TPA3251d2

[doctormord]

Btw. Cstart is not an input. It defines the startup ramp time via a capacitor and voltage comparator.

 

[gmarsh]

Yeah, the thought there is "I wonder if I can prolong the startup time, to avoid any turn-on pops, by holding that cap at 0V..."

They don't explicitly recommend a value for the cap, but the single ended example with series caps on the output speakers has a 470nf cap, and the BTL example has a 10nf cap in the same spot.

It's in there for experimentation purposes, it'll probably be DNP or removed entirely from the final board.

 

[doctormord]

Well, do they they charge the bias-caps when holding this pin low?

I use the reset pin for that.

 

[gmarsh]

Current schematic (component values still aren't set yet)

And current state of the layout. Got the size down to 120x80mm, I think I'll keep it there.

 

[a3cervo]

Subscribed to this thread. thanks for keeping us posted on your projects gmarsh & doctormord.

 

[mravlca]

@gmarsh and @doctormod: Thank you guys for explanation on implemented input circutry.

 

[alexanderson]

thanx to all for this interesting thread.

 

[Nisbeth]

Nice work Gentlemen!

/U.

 

[doctormord]

First fit.

 

[camelator]

Whaou!
Amazing job

 

[camelator]

on the graphics for THD page 1, there is a kind of rise at 10Khz, do you have comments on it? Where it comes from? possible impacts on sound quality?

 

[doctormord]

Yeah, it will sound horrible!

(You'll see this rise in every datasheet from almost every vendor. It is due to the modulation scheme. Have a look into the datasheets of TPA3116/3118/3132D2, TAS-series, MAX9744/04/09 etc. )

 

[camelator]

Got it...
Well you know doctor mord, I am not as good as you,....
Your job is really really amazing

 

[doctormord]

Don't worry, this was meant as a joke.

I'm afraid, we hadn't this measured at the MAX9709 board:
MAX9708/MAX9709 25W/50W class D - measurements | #360customs

 

[deandob]

Any (subjective) comments about what this new chip sounds like, especially in comparison to the other popular class D chips (eg. Tripath)?

I'm looking at this chip for a 2.1 amp to drive a FAST setup (desktop speakers using Jordan full range drivers and a subwoofer). This chip seems ideal for this application (lower power in SE mode, decent power to drive a subwoofer in BTL mode).

Will likely knock up a PCB for it over the next few weeks.

 

[5th element]

http://www.icecomponents.com/c-112-class-d-inductors.aspx

 

[jean-paul]

Don't worry, this was meant as a joke.

I'm afraid, we hadn't this measured at the MAX9709 board:
MAX9708/MAX9709 25W/50W class D - measurements | #360customs

So MAX9709 is a very good chip. A state of the art PCB for that one would fit the bill then ?

 

[5th element]

So MAX9709 is a very good chip. A state of the art PCB for that one would fit the bill then ?

I certainly dont see anything special about it.

 

[doctormord]

Regarding what specs/needs?

 

[5th element]

It just looks like any run of the mill single chip class D amp. Yeah it's got spread spectrum switching but in any hifi application you'll be using an output filter anyway.