Is this written by an AI? If not, I salute your language in writing. If it was, you should have stated that by using quotation marks and a reference. This is the way to handle AI support in the "new world intelligence" by humans.... 😉
"Firstly, it is indeed a feedback loop, and its primary purpose is to maintain the oscillation of the Sigma-Delta amplifier." Well, as this is not only the oscillation carrier but also the information carrier, isn't it hard to distinguish? Low passed or not....
But does it matter really - does it measure good? Does an implementation sound good?
Is it cheap, stable and great energy efficiancy?
//
"Firstly, it is indeed a feedback loop, and its primary purpose is to maintain the oscillation of the Sigma-Delta amplifier." Well, as this is not only the oscillation carrier but also the information carrier, isn't it hard to distinguish? Low passed or not....
But does it matter really - does it measure good? Does an implementation sound good?
Is it cheap, stable and great energy efficiancy?
//
Designing a high-power version of the amplifier using the LMG1210 and GaN FET EPC2001C as the driver stage, with a 48V single power rail, delivering up to 150W into an 8-ohm load. Are there any recommendations for better GaN FETs or drivers?
This is a mono board - right? Size seems small... what dimensions?
Good luck!! Please do report findings and progress - I think this is an interesting project. What seems to be a design that don't have the rising distortion at higher frequency is one aspect that is very promising. Would be interesting to see the ordinary measurements i.e. frequency sweeps of FR, distortion, noise, phase and load dependency.
I'm regret I can't support in component choices...
//
Good luck!! Please do report findings and progress - I think this is an interesting project. What seems to be a design that don't have the rising distortion at higher frequency is one aspect that is very promising. Would be interesting to see the ordinary measurements i.e. frequency sweeps of FR, distortion, noise, phase and load dependency.
I'm regret I can't support in component choices...
//
designed a aluminum house for the board for heat dispassion and shielding. the overall dimension is 68mm x 47mm x 9.6mm, it is pretty small.
Last edited:
actually, this project is amp module, an external filter can be connected to. I am trying to set the modulation frequency above 3MHz, no filter is needed, if the frequency can't reach as high as expected, then a external filter can be considered to filter out the high frequency components.
Last edited:
Wouldn't it be prudent with a filter irrespectable of switching frequency? And most efficient if it is as close to the generator as possible?
There are probably som RFI rules that this module breaks by its own and building an amp that adheres to the rule will be harder if the filters is not on the same circuit board / ground plane?
//
There are probably som RFI rules that this module breaks by its own and building an amp that adheres to the rule will be harder if the filters is not on the same circuit board / ground plane?
//
The design screams that you have RF background but these words tell the opposite. I'm confused. 😆
Great point! this is the same question I thought about. to maintain the compact size of the module, the output connector (the rating current is 15A) has 3 contacts, 2 of the sides is for the audio output, the middle one is ground. if put the external filter close to the module, I think there is not too much difference than integrate the filter into the module, if keep the area of the filter loop as small as possible.
RF behaves strange - there could actually be a considerable difference between a filter residing on a ground plane and hanging just outside.
//
//
I got a 100w single ended amp.to do 2.2 mhz in self oscillation mode. Sounded great but had very little overload capability. 5uH choke killed the carrier. Freq went as low as 900khz at clipping but it was class A performance below 70% power.
For the self-oscillation class D AMP, the maximal PWM duty cycle should not larger than ~80%.
Add 2 level LC filters. 2.15uH+1uF and 2.15uH+0.33uF, variation of responds at 20kHz is less than 0.8dB with 2Ohm to 8Ohm load. 3dB cutoff is around 60kHz.
Looking good but I'm not knowledgable in implementation. Does this board have a load/impedance dependent frequency response? Variation described above is for filter section (stand alone) only - right?
Some designs bring the inductors inside their feedback loop in order to get them out of "the equation".....
There has been a number of inductor comparison shoot-outs made - with, as I recall it, quite some differences?
How much longer did the unit become?
Why did you choose to make it double? Was a steeper filter needed? Typically we see two inductors per channel?
//
Some designs bring the inductors inside their feedback loop in order to get them out of "the equation".....
There has been a number of inductor comparison shoot-outs made - with, as I recall it, quite some differences?
How much longer did the unit become?
Why did you choose to make it double? Was a steeper filter needed? Typically we see two inductors per channel?
//
The frequency response depends on the impedance of the load because of Q factor of the LC filter. lower load has lower Q factor and the frequency response near the cut off frequency different accordingly. The LC filter design for this AMP module works with load from 2 to 8 Ohms, the frequency response at 20kHz is +- 0.8dB. BTW, the Class D AMP without LC filter is independent from load.
Indeed, the inductor is a key part for Class D AMP, it may introduce significant distortion if not select the right one especially for high frequency. the inductor I selected here following some disciplines: lower height, high current, low DC resistance, high Self-Resonant frequency etc. 2 filters are designed is for better phase responds (although phase responds is not important for the audio) and steeper filter curve too.
the board is longer by 23.4mm, the housing will be longer accordingly.
Indeed, the inductor is a key part for Class D AMP, it may introduce significant distortion if not select the right one especially for high frequency. the inductor I selected here following some disciplines: lower height, high current, low DC resistance, high Self-Resonant frequency etc. 2 filters are designed is for better phase responds (although phase responds is not important for the audio) and steeper filter curve too.
the board is longer by 23.4mm, the housing will be longer accordingly.