I am using the MC34063ECD-TR to take 5V and make -12V. I'm using the schematic as recommended in the datasheet for this application (with the exception of using a 100uH inductor instead of a 90uH inductor). It is working nicely electrically - I get a pretty clean -12 Volt rail from it. The problem is I also get a loud audible hiss from the inductor. I can press on the inductor with my finger and it dampens the hissing sound. It's around 4 to 8 kHz based on my phone's FFT app, but to my ear it sounds like white noise.
Another problem is I get a lot of radiated emissions that is really messing up my analog gain stage. I have amplifiers taking a 30 kHz signal and amplifying by 110 dB (I'm taking 25uV and amplifying it to about 7.2 V). The amplification works VERY nicely under 1 of 2 conditions:
1) I depopulate the MC34063 (telling me it's either conducted or radiated emissions from the MC34063 interfering with the gain)
or
2) I cover the gain stages with a grounded metal box. (confirming it's actually 100% radiated emissions, not conducted). It's actually pretty neat - I can look at the output of the gain stages and as I cover it I can see the noise get smaller and disappear entirely.
I'm thinking this problem could be coming from a "bad layout". My question is: What pins of which parts need to be located physically close together to minimize the switching current loop for the MC34063's inverting DC-DC Converter using the datasheets recommended application schematic.
