Remove AC hum from FM transmitter power supply

[betwixt]

There is something else to check:

Does the hum disappear if you short out the connection to the microphone AT THE TRANSMITTER END of the cable? If it does, make sure you are using a screened audio grade cable to the microphone and the screen is connected to ground and the negative side of the microphone. If an unsuitable cable is used of the screen/inner are reversed you will get capacitively coupled hum at the input to the preamp stage.

Brian.

 

[boylesg]

There is something else to check:

Does the hum disappear if you short out the connection to the microphone AT THE TRANSMITTER END of the cable? If it does, make sure you are using a screened audio grade cable to the microphone and the screen is connected to ground and the negative side of the microphone. If an unsuitable cable is used of the screen/inner are reversed you will get capacitively coupled hum at the input to the preamp stage.

Brian.

Thankyou susan and Brian....greatly appreciate your patience.

So any oscillation in the supply voltage, e.g. due to ripple, will be amplified and I will hear it?

So the caps suggested by audio don't seem to make any difference - I soldered then to the PCB caps on the back side of the board.

And the microphone cable had a single insulated core with some form of braided conductor around it - similar to coax cables.

I am assuming that is what you mean by screened?

And the braid forms the second 'wire' that is connected to to ground on the PCB and the electret.

 

[Audioguru]

Instead of the 12V switched-mode power supply with a voltage too high, you should try a normal 9V power supply that is not switched mode but has a fairly heavy low frequency transformer. If the supply voltage is regulated at 9V then it will work best.

 

[boylesg]

Instead of the 12V switched-mode power supply with a voltage too high, you should try a normal 9V power supply that is not switched mode but has a fairly heavy low frequency transformer. If the supply voltage is regulated at 9V then it will work best.

I tried a 9V iron core transformer type plug pack (you can tell by the weight) and removed my linear voltage regulator (9V) I had.

Still get loud hum over the radio though.

 

[FvM]

I guess you don't want to remove the linear voltage regulator, why did you?

Providing sufficient voltage margin for the regulator is required however. If you are not able to provide a clean DC, filtered and regulated according to state of the art, give it up.

 

[boylesg]

I guess you don't want to remove the linear voltage regulator, why did you?

Providing sufficient voltage margin for the regulator is required however. If you are not able to provide a clean DC, filtered and regulated according to state of the art, give it up.

I was powering the circuit from my 12V SLR...which is why I added the linear 9V regulator.

Without thinking too much I just bought a 12V plug pack despite 9V ones being available.

But I had another 9V plug pack laying around that I had salvaged at some point.

Since I can't get rid of the hum from any plug pack, I have the circuit powered from 4 x NiMH D cells in series with a DC-DC boost converted that takes the voltage up to 9V.

I would still like to get to the bottom of this hum for future reference though.

Oh I have also noticed that, if the circuit is to close to any mains power chords, I also get the hum on my mobile phone FM app.....not as intense though.
What I might do (not now though) is fire up the circuit with the plug pack and use my oscilloscope to try and identify this hum - will post some photos of the screen.

Might make it easier for you experts to diagnose the problem.

 

[Audioguru]

I was powering the circuit from my 12V SLR...which is why I added the linear 9V regulator.

We were wondering how much filtering there is in the power supply and now you say you had a voltage regulator that is supposed to completely kill power supply hum! Maybe the input to the voltage regulator had enough hum to cause it to dropout which makes it worse.

Oh I have also noticed that, if the circuit is to close to any mains power chords, I also get the hum on my mobile phone FM app.....not as intense though.
What I might do (not now though) is fire up the circuit with the plug pack and use my oscilloscope to try and identify this hum - will post some photos of the screen.

Any high gain audio circuit picks up hum when it is near a mains power cord unless the circuit is inside a shielding metal box connected to the circuit ground.
This circuit has 47k plus 4.7K from 9V powering the mic. An electret mic needs a few volts across it and draws 0.5mA so if the resistors have 0.5mA then the mic voltage is 9V - (0.5mA x 51.7k)= way less than zero. Therefore the mic does not have enough voltage and the 47k resistor value is so high that it allows mains hum to be picked from a nearby mains power cord. The 4.7k resistor powering the mic and preamp should be 1k ohms and the 47k resistor powering the mic should be 10k ohms.

Before you change parts values:
1) Please post a photo of the FM transmitter circuit board so we can see how far apart the parts are.
2) Post a schematic of the 9V voltage regulator you used.
3) Measure the DC voltage across the mic.
4) Post a photo of the 'scope looking at the output of the 12V power supply.
5) Post a photo of the 'scope looking at the output of the voltage regulator which is the power to this circuit.

 

[betwixt]

It can be a surprisingly complicated problem to diagnose.

You might find an oscilloscope, at least used in the normal way, actually makes the problem worse. The hum is the result of a difference in 50Hz (or harmonics) voltage between two parts of the circuit. The difficulty is in finding where those parts are, and there are probably many.

The most obvious source will be a ripple in the DC voltage from the power source but it is usually very small, especially when the current you are drawing will be tiny. A switch mode supply generates it's own ripple but it is generally at switching frequency (>20KHz) and a low frequency implies the AC input side of the switcher circuit is faulty. In your schematic, any ripple on the supply will not only change the frequency (FM) but the amplitude as well (AM).

The other major source will be from capacitive coupling to other sources. The whole building will be filled with a complicated electrostatic and electromagnetic fields created by currents flowing in the power wiring. Some of that gets conducted in to the wiring of your transmitter. It will be indiscriminate, getting in everywhere and the only way to stop it is to completely shield your whole project. Although it may seem difficult to grasp the concept, it is quite likely that there is considerable AC voltage coming out of the adapter, whether a switch mode or iron transformer type. The voltage ACROSS the output wires may be steady DC but consider that both the wires may be floating in unison at high voltage AC. The amount of isolation between input and output of the adapter may be adequate to make it safe to touch but that doesn't mean the voltage isn't there. Even a tiny amount of high voltage coupling will introduce hum, especially in such a sensitive circuit. I'm not sure if your country uses AC plugs with an Earth pin but it would be useful to try linking the negative side of the output directly to Earth to prevent the output floating.

If you want proof of the amount of coupling by proximity, use your oscilloscope and touch only the tip of the probe. The trace will show an approximate sine wave, think about where that voltage comes from. You might find your body is actually carrying quite a high AC voltage! Then touch the oscilloscope ground as well and note the difference.

Brian.

 

[boylesg]

It can be a surprisingly complicated problem to diagnose.

You might find an oscilloscope, at least used in the normal way, actually makes the problem worse. The hum is the result of a difference in 50Hz (or harmonics) voltage between two parts of the circuit. The difficulty is in finding where those parts are, and there are probably many.

The most obvious source will be a ripple in the DC voltage from the power source but it is usually very small, especially when the current you are drawing will be tiny. A switch mode supply generates it's own ripple but it is generally at switching frequency (>20KHz) and a low frequency implies the AC input side of the switcher circuit is faulty. In your schematic, any ripple on the supply will not only change the frequency (FM) but the amplitude as well (AM).

The other major source will be from capacitive coupling to other sources. The whole building will be filled with a complicated electrostatic and electromagnetic fields created by currents flowing in the power wiring. Some of that gets conducted in to the wiring of your transmitter. It will be indiscriminate, getting in everywhere and the only way to stop it is to completely shield your whole project. Although it may seem difficult to grasp the concept, it is quite likely that there is considerable AC voltage coming out of the adapter, whether a switch mode or iron transformer type. The voltage ACROSS the output wires may be steady DC but consider that both the wires may be floating in unison at high voltage AC. The amount of isolation between input and output of the adapter may be adequate to make it safe to touch but that doesn't mean the voltage isn't there. Even a tiny amount of high voltage coupling will introduce hum, especially in such a sensitive circuit. I'm not sure if your country uses AC plugs with an Earth pin but it would be useful to try linking the negative side of the output directly to Earth to prevent the output floating.

If you want proof of the amount of coupling by proximity, use your oscilloscope and touch only the tip of the probe. The trace will show an approximate sine wave, think about where that voltage comes from. You might find your body is actually carrying quite a high AC voltage! Then touch the oscilloscope ground as well and note the difference.

Brian.

Aus Brian - yes we do have earth pins on a wall sockets.

The only way I could earth the neutral pin is via a wire ling on the back of the plug pack itself.....not sure that would be safe.

What about earthing the GND on DC side? Although that would have to be via a separate wall socket.

Suppose that why all AM/FM radio, powered from mains sockets, do all this conveniently inside the unit itself.

 

[betwixt]

Suppose that why all AM/FM radio, powered from mains sockets, do all this conveniently inside the unit itself.

Exactly! In your case it's the transmitter but the principle is the same.

I think it would be best for you to experiment, try a wire link from GND to the Earth pin and if it improves things, perhaps modify the adapter or plug it in through a short extension cable that gives you access to the Earth connection. Unfortunately, many adapters have plastic Earth pins these days to make them cheaper to manufacture and it obviously prevents an electrical connection.

You might (stress MIGHT) be able to attach a temporary Earth connection to a water pipe if it is copper all the way into the ground. Sadly, if your building was erected in the last 20 years or so, the chances are you have plastic pipes at least some of the way and hence no conductivity.

Brian.

 

[Easy peasy]

Your simple FM transmitter has very little 100Hz rejection, power it from a 9V battery, and see if the hum goes up when you move your hand closer to it.

no doubt the hum will go up if you move the ckt closer to any wire carrying 50Hz mains current - this is because there is no shielding on your simple ckt with its simple construction.

A power supply with 100Hz ripple will make the problem worse, also due to the Common Mode 50Hz effect on a switch-mode type rather than a 50Hz transformer type power supply.

This is why such simple RF xtor ckts are not used int he professional / industrial sector ....

 

[boylesg]

Exactly! In your case it's the transmitter but the principle is the same.

I think it would be best for you to experiment, try a wire link from GND to the Earth pin and if it improves things, perhaps modify the adapter or plug it in through a short extension cable that gives you access to the Earth connection. Unfortunately, many adapters have plastic Earth pins these days to make them cheaper to manufacture and it obviously prevents an electrical connection.

You might (stress MIGHT) be able to attach a temporary Earth connection to a water pipe if it is copper all the way into the ground. Sadly, if your building was erected in the last 20 years or so, the chances are you have plastic pipes at least some of the way and hence no conductivity.

Brian.

You know what I could do potentially is the following.....

1) Get one of my salvaged 9V DC plug packs.
2) ***** it open and remove the circuit board and transformer
3) Obtain an empty Jaycar plug pack enclosure
4) Mount the transformer and circuit board in the above
5) Do the earth connection inside the new plug pack.

All wall sockets have a standard pin config, so I guess I can use a multimeter on one of my spares to work out which side is neutral. I can never remember just looking at the wall socket in the wall.

Don't want to connect the active to ground

 

[betwixt]

All wall sockets have a standard pin config, so I guess I can use a multimeter on one of my spares to work out which side is neutral. I can never remember just looking at the wall socket in the wall.

Warning!
The neutral has nothing to do with it, in fact you can reverse live (aka 'line') to the transformer and it will make no difference as the voltage alternates (AC) anyway. What you need to link is the Earth pin on the AC socket to the Ground side of the DC output. Live and neutral to the transformer primary, Earth to the DC output. Whatever you do, do not connect anything to either the live or neutral wires themselves as it poses a serious electrocution risk.

Brian.

 

[Audioguru]

I have many radios powered from the mains and most of them do not have an "earth" connection. They produce very low audio frequencies but no hum. Their circuit boards are not inside a shielding metal case. The circuit boards have compact layouts so there are not long traces that are antennas that pickup mains hum.

 

[boylesg]

Warning!
The neutral has nothing to do with it, in fact you can reverse live (aka 'line') to the transformer and it will make no difference as the voltage alternates (AC) anyway. What you need to link is the Earth pin on the AC socket to the Ground side of the DC output. Live and neutral to the transformer primary, Earth to the DC output. Whatever you do, do not connect anything to either the live or neutral wires themselves as it poses a serious electrocution risk.

Brian.

Oh crap, I could have sworn you said something about the neutral pin.

Earthing the DC GND makes more intuitive sense to me - that is what some do with tesla coils.

I have not done anything yet - just considering trying it at some point but just understanding what you are proposing at this stage.

 

[d123]

Hi,

AC plugs have no polarity as such, boylesg, have you ever had to think about which way to plug one in? - Another matter is which pin in the plug socket is the live wire and which is the neutral. The polarity-tester screwdriver or whatever it's called can tell you that point. Think of AC as going backwards and forwards through both wires.

Just want to add, and spoil the fun a little: Unless you have done the building wiring yourself, who's to say what interesting things the Earth cable does, or even how "Earth" is created? - Earth may not be at 0 potential and in some installations can even be hazardous - an unwitting person may even be accidentally connecting a Lord knows how live wire to the DC ground.

It is not really advisable to connect AC Earth to DC ground unless you genuinely know what you are doing. Just a thought. By all means do connect your circuit up, millions of other pèople do without checking anything and survive, just be aware of the bigger picture that Earth may not be (entirely) Earth unless certified so.

 

[betwixt]

The problem here though is that the output wires of the adapter are 'floating' at some potential above Earth and therefore any capacitance to objects nearer Earth potential (most things) conducts some signal into the transmitter wiring. It is true that a wrongly wired AC outlet could be dangerous but I assume regulations in Boylesg country cover safety checking.

Earth may not be 0V, in fact measuring from one physical ground connection to another nearby will show some voltage difference but I'd bet it's a lot less than the adapter output which could be at hundreds of volts above Earth.

Brian.

 

[Easy peasy]

refer to post # 9.

 

[boylesg]

Hi,

AC plugs have no polarity as such, boylesg, have you ever had to think about which way to plug one in? - Another matter is which pin in the plug socket is the live wire and which is the neutral. The polarity-tester screwdriver or whatever it's called can tell you that point. Think of AC as going backwards and forwards through both wires.

Just want to add, and spoil the fun a little: Unless you have done the building wiring yourself, who's to say what interesting things the Earth cable does, or even how "Earth" is created? - Earth may not be at 0 potential and in some installations can even be hazardous - an unwitting person may even be accidentally connecting a Lord knows how live wire to the DC ground.

It is not really advisable to connect AC Earth to DC ground unless you genuinely know what you are doing. Just a thought. By all means do connect your circuit up, millions of other pèople do without checking anything and survive, just be aware of the bigger picture that Earth may not be (entirely) Earth unless certified so.

Yes I do know that.

But Brian mentioned something about Earthing and the neutral pin.

I apparently I incorrectly linked the the two and (to be honest) I didn't give much thought to it given that Brian is clearly very experienced.......possibly he just inadvertently mixed up his terms AC and DC terms and said 'neutral' when he meant 'GND'.

If you explicit say 'GND' and '-' then I immediately associated that with DC since that term is often used in DC schematics.

If you say 'neutral', 'earth' and probably 'ground' then I immediately associated these with AC circuits and AC wall sockets.

That is why I later commented that earthing the neutral pin did not make much intuitive sense to me.

When Brian later pointed out my incorrect assumption about his comment, and explicitly clarified about Earthing the DC GND pin, it did make intuitive sense to me since I have been dabbling in Tesla coils also.

Look I am not going to do any of this until I have a clear understanding about what it is I am doing.

And in hindsight I am not sure I will even bother because powering my little FM transmitter run off a wall plug means that it will be less conveniently portable for Mum.

I have it running of 4xD cells with a DC-DC boost converter module to 9V.

I have another FM transmitter kit I bought of ebay to try and it was running off a 9V battery for half a day without any signs of the battery losing power.

So I suspect my 4xD cells will last for ages anyway, and they are rechargeable.

- - - Updated - - -

Hi,

AC plugs have no polarity as such, boylesg, have you ever had to think about which way to plug one in? - Another matter is which pin in the plug socket is the live wire and which is the neutral. The polarity-tester screwdriver or whatever it's called can tell you that point. Think of AC as going backwards and forwards through both wires.

Just want to add, and spoil the fun a little: Unless you have done the building wiring yourself, who's to say what interesting things the Earth cable does, or even how "Earth" is created? - Earth may not be at 0 potential and in some installations can even be hazardous - an unwitting person may even be accidentally connecting a Lord knows how live wire to the DC ground.

It is not really advisable to connect AC Earth to DC ground unless you genuinely know what you are doing. Just a thought. By all means do connect your circuit up, millions of other pèople do without checking anything and survive, just be aware of the bigger picture that Earth may not be (entirely) Earth unless certified so.

While I take it that this is why Tesla coil 'earths' and house 'earths' have to be a minimum depth?

 

[betwixt]

While I take it that this is why Tesla coil 'earths' and house 'earths' have to be a minimum depth?

From an electrical grounding point of view, it would be ideal if the whole planet was made of superconducting material (cue Star Trek...) and everywhere had the same reference voltage to work against. However, that isn't even close to reality and the best we have is whatever is beneath our feet. Being kind to our planet, the moisture and dissolved conductive elements in the crust are reasonably good conductors, at least in most inhabitable places. The more contact area your house (or Tesla coil) earth has with the physical ground, the more likely it is to have good conductivity to elsewhere. The minimum depths for Earth connections are to maximize the likelyhood of a good connection.

You probably also have a wired connection from the AC outlet Earth pin to your nearest distribution transformer where it will be linked to neutral and also given a good connection to 'real planet' Earth rods. This is for electrical safety purposes and to allow safety devices (RCCB, ELCB etc) to operate but it does provide you with a reasonably clean ground reference.

Taking your Tesla coil as an example although the same principle applies to your transmitter, it generates high voltage between two places. If you didn't Earth one of them, it would be as 'live' as the other. You probably don't want high voltage arcs from the bottom of the coil, probably taking the easiest electrical route which would be through the coil primary and driver circuits!

Brian.