[doobooloo]

Hello,

I was wondering if it would be valid to start a thread where we can start discussing some minor/major changes to the original PPA board as new boards are apparently being ordered. (Just two weeks ago the boards I got were 1.0A, today I received 1.0B boards...)

My two requests would be:

1. Larger, rectangular C7 space to accomodate for >0.12uF Polypropylene box caps (6.0x17.5 mm to accomodate up to 0.39uF BC MKP416 series caps would be great)... this would probably require some creativity as I don't see much free real estate available to accomodate such a layout in the current setup unless there is some considerable shifting of components to accomodate for this. But it would be REALLY nice (for freaks like me) to not have to wire up giant caps somewhere else on the board just to get the kind of bass boost I really want.

2. A cermet trimmer to replace R8 for easily adjustable class-A biasing current (and also to avoid the hassle of getting two gazillion resistor values just to match the draw currents)... The square cermet trimmer footprints aren't that much larger than a resistor - I would assume that it wouldn't be that difficult to add this feature in the next revision perhaps?

These two issues are ones I'd love to see addressed in the new revision!

Any ideas or further suggestions?

 

[Voodoochile]

Ditto Doobooloo's list

another (9th) 12.5mm electrolytic pad near front (it does not have to be populated of course, but will be a nice option)

minor jiggle to the smaller electrolytic landing pad to receive 10mm cans

 

[doobooloo]

Actually, a 9th C1 cap would interfere with a headphone out jack - the Neutrik barely fits, other options won't even fit with one or two C1s from the front unoccupied...

Well I guess it doesn't hurt to have extra space.

 

[morsel]

PPA suggestions are always welcome. A few questions:

Why would you want a 9th C1? I think 8 is already overkill. The only reason there are 8 is for Elna Cerafines or Silmics, Even with the Elnas it is a lot of capacitance considering the high PSRR. Have you tried 4 or even 2 .vs. 8 to see if you can hear or measure a difference? Kurt and I could neither hear nor measure a difference between 2 Panasonic FC .vs. 8 Elna Cerafine C1 caps.

The cermet trimmer idea is appealing at first glance, but it would require major rearrangement of the board, as there is not enough room for them atm. It is not necessary for the bias currents to be identical, but if you want them close, choose a 2N5484 and 2N5486 combination along with a fixed value of R8, and you should be in the right ballpark every time.

I don't think there is room for 300x700mil C7 caps.

Don't take these replies as a final "no", I'm responding genuinely with willingness for further dialog on these points.

 

[Jupiter]

morsel,

what about Voodoochile's suggestion for 10mm C4s?

 

[morsel]

Ah, missed that one. C4 is designed for 300mil 100uF FC grade electrolytics. What part do you want to fit there instead? 150uF will fit, but less than 100uF will be fine too. We already stretched that spot a while back. I think the current caps are pretty big as is.

 

[morsel]

I think a reasonable litmus test might be this: If a major disruption, rearrangement, or compromise of the layout is required to implement a change, is there measurable or audible benefit to the change, or does it provide some real world tangible added value or feature, that outweigh the negative aspects of the change?

 

[Voodoochile]

It would be nice to be able to fit a Cerafine or Muse KZ or Blackgate FK there. I can fit a Muse Fine Gold 100uF cap there, but that is the only one I can find that is not an Panasonic FC.

The Panasonic FC is an excellent cap, but if you are using something nicer still, it would be nice to be able to fit a cap from the same line. That's all.
Hardly a requirement, I realize, it's just that you could really fit them there now if you tweak the leads a bit, even without doing the doobooloo hi-low mounting. Seems that it is very roomy for 8mm, and just barely snug for 10mm... so I was thinking you could fiddle the pads a bit to sneak the 10mm in there without lead forming. But I was not there during the endless board tweaking.

As for the ninth cap spot, it was as much because there is room for it as soon as you depart from a eurocard case, and you could of course omit it if you were using the eurocard case. If you use the Blackgate FK caps, you can only fit 8x220uf... so a ninth cap would be nice, and surely there is room. In fact, I have used the alternate power con pads to accept a ninth cap, applying power from the underside.

So don't get me wrong- I'm not saying you need it, or even should have it, just that if the room existrs, why not offer it? I have already seen plenty of PPAs build with only three or four large-capacity FC series caps, so no one is afraid to leave some locations empty. Below is a cap shot showing the Nich FG 8mm:

 

[aos]

As I said before you don't really need to buy a large assortment of resistors beforehand. Just get a pot in that general range, trimmer or full, and as long as you have a breadboard and some power source and an ohmmeter you can determine the resistors for each JFET set easily, then buy only resistors you need. Only if you're a big-time DIYer you may want to have a large selection of resistors as leftovers. Changing resistors is even a bigger hassle than trying to find matching JFETs in a bag. Your amp is not going to burst out in flames just because you didn't make it exactly as tangent suggested - there are often several ways to measure the same thing, so choose what's more convenient, cheaper, or easier for you.

Having trimpots in the amp itself would be a right thing to do if cost and size don't matter. It would also allow you to change bias without (re)soldering anything.

 

[morsel]

Regarding C4, I just checked the Cerafine chart and found that 300mil (8mm) caps have a 150mil lead spacing, but 400mil (10mm) caps have a 200mil lead spacing, so they are not compatible. We would have to add extra pads for the alternate lead spacing as well as reposition the caps without precluding the use of a DIP socket for the neighboring opamp resistors.
Quote:

If you use the Blackgate FK caps, you can only fit 8x220uf... so a ninth cap would be nice, and surely there is room.

Regarding the 9th C1, I'll ask again, why do you want or need more than 8? Even with 220uF caps, 4 or 5 should be plenty, let alone 8. Have you actually listened to and/or measured the difference between 4 and 8 as I suggested? A 9th would be in the way of the headphone jack, and more importantly, it will encourage people to believe they actually need to populate it, and that amps without that many caps are somehow inferior. Convince me it is not a waste of good capacitors to have that many on board. More is not always better.

 

[ppl]

I actualy Like The Adj Bias Trimpot and 10 mm op amp electrolytics these would more easaly acomidate those of us that wish to use audiophile caps and like to have the Bias current on each channel to within 1% just to sleep better if for any other reason.

 

[morsel]

I pushed parts around a bit in ExpressPCB. At cursory glance, the TLE would have to move farther away from the ground plane to make room for larger caps, which would in turn block use of a DIP socket for the opamp resistors. I'm not sure if there are any tiny resistor sized trimpots out there, but I am open to suggestions. The stock Bourns part from Digikey is too big to fit in the R8 location.

 

[ppl]

What about the BC components on pg 840 top only CTw94 series only 2.5mm thick with inline dip pins. Morsel i sent you a abstrack of the PCB file with this pot installed. Note that the Digi key pot recomended above has inline pins and the one on the board i sent ya has triangel pinout but you should get the point.

 

[doobooloo]

I've actually tried both the CTw94 and Bourns parts as R8 substitutes, and yes the CTw94 is definately much thinner and I don't see any advantages of the Bourns part over the CTw94.

I sort of like the Bourns part for now, though, because it has just enough standoff from the PCB so that I can bend the legs so that they'll fit in the R8 holes and still be flush agains the board.

Almost looks like it was meant to be there.

Regarding the C7 - anyone know of any place other than epassives where I can get the 0.33uF MKP2 PCM-5 polypropylene caps in relatively small quantities? This part would barely fit (8.5mm x 7.2mm) in the C7 layout without being obtrusive.

Apparently at epassives they only have upto the 0.22uF values, which I am not very interested in.

I'm going on a retreat today, coming back Sunday afternoon - when I get back I'll post some pics of how the trimpot thing worked out for me. I think even with the current layout, it can be done to look very nice and professional, but of course, with an actual layout for the trimpots it could be even better (and save the anal builders like me some trouble haha)...

It just feels so good to have the bias currents nearly perfectly matched...

i'm such an anal anus.

Also, is it possible to make the ExpressPCB files for the PPA available to public so that we can also play around with it and contribute directly to the board design...? I don't know if this issue was brought up before but seeing that the battery board's ExpressPCB files are up, I'm itching to be able to play around with the files...

 

[guzzler]

Quote:

Originally posted by morsel I pushed parts around a bit in ExpressPCB. At cursory glance, the TLE would have to move farther away from the ground plane to make room for larger caps, which would in turn block use of a DIP socket for the opamp resistors. I'm not sure if there are any tiny resistor sized trimpots out there, but I am open to suggestions. The stock Bourns part from Digikey is too big to fit in the R8 location.

is the loss of the socket for the resistors particularly major? I can see the downside of moving the TLE, but surely in a high-end amp like the PPA, socketing the resistors is a less than ideal connection method and so shouldn't really be an issue here??

g