Maker Pro
Maker Pro

Clicking Noise from DC-DC Converter powering audio circuit

I am building a DC-DC converter to power vacuum tubes with the higher voltages they typically use, all from a 12V supply. This current project involves a small pentode amplifier (technically beam sheet tube, 6ME8) that will work with about 30-60 volts on the anode resistor. I'm using the MC34063 as shown in the uploaded schematics.

The tube is breadboarded itself, along with the converter I designed onto a PCB, and another that I have on the breadboard, such that either can be alternately connected to the tube. The PCB board can operate from 50 to 200 volts, while the breadboarded version is designed for 30 volts, and doesn't have an external transistor, but more filter capacitance.

Regardless of design, no matter the amount of filter capacitance, the grounding offered by a ground plane on a PCB, both circuits produce the same clicking noise of about 1 to 3 hz or so. I'm not sure if I need to switch to a true PWM design, such as the UC3843. Any tips or suggestions appreciated.
 

Attachments

  • IMG_0314.jpg
    IMG_0314.jpg
    145.3 KB · Views: 15
  • Screen Shot 2021-12-27 at 8.00.48 PM.png
    Screen Shot 2021-12-27 at 8.00.48 PM.png
    278.6 KB · Views: 17
  • IMG_0317.jpg
    IMG_0317.jpg
    385.8 KB · Views: 12

Harald Kapp

Moderator
Moderator
Check the output of the MC34063 for continuity of operation. Does the gate drive switch continuously or does it show the 1 Hz - 3 Hz as a periodic interruption? Showinbg maybe (just an idea) a trigger a saftey function that stops the oscillator (e.g. current limiter)?
Also it is good practice to add a bypass capacitor directly to the Vin pin (VCC) of the chip as shown e.g. in figure 10 of the datasheet.
 
Oscillates at about 80 kHz, but over the interval of low frequency. I believe it's operating in discontinuous mode, only activating to charge up the inductor.

The breadboarded one is fully decoupled, if you are talking about the kicad design then the bypass is connected to pins 1, 7, and 8.
 

bertus

Moderator
Hello,

How much current do you expect to get from the circuit?
Resistors R1 and R2 have a quite high resistance.
Try to lower them to say 100 Ohms (or even lower) and see how the circuit reacts.

Bertus
 
The current tube only uses about .500 mA, so the output filter resistors are fine. Leaving them at 1k-3k allows for good filtering.
 

bertus

Moderator
Hello,

Then try to move the regulation loop from C3 to C1 (right after the diode) and see what happens.

Bertus
 
Removed the cap on pin 7 on the pcb and tried bypassing the input with a 2200uf capacitor, and the output with a 220uf cap. Instead of a click, it is now a buzzing noise at higher frequency, about 200Hz.
 
The hand drawn schematic is what I'm currently working with. I mentioned the pcb because it has a ground plane, I was basing it off this schematic: https://gra-afch.com/how-it-works/power-supplies-for-nixie-clocks/

The link is for a design that powers nixie tubes and indicator tubes, for which noise isn't an issue.

The breadboard is a re-design of the PCB. Connected the 220K resistor to the right most output capacitor and it resolved all the issues beautifully. Turns out even having a ground plane won't mitigate circuit design oversights. The supply I have is noise and hum free, I'll be measuring and reporting back on why it works so well.
 
Top