Roll Call: Who's building, built, or thinking of building a beta22?
Nov 20, 2013 at 8:43 AM Post #3,091 of 3,218
>  4 boards for balanced won't fit.
Maybe but he's talking 2-channel / passive ground.  A nice way to go - no need for expense/complexity/size of "extra" channels nor the impact of going balanced on source gear, cables, headphone wiring, etc.
 
Will say you'd be hard-pressed to find a better supported (amb forums, shop, glassjaraudio kits, etc) SS DIY reference amp for dynamic cans than the β22.  Lots of experience with the amp posted around as well.  For me a 2-channel β22 is part of my end-game setup & I've built/listened to a lot of amps.  If you go β22 recommend using 1R R34/R35 (esp. w/TRS jack) & adding zobel network (to avoid perhaps having to later).
 
Whichever way you go from our PMs I bet an σ22 is going in the little box....  Something like this:
 

 
Nov 20, 2013 at 7:58 PM Post #3,092 of 3,218
You betcha! :). I've got to read up on the details of the circuit config...sounds like I have some tailoring options to lock down.

Yes, just a plain 2 channel to start. If I integrate the heatsinks I might try to get clever.....but for now I just want to get a working system built.

Yes, both cases are from Chinese eBay sources.

>  4 boards for balanced won't fit.
Maybe but he's talking 2-channel / passive ground.  A nice way to go - no need for expense/complexity/size of "extra" channels nor the impact of going balanced on source gear, cables, headphone wiring, etc.

Will say you'd be hard-pressed to find a better supported (amb forums, shop, glassjaraudio kits, etc) SS DIY reference amp for dynamic cans than the β22.  Lots of experience with the amp posted around as well.  For me a 2-channel β22 is part of my end-game setup & I've built/listened to a lot of amps.  If you go β22 recommend using 1R R34/R35 (esp. w/TRS jack) & adding zobel network (to avoid perhaps having to later).

Whichever way you go from our PMs I bet an σ22 is going in the little box....  Something like this:


 
Nov 26, 2013 at 8:32 AM Post #3,093 of 3,218
Think its been a while since there's been detailed activity here for an amp that uses A/C mains.  Four or five people contacted me about the case & may be building σ22s. 
If any are newish builders respect the importance of safety, I'll mention a few particulars of my build(s) below. 
 
There are good threads/posts/URLs that detail A/C/mains safety in DIY & builders new to it should certainly seek them out.  E.g. in these builds I used insulated connectors to IEC, a proper chassis ground stack (dedicated screw, lock washers/lock-tite, ring lug, nuts), removed anodizing at chassis ground screw & at parts of enclosure panel mating areas + underside of screws/threads & assured the entire σ22 exterior contacts safety ground.
 
If building σ22 and/or β22 it pays to carefully review their related info on amb's site & triple-check all your work before powering up.  E.g. a swapped/grounded MOSFET, swapped/maladjusted trimpot, unintended ground/short, slipped probe, pierced solder mask during adjustment/testing are some things to avoid. 
 
Anyway, in case anyone is considering 2-channel β22 in same case as above know there's significant time overhead to do it (& probably some cursing as well:)
 

 
It's a nice quiet build.  Note copper underside of ε12/zobel network/RK40 board is the star ground w/only connection to chassis ground at its mounting screw by headphone jack.  Every "module" can be removed w/o soldering which came in handy for tweaking/lowering the ε12 latch LED brightness:)  Runs cool & allows trying higher biasing.  BTW I post some other build & parts details here:
http://www.amb.org/forum/post-pics-of-your-builds-t44-690.html#p21728
 
Nov 26, 2013 at 9:44 AM Post #3,094 of 3,218
 
Beautiful build.

Great attention to detail.
 
p.s. Your table look just like mine 
wink_face.gif

 
Nov 26, 2013 at 5:47 PM Post #3,098 of 3,218
Thanks guys.  Yep, took quite a while to plan/design/execute the layout - very happy w/the results though.  Even w/after all the builds I've done its difficult to estimate in advance how long a build will actually take:)
 
>  p.s. Your table look just like mine
Heh, yeah its a an old, cheaply "veneered" pressboard repurposed desk but it does the job.  I should take a shot for the pics-of-your-workspace thread.  It's tight, like my builds, but well organized:)
 
>  Is that a Noble pot?
No, its a NOS Alps "Black Beauty" RK40.  Fun fact: Only upon creating a Diptrace lib component for the RK40 did I realize the 27 in RK27 & 40 in RK40 are the pots' widths in mm.  And a lot of Alps pots have passed through my hands:)
 
Am enjoying the amp at the moment but may give someone else the opportunity:)  Now that I've the patterns/process for this particular build down I could spend a couple/few dozen hours making another.
Note: Next time I'd use allen cap screws for heatsink / "L" bracket attachment - easier clearance WRT β22 & σ22 components.  Also though I noted ease of "module" removal, the liberal use of screw terminals sure made bench testing/assembly easier as well.
 
Nov 27, 2013 at 3:08 AM Post #3,099 of 3,218
  Thanks guys.  Yep, took quite a while to plan/design/execute the layout - very happy w/the results though.  Even w/after all the builds I've done its difficult to estimate in advance how long a build will actually take:)
 
>  Is that a Noble pot?
No, its a NOS Alps "Black Beauty" RK40.  Fun fact: Only upon creating a Diptrace lib component for the RK40 did I realize the 27 in RK27 & 40 in RK40 are the pots' widths in mm.  And a lot of Alps pots have passed through my hands:)
 

i´m really curious about this pot.
i´ve always heard good things.
it´s a shame that not exists a quad version
 
Nov 27, 2013 at 6:56 AM Post #3,100 of 3,218
Ah the black beauty was a definite step up from the RK27 standard but still behind the decent steppers out there and the smaller Alpha pots. ALP's is in a way similar to 80's Philips, discontinuing things here and there with zilch backup stock or very very limited quantities.
 
Dec 5, 2013 at 12:06 AM Post #3,101 of 3,218


My B22 is coming along, this is a pic of the set up phase for one of the boards. I found that mine was no exception, setting the bias current and offset voltage took quite a bit of fiddling. It takes a long time for the temperatures to stabilize. On the second board I set the initial current quite high and let it set for ~20 min before making any further adjustments. This seemed to simplify the process, though I would not recommend doing so without having a reference ( measurements from a known good board ).
 
I have the L and R boards finished, the Gnd board is almost done but i am waiting on a heat-sink.  I couldn't resist, I went ahead and wired it up in a 2 channel passive ground configuration and gave it a listen this afternoon. 
I was using my "test" headphones, a set of bose, and using my iphone as the source. It worked great. I wont go into how I thought it sounded, as both the cans and the source are not anywhere near ideal, and my experience with silmic caps is that they need a good dose of burn in, but it didn't do anything wrong and seemed to do a lot right
dt880smile.png
.
 
Hopefully I will be able to get the Gnd board finnished soon and start working on casing it up.
 
preliminary specs:
 
3ch b22, single s22 PS @ 48v ( 100 va ) , bias current ~160ma ( .075mv through .47 ohm resistors ) .
 
Dec 5, 2013 at 10:57 AM Post #3,103 of 3,218
It is usually said that if you plug/unplug your TRS quickly then you will probably be ok.  I didn't take any chances and always powered down if I was going to switch out the headphones though.
 
Dec 5, 2013 at 11:07 AM Post #3,105 of 3,218
I've read that balanced terminated outputs are fine to swap while the amp is on. Otherwise, SE should be turned off and then swapped. Has something to do with grounding, would have to go back and check out where I read that. 
 

Users who are viewing this thread

Back
Top