Let me first just say this is incredible. I never would of thought it would be possible to cram a desktop amp (especially one like a PPAv2) in to a portable enclosure.
As for your power caps, Panasonic FM's (improved version of the FC) are often used as power caps, as they perform well and have very low ESR. Also, as far as capacitors go, power caps makes the least difference in sq, so there may not be any discern-able difference between ELNA's the the FM's. Assuming that you are using 16v caps, a 470uf Panasonic FM is the same size as the 100uF ELNA's, and so you could get away with only 4 caps, leaving LOTS of space.
Obviously a larger battery is priority, but that kind of space leaves you a lot of freedom, like a gain switch, or if you're feeling really ambitious, you could try to cram in a linkwitz cross feed circuit
But really, this is already outstanding. Props to you!
Thanks for the kind words and advice. I tend to agree about the caps.
I have been fighting the flu for the last few days, so i haven't been able to make much progress on this lately.
We also started a new project at work, so that has cut into my "real" work a bit....
However, I did get my shipment from mouser, so hopefully I will have another amp built this week. As I stated before I want to try the 827 op amps in this one.
I also received some more cells to play with from all-battery, good news and bad news there. I found some sanyo prismatic cells, 1150ma that would be a much better fit with the current board design. They are the same width, 5mm shorter in length and slightly thicker. according to the spec sheet, 3 of them stacked would just fit in the case, and the extra 5mm on the end would leave plenty of room for the battery PCB. As a bonus they are also much cheaper than any of the other cells I am trying. Unfortunately they don't fit, just slightly too tall . It may be possible to machine the enclosure to allow them to fit, I will explore that as time permits.
The good news is that the larger 1250ma cell will fit with the new board layout, and if I use the FM caps there should be plenty of space to work with. Board layout with this design is not quite as sexy, as the main power caps are moved away from the output transistors a bit, but i highly doubt that it will adversely effect the amp as there are very generous internal power plains.
The Harwin pins and sockets I got from mouser look like they may do the trick for allowing the op amp board(s) to be removable. The sockets are gold plated and very low profile and fit in a 1mm hole , the pins are .5mm, tin and will need to be bent 90 degrees and trimmed slightly, but should work. They also have pins that are gold plated, but mouser doesn't stock them. The drawbacks to doing this are the cost of the parts, the pins and sockets are not terribly expensive, but there are a lot of them so it adds ~$8 to the BOM. It also requires two different sized holes for the op amp mounting, this would only be an issue if one elected not to use the hardware. Instead of using a standard .025" right angle break away header (as i am using right now), one would probably need to use individual component leads that were snipped off the caps, and solder them in 1 at a time. This doesn't sound like a big deal but I have done it, it can be a pain.
As for adding extra circuitry (variable gain and cross feed etc), a Cross feed circuit could be possible, i will explore that. A gain switch would be very difficult given the op amp boards layout. It might be possible to add two very small switches to it, but that would really be pushing things. The front panel is very tight right now, that is also a problem.
I don't want to get any ones hopes up, but i am also in the very early stages of designing a digital POT system that would allow control of not only the volume, but the gain and BB from a single encoder. It would use an AD5206 digital POT (or similar), this is a single chip with 6 addressable 256 position pots. I have a chip to play with and am working on the code to control it, I have also started on a circuit board that should still fit with the current board design scheme and case. But this is still very early. I am excited about it, but there is a long way to go. I may also try to get rid of the other 6 manual pots used to set the op amp and buffer bias as well and use digital POTs there as well, that way bias settings could be set, and measured, without having to open the case, or even require a volt meter.