Power supply for simple FM transmitter - MAINS HUM

Now that i understand FM transmitters better i built my own variation on the 1-transistor TX, and i need a power supply for it. Last time i fooled around with such things i would get HUM, HUM and MORE HUM, so i need a little help.

My transmitter is basically a spin-off of this design: but i skipped the pre-emphasis (i do it in software) and the tuned circuit has different values, i just used whatever i had around.

It works and sounds great, and has excellent stability for just one transistor, it's far better than the 3-transistor TX i used to have, that drifted frequency even if i just moved its antenna around (now i understand why, the antenna was a part of the tuned circuit in that tx). I wanted to add a buffer stage but it seems to do without, since i only need it to cover my house. I do have ONE problem with it - frequency drifts downwards about 0.05MHz every half hour, due to my cheap 9 volt battery draining. That is no problem for an analog radio but my Yamaha PLL tuner has to be readjusted when it happens.

So i want to make a power supply for this thing. Trouble is, if it gets within 50cm of anything plugged into the mains, it gets lots of hum. On my old TX i had a 10000uF cap followed by a LM317 and that wasn't enough, i would still have hum at about -20dB which is quite noticeable. I'm thinking a capacitance multiplier power supply would be better? Followed by a discrete regulator instead of a LM317 maybe? I noticed that the 317 does add a bit of hiss of its own.

I'd appreciate any insight.

Your problem sounds more like an earth loop than hum through the power supply.

Keith

What earth loop since it is NOT plugged into anything? Merely placing a 230v AC cord next to the battery-operated device, with signal from my PDA so that is battery operated as well, induces hum. If i plug the charger into the PDA i'll get hum again.

I can only guess at you problem from the limited information you have given. From your latest post I would say poor construction and screening is the most likely cause.

Keith

keith1200rs said:

rom your latest post I would say poor construction and screening is the most likely cause.

Keith

Click to expand...

Screening, or lack thereof, could be the issue. I'll address that and see what happens. By the way, there has been no drift in the past two hours, guess the battery has stabilized. Since i have fixed my frequency i guess i'll dab some glue onto the inductor and screen everything. I have a shield from a TV tuner, i guess that will do.

Oh, next question. If i limit the power supply current as not to damage the battery, could i use the battery as a very big capacitor? It seems like a good idea to me.

You can use a rechargeable battery as a "capacitor" (or vice versa!) but I really don't see the point in your case. A regulator should easily be able to provide very good 50/60Hz rejection so any hum problems won't be from there and so won't be solved with a battery/huge capacitor.

Keith.

did you used shielded wires for inputs to Tx ?
this my wild guess.

if , so , did you ground it properly?
srizbf
11thmay2010

I do have shielded cable now. But i did not on my old TX... Maybe that's where all the hum was coming from. But anyway, i added a buffer stage and instead of doing what it's supposed to, it does the very opposite - the buffer should help the oscillator stay on frequency, instead, short-term drift is now worse.

It's perfectly fine on my analog radios but no chance getting good reception on my digital Yamaha tuner.

I think that before this thing will be worth a nice power supply, it'll need a PLL. I just found a CD4046 so that's what i'll be doing next.

hum pickup , one of the main cause , is unshielded
inputs.

and grounding of shield should be proper.

the osc drift , mentioned after adding a buffer ,
may be due to dc voltage slow varying in the outpu of buffer.

if it is direct coupled then that may be the cause.

after seeing the complete circuit diagram of the setup,
only then , can come to any conclusion.

cheers for your experimental enthusiasm ,

srizbf
12thmay2010

This is my schematic. The values in the tuned circuit may not be the same as in the real thing but they're close enough. The buffer is coupled to the oscillator by a second coil next to the one in the tuned circuit, basically acting as an air core transformer. In theory this should put very little load on the oscillator so it should not drift. In theory that is. The antenna is coupled to the output of the buffer by a capacitor so again i don't see the source of the drift. The 75 ohm resistor in the schematic represents the antenna.

The coil in the tuned circuit has 5 turns with 5mm internal diameter, the one coupling the buffer has 9 turns with the same diameter. Maybe i should try reducing the number of turns of the coupling coil? If i do not connect the antenna there is no drift at all.

I'll post a few pictures of the actual device soon.