[Cmoyamphelp]

okay got it done,put in 25V 470uf caps and .47uf caps as well and but in a opa2107ap,what next? any other way too make it sound even better?(it's already great but i feel like adding more too it,

 

[n_maher]

Honestly, I wouldn't invest a whole lot of time or money into a Cmoy. The simple fact is that is that it's little bit limited and the more you try to tweak the more you're just making it like a totally different amp. Example, add a buffer and a TLE and you're starting down the road to a A47, add output buffers and reverse voltage protection and whoops, you're basically building a MINT or a PIMETA.

So my advice would be to use the Cmoy as a stepping stone and go ahead a try a totally different amp next. The next easiest (although not cheapest) would be a PIMETA. You can build a stripped one for around $75.

My 0.02,

Nate

 

[Cmoyamphelp]

what one would give me the best sound and still fit in a mint tin??

 

[Oink1]

Quote:

Originally Posted by Cmoyamphelp what one would give me the best sound and still fit in a mint tin??

I believe the next amp that would fit in a mint tin is a mint - Pimeta's are too large

 

[Pars]

You could make that amp into an A47 with the simple addition of another socket, opamp, and (4) 47 ohm resistors (2 where your current R5 jumpers are). Changing out the two voltage divider resistors you currently have for a TI TLE2426CLP railsplitter would also be a good idea. $2 for the TLE, $0.50 for the socket, $1 for the 4 resistors and not sure what opamp you are using, but say another $4 for that... = A47. Sounds alot better than a CMoy, rivals a MINT in a lot of ways, and a lot cheaper (the MINT probably is a better amp, tho). Basically the second opamp is configured as a unity gain buffer, feeding into your currently existing feedback loop thru a 47 ohm resistor.

Those who think that an A47 requires the 5-pin BUF634 for the PSU had better go back and read thru Apheared's stuff (inventor of the A47). He never used one (AFAIK); for the original A47 Apheared actually used two battery packs with the virtual ground coming from between them (ala Grado RA-1 or whatever). The A47 architecture that you see posted on the net is not from Apheared, but was a tweaking of the Lavollis architecture (Ben someone, name escapes me right now). Furthermore, at the time that that was done, it seemed to be common logic that a TLE could not source enough current on the ground to be effective past a single opamp design. Tangent himself has since stated that that is not true (check Tangent's ground article that he referenced in the original thread).

At any rate, search for A47 on headwize. Realize that Apheared used two dual opamps, using one for each channel. You can achieve the same thing by leaving your existing opamp handling both channels, and using the second one as the buffers, one per channel. Just change the pin numbers accordingly. I've built 10 or so that way.

 

[Earwax]

Have you looked at Tangents tweaks page? Using the TLE2426 for virtual ground helps and it's an easy change.

I've been tinkering with cmoys for 6 months and I don't feel like I've gotten to the end of the road on what can be acomplished with the simple one opamp design yet. For instance, Mint/pimeta tweaks like class-A biasing can be added to a cmoy.

Even though I have a Mint and Pimeta, my tweaked-out cmoy is usually what I carry around because of it's small size and insanely good battery life. Goes well with my minidisc player.

[P.S. For headphones of 32 Ohm and less, do go with a buffered design like Mint or A47, high current for very low Ohm phones is a weak spot of the single opamp design]

P.P.S. the latest version -

 

[tangent]

Quote:

class-A biasing can be added to a cmoy

Not practically. Class A biasing works in a buffered headphone amp because the op-amp's "load" is the buffer input, in the multi-megohms range. Since Class A is defined as having greater standing current than the load requires, it takes a miniscule amount of current to keep the op-amp in class A in that configuration.

But when the load is headphones -- 5 orders of magnitude heavier load! -- the required current is up in the tens of milliamps. At the very least, this is going to add simple load-based distortion. It's also likely that you will start to turn on the output current limiter circuitry in the chip, which also makes it sound worse.

 

[Earwax]

I goofed by talking about class-A, what I'm doing is still class AB, but adding a couple ma of current, biasing the opamp further towards A, but it's nowhere near full class A operation. It's enough to help some opamps, especially some BB, the AD opamps I tried didn't benefit.

 

[Pars]

Quote:

Originally Posted by Earwax Have you looked at Tangents tweaks page? P.P.S. the latest version -

Nice job on the board, Earwax! I see shades of a certain clone in there as well (Solens). What pot is that btw, and where from?

I've been tweaking Guzzler's design into an A47, and have etched one board (turned out fine, tho I didn't have a TO92 cased TLE around to finish it). Next one will be a double layer board with ground plane, setup for the RK09 Alps pot.

Thanks,

Chris

 

[Earwax]

The pot is G13285 from Electronic Goldmine , a 10k alps is what they were shipping even though it doesn't exactly match the photo. A bit large for a mint tin, and not great tracking, but seems entirely neutral in sound.