Why radios hate SMPS?

[ali6x944]

hello everybody,
i was working on a battery powered AM radio reviser/transmitter, i want to transmit for a minimal distance of 20m and across all frequencies in the AM band.
the batteries I'm going to use are 9V pile construction type -max Wh out is 400 mAh- Eveready Silver 9V (6F22) datasheet :
data.energizer.com/PDFs/6F22SIL_EU.pdf
so not a hell a lot of time before i drain it out if i want to get farther rang, so i used a 34063 in the boost and the inverting configurations
to get more power...
but when i run it kills my radio transmission, and the only thing i hear is a hushing sound!
turn it of and replace it with an 7805 regulator, the thing works amazingly but i get a rang of just 10cm!
please help!

[orolo]

SMPS generate quite a bit of switching noise which can get into your circuit: first, through the power supply, unless you filter it out really well, and second, through magnetic coupling, unless you shield both your transmitter and your supply as well as you can. Even with filtering and shielding the noise may not go away, and many people would suggest you to try mains transformers + linear regulators for noise sensitive applications that require moderate power.

If you want to stick to SMPS, try the following: shield both your transmitter and power supply with grounded metal boxes. Liberally filter both the input and output of your SMPS with ferrite beads and capacitors. Watch the output of your SMPS with an 'scope: if there remains significant noise at about 33Khz (and harmonics), consider some additional filtering, perhaps a capacitance multiplier. If your power supply is far from the transmitter, twist the supply wires to reduce pickup. To be honest, I've never stuck too long with SMPS after getting bad performance in transmitters, so I never tried to push the idea. I switched to linear regulators.

The fact that with a 7805 you get lots of less power may depend on your circuit, or the antenna you are using. At these frequencies, an antenna needs to be quite big to be efficient.

[dom0]

SMPS not only radiate / conduct energy near their switching frequency, but also at much greater frequencies (dozens, even hundreds of MHz) as a result of damped oscillations that occur due to the switching action itself. Getting these out of a system is a lot harder than starting over and getting a properly designed, low interference SMPS.

[ali6x944]

so i have to drop the use of SMPS for this design?

[Jeroen3]

No you have to properly design it, and shield it.
Watch some teardowns on the channel. Almost all products contains smps supplies but they don't interfere with your AM signal, or do they?

[ali6x944]

any shielding methods u know?

[Ian.M]

You'd be surprised how much RF 'hash' common household gadgets, appliances and light fittings with SMPSUs put out on the AM bands.  e.g. to get good reception of national marine weather forecasts on a station in the longwave band, I have to switch off phone chargers, fluorescent lighting (with electronic ballasts) and a portable 12V/mains fridge.

@O.P.  a broadband transmitter, even if it appears to be short-range when an ordinary portable domestic radio is used to receive it is extremely antisocial.  A radio amateur even several kilometers away who has a sensitive receiver and a good outdoor aerial may find your transmitter is jamming weak international stations, will have the equipment to find your location, and can easily report you to the authorities.    Also its much much easier to get the range you want with a single frequency tuned transmitter, which if you stay within the commercial broadcast bands and avoid the frequencies of national and local stations, although still probably illegal, is *FAR* less likely to attract attention.

The fact your transmitter is excessively sensitive to power supply noise indicates its a very poor design.

[orolo]

any shielding methods u know?

Put the circuit inside a metal cookie box and ground it. I'd swear I saw that very thing done in an application note by Linear Technology about some ultraprecise preamplifier. Anyway, it works.

@O.P.  a broadband transmitter, even if it appears to be short-range when an ordinary portable domestic radio is used to receive it is extremely antisocial.  A radio amateur even several kilometers away who has a sensitive receiver and a good outdoor aerial may find your transmitter is jamming weak international stations, will have the equipment to find your location, and can easily report you to the authorities.    Also its much much easier to get the range you want with a single frequency tuned transmitter, which if you stay within the commercial broadcast bands and avoid the frequencies of national and local stations, although still probably illegal, is *FAR* less likely to attract attention.

I'm ashamed I didn't point this out myself. Whenever you are transmitting, you must filter the hell out of your output signal, and then test it just to make sure. For civilty and for safety.

[ali6x944]

THANKS GUYS FOR ALL THE INFO...
IT WAS REALLY HELPFUL, THANKS ! 

[tggzzz]

i was working on a battery powered AM radio reviser/transmitter, i want to transmit for a minimal distance of 20m and across all frequencies in the AM band.

In many jurisdictions, transmitting is illegal unless you have the correct licence. In order to get the licence, you will have to demonstrate that you know how to build and operate transmitters in such a way that you do not interfere with other users of radio systems. Currently it is clear that you do not have that knowledge.

Whether or not you have a licence, interfering with other users is illegal in most jurisdictions.

I suggest you contact some local radio hams and ask them for advice.

[uncle_bob]

Hi

Switchers create hash both on their inputs and on their outputs. It is common to see designs that focus on one and ignore the other. High power transmitters create major RF fields. Again you have things running around that need to be taken care of. This all creates a third problem, switchers depend on fast signals, RF is a fast signal. Get RF into the wrong part (input, output, control) of a switcher and really nasty things can happen. Much of the testing of a transmitter is involved in tracking this sort of thing down and taking care of it. Been there / spent a *lot* of time doing it.

Bob

[Ian.M]

Its not all gloom and doom - very low power short range transmitters *MAY* be legal on certain bands in your country.

e.g. in the USA, under part 15 of the FCC rules, AM (medium wave) transmitters of up to 100mW output power are permitted without licence, but under several stringent restrictions.   Home designed and constructed ones are permitted but one must be extremely careful to make sure your design complies with the regulations as the possible fines are extremely punitive.

However there is *NO* substitute for local knowledge of the applicable laws, and if your government has authoritarian tendencies, the chances are high that they are not permitted, and even enquiring about them may not be a particularily safe thing to do.