Hi all! I'm a new electrician and I'm making a radio for myself. I have a radio receiver that works on 5 Volts and has 2 outputs of 3 Watts each. I've recently bought a D-Class amplifier that works on 12-24 Volts and has 2 outputs of 80 Watts each. I have a problem with the EMI effect. On "not so good" stations, when I connect the speakers to the amp's output, the radio signal falls. I can see it from the radio board and I can also hear it from the speaker. If I connect the speakers directly to the radio board, the signal is perfect. I've read that this is a common EMI problem and I need an output filter. I've also heard about the ferrite chokes but I don't know how to choose the right ferrite or how to use them. Can you please help me find a solution? It means a lot to me. Thank you very much!!!
EMI (ElectroMagnetic Interference) is an important problem for the combination radio -> class D amplifier. Could you tell what class D amplifier you bought?
Your class D amplifier should already have an output filter, in particular if it is 24V supply/80W output. The problem you observe is likely to be the radiated noise (by air) that partly paralyzes your radio receiver. A class D amplifier generates square-wave signals in the range 300KHz-600KHz and, as they are square-wave shaped, they contain a lot of higher frequency components. Perfect to disturb a radio receiver. Radiation shielding may be the solution to your problem.
Try the following:
1) Use the radio with a less strong radio station and the speaker connected to your radio. Connect a second speaker to the class D amplifier (a class D amplifier should only be turned ON with a load connected). Put your class D amplifier some meters away from the radio and turn the class D amplifier ON. It should make your radio reception less good. The closer the class D amplifier is to the radio, the worse the radio reception should be.
2) Use the headphone output on your smart-phone as input to the class D amplifier and make sure that the class D amplifier works well.
Your class D amplifier should already have an output filter, in particular if it is 24V supply/80W output. The problem you observe is likely to be the radiated noise (by air) that partly paralyzes your radio receiver. A class D amplifier generates square-wave signals in the range 300KHz-600KHz and, as they are square-wave shaped, they contain a lot of higher frequency components. Perfect to disturb a radio receiver. Radiation shielding may be the solution to your problem.
Try the following:
1) Use the radio with a less strong radio station and the speaker connected to your radio. Connect a second speaker to the class D amplifier (a class D amplifier should only be turned ON with a load connected). Put your class D amplifier some meters away from the radio and turn the class D amplifier ON. It should make your radio reception less good. The closer the class D amplifier is to the radio, the worse the radio reception should be.
2) Use the headphone output on your smart-phone as input to the class D amplifier and make sure that the class D amplifier works well.
I use this amplifier I bought from Aliexpress: 12V 24V 80W+ 80W Stereo Audio Dual channel Digital Audio TPA3116D2 Treble Bass Regulating power amplifier Board-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group
I agree with you, this amplifier has got no output filter. That is unusual with the TPA3116D2 chip. Then, the speaker cables will work as antenna for the modulation frequency (300KHz-600KHz).
Try to see if the radio reception is disturbed even with the amplifier meters away.
Try to see if the radio reception is disturbed even with the amplifier meters away.
Well, when I put the amp some meters away, the signal of the radio stations is much better! But not the best signal. Do you know what filter do I need to buy or how to make one?
For that kind of emi problem I suggest ferrite clamps - the likes you find sometimes glued to USB-cable.
Any cable connected to the class-d-amp should pass a ferrite clamp that is located close to the class-d-amp.
The best option is not one ferrite clamp/cable but a single big common clamp for all cables.
Any cable connected to the class-d-amp should pass a ferrite clamp that is located close to the class-d-amp.
The best option is not one ferrite clamp/cable but a single big common clamp for all cables.
I tested one ferrite clamp from an old power supply. I cut the wires from the supply and then used them to connect the amp's output with the speakers but something strange happened: there was no sound from the speaker! Even if I connect the speaker's with the output's screws that tight the cables on the output. It looks like this ferrite cuts the audio or something. Is there different values of ferrite? Cause I don't know about them a lot.
Input isolation helps. Just the fact of breaking the D.C. path makes a difference. These are some of the best that I have found.
JaJaBor Female to Male 3.5mm Aux Audio Noise Filter Ground Loop Noise Isolator Eliminate Car Electrical For Car AUX Speaker-in Bluetooth Car Kit from Automobiles & Motorcycles on Aliexpress.com | Alibaba Group
JaJaBor Female to Male 3.5mm Aux Audio Noise Filter Ground Loop Noise Isolator Eliminate Car Electrical For Car AUX Speaker-in Bluetooth Car Kit from Automobiles & Motorcycles on Aliexpress.com | Alibaba Group
This may be useful to reduce humming noise due to ground loop. But not in case of radio interference in the 100MHz range.
You could try making sure the amp chassis is grounded which it should be anyway for safety.
It could be residual carrier from class d amp getting onto the speaker leads and radiating RF, in which case a emi bead filter around the cable right up at the amp could help.
It could be residual carrier from class d amp getting onto the speaker leads and radiating RF, in which case a emi bead filter around the cable right up at the amp could help.
I have experienced if the D.C. path is broken to the input, the interference is much less. Less than a direct connection from the source to the front of the amp. (Or from it)
If any interference passes through, it's radiated, not through a direct path via the shield of the input cable(s). The isolation transformers provide a shunt which helps with the reduction. The input impedance of the amp is lowered, being it's connected to a coil, not just floating on a couple of caps.
Works for me.. 😉
The amp has an NS4358 chip. FM tuner right next to it. Same set up with a nasty PAM8610 board. Amp/Tuner operating from the same battery. The NS4358 board uses a voltage regulator to reduce the 12 volts to 5 volts. Works like a charm with the isolation inserted. Not at all without it..
*Note the little piece of orange/white wire from the tuner. That's the FM antenna!
*Note the little piece of orange/white wire from the tuner. That's the FM antenna!
Last edited:
I have used a very similar module here:
http://www.diyaudio.com/forums/full-range/274036-fast-class-boombox-build.html#post4319070
Supplies were isolated so there was only signal path ground connections.
http://www.diyaudio.com/forums/full-range/274036-fast-class-boombox-build.html#post4319070
Supplies were isolated so there was only signal path ground connections.
I'm using other power supply to the radio board and other for the amplifier. Is this causing the problem? And something more: how can I ground the amplifier? It's hasn't any kind of hole or something (like the radio board has).
Maybe with ferrites like Nuprime ST-10
The best are the Würth 150 kHz ferrites. I have a lot of in my systems. And Schaffner RF/EMI filters too.
The best are the Würth 150 kHz ferrites. I have a lot of in my systems. And Schaffner RF/EMI filters too.
Trust me, the best RF ferrite to attenuate < 1 MHz (aka audio band) is the Würth 150 kHz ferrite, without any doubt. And the Schaffner filters have the best attenuation graphs too.
Or you can read the specs of the ferrites like I had three years ago.
Or you can read the specs of the ferrites like I had three years ago.
Perfect! Thank you very much. Can I ask you something more? The ferrite that it's on a power supply for a laptop isn't good for a project like this? I mean, does it have different values from the one you have suggested me to buy?
You can use it but it will work with EMI, aka f > 1 MHz like the cheap ferrites. Video band.
Our problem is always f < 1 MHz.
All I have are of 8 mm. Only one is 12 mm (diameter), in the fat cable of a power strip.
You can make too:
With
Schaffner FN 9244B:
FN 9244
and Schurter IEC C13 connector. Unshield power cable is better idea.
My estimation is the ferrite is like almost 1/2 FN 9244B.
B = medical
Our problem is always f < 1 MHz.
All I have are of 8 mm. Only one is 12 mm (diameter), in the fat cable of a power strip.
You can make too:
An externally hosted image should be here but it was not working when we last tested it.
With
Schaffner FN 9244B:
FN 9244
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
and Schurter IEC C13 connector. Unshield power cable is better idea.
My estimation is the ferrite is like almost 1/2 FN 9244B.
B = medical
Last edited:
- Status
- Not open for further replies.