PPA Project Announcement
Mar 8, 2003 at 4:17 AM Post #17 of 340
This looks like a great project!

Do you intend this design to be less user-configurable than the META42? The recent distortion measurements by KurtW suggest that the extreme configurability of the META42 is perhaps a mixed blessing. I think it would be desirable for the PPA (and the META42 for that matter) to have a well-tested reference configuration that sounded good and was specified in detail.
 
Mar 8, 2003 at 4:49 AM Post #18 of 340
Yes, it is intended to be less user configurable. The testing I had hoped would happen with the META42 resistor configuration got swept under the rug and forgotten early on. This time we are not leaving such things to chance.

1k is a reasonable value for the buffer input resistors. They could be higher or lower. The choice is not critical.

Incidentally, Kurt will be joining Team PPA when he gets back from vacation in a week or so.
 
Mar 8, 2003 at 5:11 AM Post #19 of 340
Quote:

Originally posted by morsel
Incidentally, Kurt will be joining Team PPA when he gets back from vacation in a week or so.


Yay!
smily_headphones1.gif
 
Mar 8, 2003 at 7:49 AM Post #21 of 340
Issue: Ground plane over high impedance input area - definitely something I recommend you do. Production version of my PDAC does that and it made a big difference. Granted I do have sources of EMI within few cm but still, the noise at the output is unmeasurable on my oscilloscope. It might or might not be audible but it sure is measurable. 50k Alps Blue is also a good choice IMO.
 
Mar 8, 2003 at 8:46 AM Post #23 of 340
Hi Aos, as you can see from the layout we do have a mini ground plane over the high impedance area but not over the whole board. We started out with a star ground configuration but switched to the ground plane a couple of months ago.

We are seriously considering separating signal ground from opamp power ground with a 10 Ohm resistor and possibly a small cap in parallel to further reduce noise and hum. To break ground loops between the amp and other equipment we might want to put 10 Ohm resistors in series with all input connections as well.

I have tentatively split up the ground connections into two groups, let me know if you think any of them are in the wrong group:

signal ground connections:

input grounds
pot grounds
opamp input resistors
opamp feedback resistors

power ground connections:

opamp bypass caps
TLE output
ground plane
case
pot shaft and panel switches

We haven't talked much about crossfeed. There is no onboard crossfeed atm. I'm thinking it should go on a daughterboard due to space considerations. There is onboard bass boost at the moment, but not yet room for the polypropylene BC Components cap we want to use.
 
Mar 8, 2003 at 10:57 AM Post #24 of 340
hi Morsel

this project looks like it will be the ultimate opamp based amp...
eek.gif


just one question, why using bass boost ? i thought such type of implementation was not welcome in the high-end world.



m.
smily_headphones1.gif
 
Mar 8, 2003 at 6:40 PM Post #25 of 340
Quote:

just one question, why using bass boost ? i thought such type of implementation was not welcome in the high-end world.


I believe it's a ppl thing. You see the bass boost often in ppl's designs.
 
Mar 8, 2003 at 6:56 PM Post #26 of 340
Quote:

Originally posted by Possum
I believe it's a ppl thing. You see the bass boost often in ppl's designs.


PPL is a basshead!!
tongue.gif



MekanoPlastik:

You can just leave out the bass circuit, too. It's not required by law to have to intall it.
biggrin.gif
 
Mar 8, 2003 at 6:58 PM Post #27 of 340
Do you see availability of the ALPS Blue as a problem? I agree that it's an excellent choice, but it's not so easy to get. Maybe tangent would sell them on his website with the boards, as he does now with the Elantec buffers for the META.
 
Mar 8, 2003 at 7:57 PM Post #28 of 340
can't say I really call it portable, more luggable! seriously though, this thing looks awesome, and its great to see radically new topologies coming through!

nice one, dudes and dudette!

g
 
Mar 8, 2003 at 8:24 PM Post #29 of 340
Since Tangent is selling the Alps Blue there should be no problem with availability.

Time to tackle some of the issues. Comments are welcome.

You may have noticed that there are two LED modules on the board, one centered and one on the left side of the faceplate. While it is possible we might leave both of them, it would be useful to hear how people feel about positioning the power LED and the headphone jack.

My opinion is the headphone jack should be in the center of the front panel so left and right handed people who use 1 hand to pull their plug out will both be able to do so easily. Putting the LED in the center and the jack on the left means that left handed people will not be able to brace their fingers against the case to withdraw the plug, thus causing the entire unit to shift on the shelf. Some people don't care or use two hands and find the centered LED more asthetically pleasing. I think that since one of the two has to be on the side we might as well make it more ergonomic for everyone.

This leads to the issues of panel mount .vs. board mount components and custom cut panels. The pot is board mounted, that is a done deal. At the moment nothing else is board mounted and there is a 1 inch no parts zone at the front and back of the board to allow for panel mount components. This gives the most flexibility in choosing parts, but means more work during assembly. If we used board mount parts or rigidly specified panel mount components we could then offer custom cut panels.

The so called locking silver tab Neutrik jacks are board mountable but do not in fact lock. The red tab locking jacks do lock but are not board mountable. If we are going for board mountable jacks we might as well go for some of the more sleek and asthetic jacks that don't lock.

Nonlocking jacks can result in shorting the amp outputs. The amp does not have any output protection at the moment. Shorting the outputs for more than a few seconds with the volume at a high level could blow the output buffers. (Imagine yanking the cord 1/2 way out from across the room.) That leaves us with 3 options: live with the risk, use locking jacks, or provide output protection.

I think DC servo circuits like the Gilmore uses are the best thing since sliced bread, but PPL hates them. Other options include fuses, thermal resetting fuses, current regulators, or foldback current limiters, either in the power supply, on the buffer rails, or on the outputs. All of these solutions have the potential to harm the sound quality.

Of all these options, perhaps the least intrusive and complicated is to current limit the power supply and not use too much C1 rail capacitance. The capacitors would still be able to provide power for dynamic passages, but a dead short would quickly drain them and be limited by the maximum current of the supply. Going from 2 to 4 buffers per channel would help distribute load among the buffers, reducing the chance of damage while increasing the ability of the amp to handle dynamic passages, however, doubling the buffers will consume more current and make the amp less portable friendly. PPL's Not So Portable Amp? Perhaps. Maybe all that matters is that it's easy to carry around, not that it will run on rechargable batteries.

Moving to 4 buffers per channel will eat into board space, forcing a reduction in C1, which is OK as there is probably too much C1 capacity anyway.

Alternatively, we could switch to TO-220 package BUF634 buffers, which cost twice as much and don't sound as good but take less space and have built in output protection.

Isn't it interesting how all these issues are interconnected?

We are currently testing the 3 FET isolated opamp power rails configuration to determine how much if any benefit it provides over using a single FET isolated rail for all 3 opamps.
 
Mar 8, 2003 at 9:32 PM Post #30 of 340
Hi morsel,

I vote for 'PPL's Not So Portable Amp'.
biggrin.gif


I guess I'm a little disappointed that there has to be such compromises in the design. Designing for portability, limits board size, circuit implementations, etc. I would have much prefered a full blown ac PPA.

Excuse my snivelling and whinning!
tongue.gif
 

Users who are viewing this thread

Back
Top