For me it is more than enough but some seem to need 200W or more delivering amplifiers. Small amplifiers are more elegant and designing for real world power levels makes things possible. Better design for quality Watts one really uses IMO.
One can try out lower supply voltages than 26V. Makes linear power supply design coming in acceptable area. Mine is 20V with LT4320 rectifier and 17V 5.6A transformer. Too bad the transformer is whining so the SMPS is back.
Never bothered with manipulating PMP modes as the chip changes modes automatically itself. I hope the Aiyima A8 has the optimal modes for sound quality instead of efficiency covered in firmware.
One can try out lower supply voltages than 26V. Makes linear power supply design coming in acceptable area. Mine is 20V with LT4320 rectifier and 17V 5.6A transformer. Too bad the transformer is whining so the SMPS is back.
Never bothered with manipulating PMP modes as the chip changes modes automatically itself. I hope the Aiyima A8 has the optimal modes for sound quality instead of efficiency covered in firmware.
Last edited:
The double mono MA12070 PBTL Can drive speakers up to 2 ohms, moreover, it's the only way to achieve the advertised power... At 8 ohms, it will be on par with a simple Ma 12070.
The customers that have loudspeakers reaching 2 Ohm are likely not MA12070 amplifier customers 😉
The dual mono MA12070 PBTL amplifiers maybe are gadgets with a feature to boost sales. Some day one will be sold with a tube.
The dual mono MA12070 PBTL amplifiers maybe are gadgets with a feature to boost sales. Some day one will be sold with a tube.
Last edited:
To be fair though, it's hard to believe a tube pre-amp using underpowered leftover Chinese tubes and a 12V wall wart can sound any good. Yet somehow it did the trick for a lot of folks.
Good tube products exist but they ain't cheap (be it Chinese or Western origin, surprisingly).
Tpa3255 can be used from split +- supplies. The advantage with it is that power supply pumping is eliminated due to its bridged configuration.
Find it somewhat peculiar that the PMP0 is slightly lower power at 1% than the other channels. Nothing major, but it seems to be something like 35W vs 38W for the others.
I agree that the average customer is not electrostatic speaker owners🙂
But still many customers connect small class D amps to old speakers with impedance that can very well variate between 3 ohm and 8 ohm with big phase angles. Think the MA series amps tacle this well compared too other cheap solutions because of their extremly high efficiency and low RDs ON. (eg. good temperature regulation in the chip during use)
My theory is that this is the reason people rate them well even with their medium (good enough) measurement values.
Looked through various datasheets of the IR4xxxm IR2xxx, and the according evaluation boards.
Basically it is copy paste all the time. The same circuit drawn in the datasheets. The same (deviating) implementation on the evaluation boards schematics.
Still thinking the feedback resistor should actually be connected to IN- and not the front side of the Rin. As drawn every time in the datasheets.
Also when in more detail the PWM mode is discussed, it is drawn like this.
Is this indeed an implementation error, that is repeated over and over? Because they copy every time the previous design?
Any opinions?
Basically it is copy paste all the time. The same circuit drawn in the datasheets. The same (deviating) implementation on the evaluation boards schematics.
Still thinking the feedback resistor should actually be connected to IN- and not the front side of the Rin. As drawn every time in the datasheets.
Also when in more detail the PWM mode is discussed, it is drawn like this.
Is this indeed an implementation error, that is repeated over and over? Because they copy every time the previous design?
Any opinions?
Proof of the pudding is.... eating it.
I removed the 47k's feedback resistors and replaced it by Dale RN60 resistors and connected it to the input of the MA5332. So according to the datasheet.
Didn't get worse, I would even say it sounds better. Cleaner. Of course also a different resistor type use, which will change sound signature also.
I can hear the details in Talk Talk's - Spirit of Eden. That's a good start.
Before I had an impression of some "noise around the edges". Don't know how to express it. This impression is gone. Thought still not that clean as the Sabaj, yet. But comes closer now.
I already had in mind removing all the resistors in the audio path. Don't know what is used, But probably just standard.
I have Susumu RR's lying around. RG's don't seem to be available at the moment. So I'll have to use the RR's.
I removed the 47k's feedback resistors and replaced it by Dale RN60 resistors and connected it to the input of the MA5332. So according to the datasheet.
Didn't get worse, I would even say it sounds better. Cleaner. Of course also a different resistor type use, which will change sound signature also.
I can hear the details in Talk Talk's - Spirit of Eden. That's a good start.
Before I had an impression of some "noise around the edges". Don't know how to express it. This impression is gone. Thought still not that clean as the Sabaj, yet. But comes closer now.
I already had in mind removing all the resistors in the audio path. Don't know what is used, But probably just standard.
I have Susumu RR's lying around. RG's don't seem to be available at the moment. So I'll have to use the RR's.
The 2k7 resistor likely was an input filter (combined with the input capacitance) or stopper resistor. Datasheet stuff generally is the minimum required for correct functioning and not the maximum.
2 changes at the same time give noise around the edges, that is correct 🙂 My guess is that it will never be as the Sabaj simply because it is not multilevel class D.
2 changes at the same time give noise around the edges, that is correct 🙂 My guess is that it will never be as the Sabaj simply because it is not multilevel class D.
Last edited:
For who is interested.
I like this amp more and more.
I really listen to the music, forgetting the amp. Getting drawn in. With the Sabaj I noticed only listing to the amp. That says a lot.
No it is not perfect, yet. But the presented picture is really really rock solid and open. It outperforms the Sabaj here by far.
This sounds like a real big amplifier. And with much more ease. Heavy continuing bass, like some Dead Can Dance tracks, are played without a blink of an eye.
Swapped back to the untouched module (I have 2). Listening to Barbara Streisand. First impression is, it is also good sounding, perhaps little less dynamically. But back this feeling, something is not right.
Then playing the same track with the modded version, it became clear that with the untouched version Barbara sounds like she had been drinking and smoking too much the night before. As with the modded version she sounds as clear as she should be.
This has potential.
Only mods so far, is changing the feedback resistors, with the 47k directly at the -IN. The 1k is connected from pin 24 to the 47k resistor. This shortens the (too big) feedback loop. So, still listening to the ceramic coupling caps and the 5532.
I like this amp more and more.
I really listen to the music, forgetting the amp. Getting drawn in. With the Sabaj I noticed only listing to the amp. That says a lot.
No it is not perfect, yet. But the presented picture is really really rock solid and open. It outperforms the Sabaj here by far.
This sounds like a real big amplifier. And with much more ease. Heavy continuing bass, like some Dead Can Dance tracks, are played without a blink of an eye.
Swapped back to the untouched module (I have 2). Listening to Barbara Streisand. First impression is, it is also good sounding, perhaps little less dynamically. But back this feeling, something is not right.
Then playing the same track with the modded version, it became clear that with the untouched version Barbara sounds like she had been drinking and smoking too much the night before. As with the modded version she sounds as clear as she should be.
This has potential.
Only mods so far, is changing the feedback resistors, with the 47k directly at the -IN. The 1k is connected from pin 24 to the 47k resistor. This shortens the (too big) feedback loop. So, still listening to the ceramic coupling caps and the 5532.