Radio board going crazy with radio transmitter

[rvalente]

Hello Mates,

this prototype has a very simple task, on the push of a buttom a recorded message is played through the HAM radio on a commercial frequency.
Its a very crude notification system.

(The seen band is HAM but, the end use will be commercial band).

The radio is being powered by a battery eliminator, just a 7808 inside the baofeng battery case.
The board is supplied with 12V from the lab power supply.

When the button is pushed the STM32 (running arduino) sends a very simple serial burst to the MP3 player, which turns on the TX transitor to the radio and feeds the audio to the radio.

The STM32 checks the end of the message by pulling the MP3 player "playing pin" and while is low, the TX is held down.

The radio is driving everything crazy (5W tx power on VHF), the lab power supply galvo goes crazy, the board STM32 thinks the audio is still playing and I'm obliged to remove the headphone jack to quiet everything down.

There is no ground plane at all since I'm using the perfboard with 0805 and enamel wire for connections and there is little to no noise suppression.

My question is: What kind of noise suppression should I use? How to calculate it?

[DimitriP]

My question is: What kind of noise suppression should I use? How to calculate it?

There is no "noise".

I almost went crazy by looking at it  
You are describing RFI,  picked up by the miles of interconnecting wires.

The first answer is better layout.
The second answer is instead of blasting 5W while testing use lower power.

I was going to say add capacitors and RF chockes along signal paths inputs and outputs but with the current layout it's a crapshoot ....

 

[thinkfat]

If you cannot reduce the TX power, connect the transceiver to a dummy load during testing.

I hope this is not the final setup, you'll have a hard time getting all the radio interference under control. I suggest you put a lot more distance between the transceiver and your board and use a choke on the connecting wire to block the RF. Or use a separate antenna that you place a couple ten meters away from your board.

[rvalente]

Hello friends,

I much appreciate your replies.

As suggested, did some things:

Added a small filter in the input power supply, just a 47uh inductor and a 10uF cap.
Added a small shield in the bottom of the board.
The antenna is not anymore in the radio, but with a 1m long cable

The behavior is now much much better.

Anyway, this is just a proof of concept. The final product will have a decent layout board with the antenna being at list 10m away from the board.

Thanks a lot!