A very good treatment of the MC34063 in buck converter mode, so lacking on YT. The noise at switch on is most likely due to pickup in the control logic. It’s really hard to get a switching regulator to work well without a well designed PCB to control stray impedances, and the multiple transitions will increase switching loss. But it does seem to be regulating well. Nice job!
Hi Mark, Thanks for the kind comments. The MC34063 is a nice chip. I’ve found that most of the time, the noise is caused from the inductor being undersized. The data sheet is a bit optimistic with their inductor calculation. The other problem is if you get a bunch of fake chips. You will see it immediately on a scope. Cheers Craig
Yeah, I found it a few years ago, good price and good multiple sourcing for production designs. Its behaviour changes quite a lot as the load varies, which more complex chips are better at handling, but unbeatable for price. And I found a few errors in the data sheet formulae. But well worth getting to know.
Hi there, If you look at the chip block/functional diagram - you can definitely check certain portions of the chip to make sure they are not blown. Cheers Craig
I'm concerned the buck regulator may not be working properly as shown. I don't believe that sawtooth pattern should be visible at the 5ms/div timescale shown. That looked like 200hz when it should be operating at 80kHz. My first suspicion is that the inductor may be grossly oversized if it's really a 100mH in place of a 130uH.
Hi Tinfever, The circuit works as it should - the problem is that I got a whole bunch of fake chips from one of the online suppliers. Once I changed the chip out for the real one - then everything looks as it should. I don't think I was using 100mH - maybe 100uH. I cannot remember that chip offhand but 100mH would give me 50.2K of impedance which sounds way too much. Cheers Craig
A very good treatment of the MC34063 in buck converter mode, so lacking on YT. The noise at switch on is most likely due to pickup in the control logic. It’s really hard to get a switching regulator to work well without a well designed PCB to control stray impedances, and the multiple transitions will increase switching loss. But it does seem to be regulating well. Nice job!
Hi Mark,
Thanks for the kind comments.
The MC34063 is a nice chip. I’ve found that most of the time, the noise is caused from the inductor being undersized. The data sheet is a bit optimistic with their inductor calculation. The other problem is if you get a bunch of fake chips. You will see it immediately on a scope.
Cheers
Craig
Yeah, I found it a few years ago, good price and good multiple sourcing for production designs. Its behaviour changes quite a lot as the load varies, which more complex chips are better at handling, but unbeatable for price. And I found a few errors in the data sheet formulae. But well worth getting to know.
Did you ever look with a scope at the signal at the cathode of the diode - pin 2 of the 34063?
Hi Egbert,
I did look have look at the cathode signal with the scope. The signal will be flat because of the smoothing capacitor.
Cheers
Craig
Is there a way of testing this chip by ohms or continuity across certain pins to find if it is deffective?
Hi there,
If you look at the chip block/functional diagram - you can definitely check certain portions of the chip to make sure they are not blown.
Cheers
Craig
I'm concerned the buck regulator may not be working properly as shown. I don't believe that sawtooth pattern should be visible at the 5ms/div timescale shown. That looked like 200hz when it should be operating at 80kHz. My first suspicion is that the inductor may be grossly oversized if it's really a 100mH in place of a 130uH.
Hi Tinfever,
The circuit works as it should - the problem is that I got a whole bunch of fake chips from one of the online suppliers. Once I changed the chip out for the real one - then everything looks as it should.
I don't think I was using 100mH - maybe 100uH. I cannot remember that chip offhand but 100mH would give me 50.2K of impedance which sounds way too much.
Cheers
Craig