So a few weeks back I got making this nice looking lab power supply, but other than running my bench LED lights, its not been used. And thus begins my world of hurt, because the other day I had a project that needed some power and after connecting it up, things just were not doing what you would expect them to do.
A circuit that should give a nice sine wave looked a mess on the scope, a receiver I powered up had this god awful switch mode racket on just about everything. So get get the pixie wrangling gear out to take a look at what was going on and it was nice a pretty picture.
As you can see from the scope output, the 12v DC was not pretty and it was making my projects not happy. So i figure, It needs a filter and I do 30 seconds of math and figure 10 to 100uh of inductance with 100uf of capacitance should be close to enough to the business.
So I make a nice looking filter and connect it up to the power and yes it attenuated the noise and ripple but it really did nothing else, all the problems were still there.
When you know you are chasing your own tail, start talking to the smart people. In my case this is Brenton, and we got talking about all aspects of the design, what modules I used what SMPS i used and a bunch of other specifics.
Well it turns out, had a fundamental flaw in my PSU design, I had left the 0v DC floating and this is a trap for young players. Because in the words of someone smarter than me “Without that connection the output is magnetically isolated and noise leaks out via capacitive coupling in the output transformer.” So after checking that 0v DC was not tied to Mains Earth, they were quickly coupled together and all my problems went away.
And as you can see from the first scope output above, the power is now rather clean, and in the 2nd scope output, is the DC after my new filter. The filter will be installed permanently in the next few days and will remain a permanent part of my lab power supply. And with that fixed, we can get onto building a receiver now for the 10w CW transmitter.




