[Daroid]

Quote:

Originally Posted by Joshatdot /img/forum/go_quote.gif From the frequency response in the Datasheets, the 2132 has a bass boost, while the 2227 is pretty flat.

Which of the graphs are you looking at?

 

[Joshatdot]

Quote:

Originally Posted by Daroid /img/forum/go_quote.gif Which of the graphs are you looking at?

Top of Page 4
OPEN-LOOP GAIN/PHASE vs FREQUENCY - http://focus.ti.com/lit/ds/symlink/opa2132.pdf

Top of Page 5
OPEN-LOOP GAIN/PHASE vs FREQUENCY - http://focus.ti.com/lit/ds/symlink/opa2227.pdf

 

[Daroid]

You don't use opamps in audio circuits in open loop configurations.

Both graphs show the gain with no feedback. The feedback limits the volume level.
By using a standard non-inverting op-amp for audio with feedback as in
http://www.ozhiker.com/electronics/a...ic/noninv1.gif

The gain is Vo/Vin=(R2+R1)/R1.

The most you'd ever need in audio opamp circuits is a 12-20 dB gain, so the graphs show that the OPA2132 amplifies this up to 800 kHz and the OPA2227 to 750kHz. The resistors R2 and R1 limits the amplifications so it will be a flat +20dB amplification from DC to 750 kHz at least.

The open gain loop vs. phase plot tells you it has a phase margin of about 45 degress which makes these opamps easy to design with as they do not show instabilities, and one does not need to limit their bandwidth when they implement them in circuits.
It also tells you that neither should be used for RF amplifications.

 

[balou]

Quote:

Originally Posted by FaxData /img/forum/go_quote.gif Also, I was curious, what're your thoughts (everyone) on replacing the opa2132pa with the opa2227pa? I notice it's actually a cheaper chip on digikey, and its voltage requirements seem similar, but tangent soft rates it higher.

I dont like the opa2132pa in my cmoy very much, bass is somewhat loud but not well defined, and soundstage isnt very big.

about the rating... the opa2132 seems to be an opamp which is very forgiving to various mistreatments from newbie diyers. no oscillation problems, and I think various overvolt protections. that might explain the higher price.

you might try some chips from analog devices. the ad8066 and ad8620 seem to be highly regarded here for portable amps

 

[FaxData]

hmm, the 2107 is backordered on digikey, at least the DIP one is... I don't think I'm ready for SOIC just yet; they just look too difficult to solder. also, I've seen posts saying the ad8620 doesn't work well in a cmoy?

I'm thinking of picking up the 2227, a TLE2426CLP and a 1000uF aluminum electrolytic cap (P5544-ND) to replace parts on the circuit I've built... looking at the diagrams, I think I've figured out a good layout for the rail splitter and the bigger cap (20mm long?!). I have to pick up some audio jacks too (accidentally left them out of the first order).

anything else worthwhile, since I have to pay the $5 from digikey? I just used the standard input caps listed at tangent (3013PH). they're huge, any suggestions for something better and/or smaller? other suggestions?

 

[Daroid]

1000uF is overkill, even 220 uF is more than enough. The circuit draws next to no power anyway. This might save you some space.

If you need physically smaller input caps you can choose between a lower voltage rating (A), lower capacitance (B), different di-electric (C)

A: Is difficult, polyester already offers the lowest voltage tolerance vs. size.
B: Possible, but you'd need to rethink the value of R2. I'd say - with the risk of slightly more white noise - choose C2 to 47nF and R2=200k or slightly higher. ONLY do this if the size really is a concern.
C: Different di-electrics features different ESL and ESR. Polyester is probably still the worst, but still the most compact solution.
The only other recommendation for better capacitors which is a compromise of the good specs of polypropylene and the small size of polyester, is polycarbonate caps. But they are dicontinued.

 

[FaxData]

hmm, I guess I don't really need to change C2... I was just contemplating it.

Also, I believe I got polypropylene caps, not polyester ones. I could be wrong though. P3104 from digi-key.

So what's the deal with C1? I've heard increasing the capacitance improves bass handling, and I've see cmoys with 1000uF caps on ebay. Is it all for nothing? I think even tangent suggested trying 470uF for "significant improvemnt" on his website... I'm pretty sure I should have the room with a little work on the board, since I will only have a single C1 after putting in the rail splitter. Not worthwhile?