagram are here: -

enter image description here

So, remove C106 and test it with a capacitance meter because there's no other way to determine the switching frequency given that the device may be damaged. You might get lucky by removing the MOSFET and seeing if the UC3842's oscillator magically springs back to life of course.

100 kHz is fairly standard for switching frequencies so if you say that RT (R106) is 15 kΩ then C106 (CT) will be about 1 nF

  • \$\begingroup\$ Thank you. I have a further question. On the board C106 is quite close to R109, R108 and C108 on which where significant burn marks (especially on R109). So I guess there is a possibility C106 got damaged as well. Now, R106 is 15k Ohm, and as this is used in an audio setting, I guess this leaves us with 2.2nF, 1nF or lower (as higher values would go near the 20kHz which is close to the human hearing upper limit freq). This would imply 50kHz or 150kHz switching freq. Are there any other components which could help me eliminate one (maybe filter capacitors on the secondary side)? \$\endgroup\$
    – dioniz
    Jan 22, 2021 at 12:23
  • \$\begingroup\$ No, I don't think so. Most folk go for 100 kHz switching as a standard so maybe assume that is the operating frequency @dioniz however, if you can find a way of measuring the primary inductance of the flyback transformer, power throughput calculations can be made to better estimate the frequency. \$\endgroup\$
    – Andy aka
    Jan 22, 2021 at 12:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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