PAM8610 breakout board has FADE hardwired to the ground so it fades in the sound on power up. I thought this was a good feature to mute undesired popping on power up of the source, but the PAM board has annoying crackling on start up itself. Was I unlucky with my purchase or is it “the feature”? See attached video for the demonstration. PAM8610 crackling on start up - YouTube
Can you asure that the undesired sounds are caused by the board itself removing all sound sources?. Perhaps your unit is failed.
This thrash is a class D without RF integrator filters, so you can expect any class of problems with sensitive devices like preamplifiers, tuners, phono equalizers, TV and the like.
This thrash is a class D without RF integrator filters, so you can expect any class of problems with sensitive devices like preamplifiers, tuners, phono equalizers, TV and the like.
Last edited:
Hi, yes it is crackling exactly like on the video without anything connected to the input. By RF integrator filters, do you mean ferrite chokes? As I understood the breakout board follows data sheet, not sure what is missing.
There are two kinds of class D amps, those who use a low pass filter to integrate the PWM pulses into an analog signal to be feed to the speakers, and those that don't do it, spraying rf elsewhere including the amp inputs. The lower cost of such units that lacks the filter is paid in rf radiated and heating in the magnetic pieces of the spk lowering its efficiency and offending any electronic device close to them. Search in the web for "filterless class d amp".
Last edited:
Thanks for explaining. I found that if I switch amplifier off the FM module is able to find more stations than with amplifier on. If I enclose it in metal box and ground it, will it reduce RF interference? EDIT those stations that are picked up with amp off are mostly just noise but still
Last edited:
Effectivelly. Not only enclose amp in grounded metal case, also try to use shielded twisted pair for spk and if possible, also ground the metal parts of the speaker itself.
Try to read about "Faraday cages", the physics inside it and how to proper make them.
I understand that those sets can be selled in desastrous countries like mine, but I can't imagine how in advanced places like yours, with stringent EMI regulations, how and why those **** is not entirelly prohibited.
Pardon by my insistence, but as ham radio from 1987 HF bands are lost because of myriads of pulsed emitters of rf like amps, smps, led lights, monitors, tv, etc, which lacks entirelly of emi filtering. Switch on a 40mts receiver nowadays is impossible because of the excessive noise in the ether.
Try to read about "Faraday cages", the physics inside it and how to proper make them.
I understand that those sets can be selled in desastrous countries like mine, but I can't imagine how in advanced places like yours, with stringent EMI regulations, how and why those **** is not entirelly prohibited.
Pardon by my insistence, but as ham radio from 1987 HF bands are lost because of myriads of pulsed emitters of rf like amps, smps, led lights, monitors, tv, etc, which lacks entirelly of emi filtering. Switch on a 40mts receiver nowadays is impossible because of the excessive noise in the ether.
The PAM8610 datasheet contains useful informations both about the suppression of power on/off noise and the radio frequency emission. I've bought a few of those 1.50 $ PAM8610 modules myself because it is a useful 8+8W amplifier that does not need any thermal management and can be easily paired with a generic microcontroller or even a classic audio source. It is basically the audio power section of a low end flat screen TV set or bluetooth speaker: audio input and volume control are still analog.
The useful part of the datasheet for DIY builders is after the "boring" technical data. To avoid the power on/off noise, the /SD (shutdown) signal should be pulled low at turn-on and before turn-off. This is the pad marked "SW" on the breakout board, and it should be driven by a microcontroller pin, or a RC relay network, or even a mechanical switch. Power consumption in shutdown mode is a few microwatts, so this chip is meant to have the power supply always on and the /SD signal is the on/off control to be used instead. That's almost a given on any power IC with this target use case.
This still does not ensure a quiet operation, because the radio frequency emission (EMI) and the power supply decoupling needs to be addressed. Both functions are required for proper operation but are missing on the breakout board. To block unwanted noise/oscillations on the power supply, a filter capacitor need to be soldered directly at the power supply input pads of the board. A ceramic capacitor for radio frequency suppression is already on the board; a standard electrolitic capacitor between 220uF and 1000uF is enough to take care of spike suppression. On the video I see an auxiliary board with volume control trimmers that has been fitted below the PAM8610 board; the capacitor may be already there but I am unable to see it.
The EMI suppression is the most difficult aspect to address properly, because a oscilloscope and some radio frequency experience are required to obtain a perfect result. Anyway, the datasheet gives two very good starting point suggestions: 1. add a ferrite bead on each speaker output terminal (and a 200pF capacitor to ground after the bead); 2. keep wire lenght from the PAM8610 to the speaker below 20cm. The EMI diagrams listed on the datasheet after this crude EMI filter shows a barely enough result, but this will remove most of the digital noise that now is radiating back on the analog input pins. Ferrite beads may be recovered from discarded power supplies. When space is not a issue, I use VK200 inductors instead of ferrite beads, because I have a large stock of VK200 bought at low cost many years ago. Maybe a member here may suggest an equivalent modern part that can be bougt from online distributors. The PAM8610 oscillator works at 250 KHz.
To reduce the noise, it is also useful to use twisted or shielded wire from the board to the speaker; keep this wire at some distance from the amplifier input.
The useful part of the datasheet for DIY builders is after the "boring" technical data. To avoid the power on/off noise, the /SD (shutdown) signal should be pulled low at turn-on and before turn-off. This is the pad marked "SW" on the breakout board, and it should be driven by a microcontroller pin, or a RC relay network, or even a mechanical switch. Power consumption in shutdown mode is a few microwatts, so this chip is meant to have the power supply always on and the /SD signal is the on/off control to be used instead. That's almost a given on any power IC with this target use case.
This still does not ensure a quiet operation, because the radio frequency emission (EMI) and the power supply decoupling needs to be addressed. Both functions are required for proper operation but are missing on the breakout board. To block unwanted noise/oscillations on the power supply, a filter capacitor need to be soldered directly at the power supply input pads of the board. A ceramic capacitor for radio frequency suppression is already on the board; a standard electrolitic capacitor between 220uF and 1000uF is enough to take care of spike suppression. On the video I see an auxiliary board with volume control trimmers that has been fitted below the PAM8610 board; the capacitor may be already there but I am unable to see it.
The EMI suppression is the most difficult aspect to address properly, because a oscilloscope and some radio frequency experience are required to obtain a perfect result. Anyway, the datasheet gives two very good starting point suggestions: 1. add a ferrite bead on each speaker output terminal (and a 200pF capacitor to ground after the bead); 2. keep wire lenght from the PAM8610 to the speaker below 20cm. The EMI diagrams listed on the datasheet after this crude EMI filter shows a barely enough result, but this will remove most of the digital noise that now is radiating back on the analog input pins. Ferrite beads may be recovered from discarded power supplies. When space is not a issue, I use VK200 inductors instead of ferrite beads, because I have a large stock of VK200 bought at low cost many years ago. Maybe a member here may suggest an equivalent modern part that can be bougt from online distributors. The PAM8610 oscillator works at 250 KHz.
To reduce the noise, it is also useful to use twisted or shielded wire from the board to the speaker; keep this wire at some distance from the amplifier input.
Attachments
Last edited:
I want decide to buy one PAM module (but2x15W version), i think will be great If install it in 2 speakers 4" subwoofer, with 12Volts adpator supplies.
Am i right (my opinion)?
Any recomendations or suggestions?
Am i right (my opinion)?
Any recomendations or suggestions?
Yes. And enclose them in a well ventilated but well shielded case, to reduce EMI to the othet devices. I insist in grounding speaker chassis too.