[pinkfloyd4ever]

Ok, for those of you who don't know, the PPAS is a surface mount (and very portable-the original PCB is about the size of a stick of gum!) version of the PPA. The original group buy for the PCBs was back in '06, and they're kind of hard to find nowadays. So just tonight FallenAngel and error401 and I have got all excited like JOJO THE IDIOT CIRCUS BOY WITH HIS PRETTY NEW PET!! Now, I personally don't have the knowledge to do PCB (re)design, but if there's enough interest in having the layout redone to turn the volume pot (alps RK097) 90 degrees and possibly change buffers from ICs to discrete buffers...I'd definitely be willing to help with logistics of ordering & distributing boards, prototyping, etc. I'm wondering though, would the performance increase of discrete buffers be worth the increase in size and presumably power consumption? Just sharing my thoughts, I honestly don't know enough about it to make any real comments on that, so how about all of you tangents and ambs who know what you're talking about? What are your opinions? I know that the PPAv2 uses ppl discrete buffers and that the sijosae buffer is smaller, but error401 or anyone else, could you elaborate on the differences between the two?

edit: I forgot that there had been a discrete Sijosae buffer add-on board layout designed, but I'm not sure if it was ever produced.

Finally, and perhaps most importantly, who would be interested in buying a production board? If there's not enough interest among head-fi diyers, it may not be worth doing. So what are everyones thoughts?

 

[holland]

I might be, other than that, what are your design goals? What transistors are you planning to use for the discrete, and what type of buffer configuration?

Since the primary target would be portable, what type of battery life is your goal (design goals)? What voltage targets, what battery configuration for the case of choice (li-po, nimh, size and capacity, etc.)?

 

[error401]

[size=small]Late edit: This thread ended up diverging pretty far from the original PPAS. I will eventually go ahead with the 'super PPAS' I proposed in the earlier part of the thread, and will dub it the PIP. It'll get a new thread when I get the ball rolling on it again. In the meantime, I've designed a new board for the original PPAS circuit; discussion of that starts on page 10 of this thread.[/size]

Mainly we were hoping to replicate the original PPAS, at least in function. There will be no significant changes to the PPAS circuit, except possibly adding a discrete buffer. I don't have a PPAS so I don't know what it's battery life was like, but I assume this design would be similar.

A discrete buffer would probably have Iq around 8mA/channel (from a hasty sim), a little less than BUF634 in high bandwidth mode. This thing is not going to win any battery life awards, but it should be small enough to fit some decent batteries in. I will design a 50mm wide board as small as I can to fit in the Hammond 1455C80x case. The pot will be onboard, all other components air-wired. Batteries are builder's discretion...I will probably be powering mine from a wall-wart.

If a discrete buffer is implemented, I will likely take Sijosae's design verbatim. Substitute BC807/BC817 (SMD version of BC327/BC337).

Edit: For anyone wondering, there are some significant differences between the reference PPA and the PPAS. They're mostly in the power supply, I'll list what I notice off hand:

- No high-current supply or individual JFET-isolated opamp rails
- Single JFET current source to bias the opamps instead of a cascode
- Single output buffer
- (Much) less power supply capacitance
- No bass boost.

 

[srserl]

I will buy a board if for no other reason than I enjoy surface mount projects, and there aren't very many of them to build. I have heard a PPAS only a couple of times, but I remember it sounding pretty good.

Scott

 

[holland]

I see. Might be hard to airwire the jacks with the board rotated due to component and board height on a PCB in a slot. The width of jacks with the pot on the same edge can come close to covering the width of the hammond.

I took a quick look at the prior layout. Add some stabilization, like rail bypass. I recall hearing of some stability issues with some configs of the PPAS, might be worth looking into.

 

[rds]

I'm in for a couple boards

 

[error401]

Quote:

Originally Posted by holland /img/forum/go_quote.gif I see. Might be hard to airwire the jacks with the board rotated due to component and board height on a PCB in a slot. The width of jacks with the pot on the same edge can come close to covering the width of the hammond. I took a quick look at the prior layout. Add some stabilization, like rail bypass. I recall hearing of some stability issues with some configs of the PPAS, might be worth looking into.

Putting the jacks on the PCB only makes this worse

. The caps will be grouped on one side opposite the pot as in the current layout. This will hopefully leave room for an output jack above the SMD components toward the centre. Inputs will have to be on the opposite panel, and you're right, it will eat up some internal space. I'm working on a PCB now, and we'll see what sort of space is going to be left.

Another issue with the current layout is that it doesn't account for the Hammond extrusion. I will need to leave an extra 4mm for tall parts on each side to accomodate the screw hole, which really hurts the layout.

Edit: I corrected my sim and arrived at 3.6mA/channel. Respectably small, shouldn't do too much damage to the battery life. 7.2mA for those, 3.7mA x 3 for OPA227s and around 1mA for an LED and we're just under 20mA. Should give around 10 hours from a NiMH 9v, just like the PPAS site suggests.

 

[MASantos]

I think that if you're going to redesign the board you could put the jacks and battery on the board just like the mini3, and also make it the size of the hammond 1455 801 case. This will make it much easier to build since there will be no wiring inside. I'm sure you be able to do it!

That would be a nice alternative for the mini3 as quality portable diy amp.

 

[error401]

Interesting idea, but not quite the scope of what I originally had in mind. That said, I'd almost prefer to do it this way. Fit a 2xAA holder and the jacks on one side, and a charge controller, boost converter and the amp on the other. More complex, more fun, and more integrated. Win! Should be able to get 50 hours or more of battery life to a charge, and charges under 4 hours. Not many portables can say that!

Anyone else want this project to go that way?

 

[FallenAngel]

I have no idea how you can make it run off 2x AA, but I have my PPAS running off 8x AAA in a cut-down 1455C1202. With the layout in my head, you should be able to fit the board and 8x AAA in C802 case

 

[error401]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif I have no idea how you can make it run off 2x AA, but I have my PPAS running off 8x AAA in a cut-down 1455C1202. With the layout in my head, you should be able to fit the board and 8x AAA in C802 case

DC-DC boost converter. I know they've got a bad reputation around here, but after using the HPDAC and HPDAC2 by dsavitsk which use this topology I'm convinced that with proper filtering and post regulation and the good PSRR of opamps and buffers we can deal with the noise they produce.

Say boost from 2.4V to 10V, and then use an LDO reg like an LP2980 or something to get down to 9V.

 

[FallenAngel]

I still have an HPDac on my desk, my girlfriend uses it with her Macbook all the time. Very cool little device. I don't know about 3V though - it'll give maybe 5V at the end?

 

[error401]

Quote:

Originally Posted by FallenAngel /img/forum/go_quote.gif I still have an HPDac on my desk, my girlfriend uses it with her Macbook all the time. Very cool little device. I don't know about 3V though - it'll give maybe 5V at the end?

Nah, it can give any voltage, though we'll be around the limits at these voltages. Maybe will have to go to 3xAAA which will hurt battery life and take more space - or lower the opamp rails. Boosting the voltage isn't a big problem. I screwed up the battery life calculation though. If we boost to 10V, that's x4 though, and to build in inefficiency, say 100mA from the batteries. Still, with 2700mAH AA's readily available, that's 27 hours from a charge.

 

[rds]

I think a higher voltage NiMH or Lithium battery is a much better idea.

 

[jbloudg20]

Quote:

Originally Posted by rds /img/forum/go_quote.gif I think a higher voltage NiMH or Lithium battery is a much better idea.

x2 its just a better idea to use a 9V and implement a charger.