If you're interested in more details concerning classD PSRR measurements then this paper is an excellent read - https://e2e.ti.com/cfs-file/__key/c...iles/6/PSRR_2D00_Open-Loop-Vs-Closed-Loop.pdf
What you say is that an ideal class A poses a constant load to the PSU. So, it should need just a small amount of filter capacitance only. However, I have seen class A amps with humungous amount of PSU caps.
Please see
[url]http://www.odysseyaudio.com/products-khartago-stereo.html
This product has 60000 uf of PSU caps. There is an upgrade option too!!
Secondly, by your own statement, in a class D, output currents follow output power. So isn't a cap bank needed to provide short term input current?
Are you implying that a humongous bank of caps is a "must" for a class A amps? From a commercial standpoint, a humongous cap is is strong selling feature, whether it is needed or not. Don't forget one factor that affects the bass response is the damping factor of your amp.
I think a better point to start is to look at your load requirement (output at speaker). Some speakers require more, some less. Estimate what will be the maximum demand, and then work backwards to see how much supply is needed.
I believe the same topic popped up recently in one of the D-amp thread and Doctormod has shown (with assumption, calculation, and numbers) that one does,'t need huge capacitance to get "good" bass response.
Regards,
There is a graph on the first page of the tpa3255 datasheet that answers this. Right at 100 watt output power is when the THD+N quickly ramps up. But it's about 0.003% at 100 watts @ 8ohms. Around 0.0025% at 50 watts. Tons of similar graphs in the datasheet.
Depends on your definition of really low.
For comparison, TI's previous-generation class D chip, the tpa3116, has 0.04 to to 0.5% THD+N at 30watts with 8-ohm load (depending on frequency). See page 9 of the tpa3116d2 datasheet. That's a lower-power chip, so above 30 watts, THD gets bad quickly.I cannot emphasize enough the importance of playing around with psudesigner rather than asking questions. It's a huge waste of everyone's time until you get to a point where you understand these basics.
A class-a amplifier draws a huge demand on the PSU with it's very large quiescent current through the output transistors and is only being refreshed at 100/120 hz depending on your location. Class-d amplifiers are much more efficient (up to 95% near max power) so are only demanding stored energy as the music demands it, which, by and large, is relatively infrequent. The 3255, in comparison to it's max output, has essentially zero quiescent current demand.
A simple understanding of power supplies and (even mediocre psrr, as is the case in pretty much every class D I've looked at) will show that "bass is better with more capacitance" is only really true in a case where your power supply is grossly undersized, which yours is not. And to generalize, if you're modulating the output materially due to sagging rails, then you've got bigger problems than bass.
Sorry for the ambiguity. The TPA3251 is obsolete as per the TI product web page.
Oops. Had a brain fart. Apparently it's just the TPA3251EVM that is obsolete, not the chip itself.
Okay, we're getting somewhere!
First off you want to find the ESR of the nichicon cap, but the datasheet here shows it in terms of loss tangent:
http://www.mouser.com/ds/2/293/e-lkg-875808.pdf
We can convert (assuming an ideal capacitor, good enough for this work) by using ESR = wC*tan(theta); ESR = 50*2*pi*10000*1e-6*0.3 = 0.94 ohms
Dielectric loss - Wikipedia
The load of the amplifier is essentially the load of your speaker (assuming 100% efficiency, which is a very simplified assumption but will give you pretty good info). Let's try 4 ohms for now.