[quicksilver96]

Saw there was a group buy a while back for a diamond buffer board, are there currently any boards of any buffer design available for purchase? I hate protoboard, but I guess I could use this as an excuse to get into board etching...

If I end up building a diamond buffer or Sijosae's Monofied Discrete BUF634, do I need to match anything with the BJTs or FETs? A little documentation on the design would be cool.

 

[AtomBoy]

Someone might sell you a board, but I think the group buy is over.

Steinchen's page on the diamond buffers for the Millett Hybrid is here:

http://www.diamondstar.de/dDB_overview.html

- tons o' links and design info.

There's also discussion here:

http://www.diyforums.org/phpBB2/view...277be382b1e34a

among other places.

 

[quicksilver96]

Hell yes, that first page is exactly what I was looking for. TY

 

[mono]

Quote:

Originally Posted by quicksilver96 Saw there was a group buy a while back for a diamond buffer board, are there currently any boards of any buffer design available for purchase? I hate protoboard, but I guess I could use this as an excuse to get into board etching... If I end up building a diamond buffer or Sijosae's Monofied Discrete BUF634, do I need to match anything with the BJTs or FETs? A little documentation on the design would be cool.

I took a handfull of JFETs and picked out two that gave closest to my target CCS with a common value resistor (I had a lot of 1K at the time). The Sijosae Monofied could use 2N5485 or 2N5484 or another JFET instead, and you'd just use the resistor you wanted to achieve the target CCS rate. Easiest would probably be to put the resistor in the test jig while you measure the IDSS.

As for protoboard, the Sijosae Monofied layout can be done on 0.1" hole spaced protoboard, BUT with the etched board I could at least solder the pins on the bottom to a little bit of copper on the etched board, you'd have to secure your pins another way. Also I'd intended to use them in a pimeta where the available space was limiting so for that application I reducing the spacing of some holes to less than 0.1". The actual etching pattern is on the picture of it, just scale the DPI so that most of the holes are 0.1" spaced, particularly those that plug into the DIP socket (but again, some will be less). Also worth noting is that on the Pimeta, that buffer layout will only fit in the ground channel with 10mm diameter electrolytic caps beside it, not 12.5mm.

I matched gain (hFE) on the transistors with a cheap little DMM.

 

[vixr]

YGEM

 

[quicksilver96]

After reading everything on the DB page, it seems a little more than I want to undertake for just the pimeta (maybe I'll build a ppa, m^3, or MH after I build the ppas). The monofied has a nice picture already for board etching and a circuit that is a bit less complex than the DB. I don't necessarily need all the features of the DB design. Also, the monofied looks pretty inexpensive and if I don't like it, I'll just toss the buf634s back in it.

Quote:

Also worth noting is that on the Pimeta, that buffer layout will only fit in the ground channel with 10mm diameter electrolytic caps beside it, not 12.5mm

Yeah, I did notice that. Any problems leaving a buf634 on the ground channel and running discretes on the L/R channels? I might end up building a little daughterboard with all three channels on it anyway, depending if I keep the AD843s on a browndog 2->1 on the L/R opamp socket. Since it is a home unit in an oversized case, I have plenty of room to topple the caps out of the way also.

Edit - vixr: YGEM

 

[mono]

Quote:

Originally Posted by quicksilver96 After reading everything on the DB page, it seems a little more than I want to undertake for just the pimeta (maybe I'll build a ppa, m^3, or MH after I build the ppas). The monofied has a nice picture already for board etching and a circuit that is a bit less complex than the DB. I don't necessarily need all the features of the DB design. Also, the monofied looks pretty inexpensive and if I don't like it, I'll just toss the buf634s back in it.

FWIW, after building with BC327/337, I've now built with 2N5087/5088 & 2N3904/3906, and found the BC327/337 sounded better, more dynamic.

Quote:

Yeah, I did notice that. Any problems leaving a buf634 on the ground channel and running discretes on the L/R channels?

No that'll work fine, though you could just use the 10mm caps on those last two cap positions instead of swapping out all 4.

Quote:

I might end up building a little daughterboard with all three channels on it anyway, depending if I keep the AD843s on a browndog 2->1 on the L/R opamp socket. Since it is a home unit in an oversized case, I have plenty of room to topple the caps out of the way also.

I thought about a daughterboard for all 3 buffers and L/R single channel opamps but that didn't take care of the ground channel unless the daughterboard was overtop of the output pins which I wanted to avoid. Plus, I already had it modular to fit in all 3 positions. I think all 3 positions should be filled if possible, including ground.

I have two AD843 on a homemade browndog-like adapter and the buffer board pictured fits next to it on a Pimeta, but barely. The dashed lines around it are cut lines and need to completely disappear for best fit, including against the larger Alps Rk27 pot if you used that instead of RK09 or the Panasonic %.

Shoot me a PM with your address and I'll send you a free set of boards. I hadn't decided if I was going to make another run of these same boards and/or to make a daughterboard then use the remaining single boards for ground channels.

Unfortunately in retrospect it seems I didn't bend the Pimeta's integral CCS JFET leads enough or push them down hard enough before soldering so the Pimeta testbed I was using may need a minor operation to fix that before I can get enough clearance for a daughterboard like that on it, unless it stood farther off the mainboard than I want it to... but then again, it's not like my JFETs are sticking up very high either, probably about the same as most Pimetas.

 

[vixr]

No sweat... Let me know if you change your mind about the DBs... Rob

 

[Pars]

Quote:

Originally Posted by mono FWIW, after building with BC327/337, I've now built with 2N5087/5088 & 2N3904/3906, and found the BC327/337 sounded better, more dynamic.

I'd agree. I built a set with Dynalo transistors (2SA1015 / 2SC1815) that sound good, but the set I built with 2N3904 / 2N3906 doesn't sound very good. If you are in the US, you'll notice that the BC327/337 aren't real easy to come by. I bought a bunch of them (Philips -25s) off a UK ebay seller, and these are my favorites that I have tried, although the SA1015 / SC1815 boards sound pretty close. If you want some of the trannies, I'd be happy to send you a few.

 

[quicksilver96]

I think that mouser carries the BC327/BC337:
BC327 Fairchild -40/Vishay -16

BC337 Fairchild -40/Fairchild -16

Are these not what I am looking for? Does it make any difference in this project whether I use a -16 or -40 as long as both of the complementary pair are the same?

Quote:

No that'll work fine, though you could just use the 10mm caps on those last two cap positions instead of swapping out all 4.

I'll just topple the last 2 over, there is nothing for a good inch to that side of the board anyway. Figured if I am doing the L/R, I might as well do the ground also.

Quote:

I have two AD843 on a homemade browndog-like adapter and the buffer board pictured fits next to it on a Pimeta, but barely. The dashed lines around it are cut lines and need to completely disappear for best fit, including against the larger Alps Rk27 pot if you used that instead of RK09 or the Panasonic %.

Yeah, I have that RK27 in there too, so the monofied sijosae will fit with an 843 pair and an RK27 if the board is trimmed close?

I had to put a slightly taller mil-max dip socket in the L/R position to accomodate the adapter clearance over the surrounding FETs. I might have to rework a lot of this whole area in order to use the buffers and 843s. I may just have to abandon the 843s if it means not being able to use a discrete buffer. I'll figure a way to cram it all in there...

 

[mono]

I think I used the -16 variety and they had hFE around 200.

Yes it is possible to fit a pair of single channel opamps, and RK27 but since I don't have the real browndog I can only say the two opamps themselves fit. If you built the buffer with the transistors leaning too far towards the pot, it might not fit or if there were enough slack in their leads they could be bent away from the pot. It will be close but is possible.