beta22 phase1

Jan 25, 2010 at 6:02 PM Post #16 of 57
Quote:

Originally Posted by jh4db536 /img/forum/go_quote.gif
this was a little bit more difficult than i thought. for some reason i had a really hard time with the box caps, the damn solder wouldnt flow correctly i dont know why (no matter what i did it flowed upwards and was repeling the circuit board even though gravity was on my side).


Soldering anything to the ground plane on these large PCBs is always a hassle...... they just suck so much heat out of your iron.

Try increasing the temp on your iron and get some flux on the board first, and you should be A-O-K.
 
Jan 29, 2010 at 7:53 AM Post #17 of 57
i am definitely getting better at this
smily_headphones1.gif

i am saving the transistors for last since i plan to match them up real good when i get my fluke...this harbor freight dmm sucks for accuracy me thinks.
i barely have time for this hobby work is too much
my 'wip'
IMG_1371.jpg

IMG_1370.jpg

IMG_1367.jpg

IMG_1368.jpg

IMG_1369.jpg
 
Jan 29, 2010 at 9:36 AM Post #19 of 57
@jh4db536 - The kit is from Glass Jar Audio?
 
Jan 29, 2010 at 4:46 PM Post #21 of 57
Quote:

Originally Posted by jh4db536 /img/forum/go_quote.gif
Jeff is awesome. the idea of making a b22 wouldnt be feasible for me if GJA wasnt around.


Considering the unbelievable amount of time I spent ordering all the pieces to assemble my ongoing b22 project, I really regret I wasn't aware of the existence of GJA !!! Seriously.. selecting and putting in the basket all the components required, and also finding turnaround when mouser or digi key were out of something... I've spent several evening doing this !
Sure I gained some knowledge on what is available on the market, differences between components and so on... but what a wasted time !
confused_face.gif
 
Jan 29, 2010 at 9:21 PM Post #22 of 57
I fully sympathize with you LeMat, as I went through the process on my own as well. I am certain I would have saved a significant amount of money had I gone through Jeff. At the same time I am happy with how much I learned having put in the work for it.
smily_headphones1.gif


@jh4db536- Keep it up! I'm enjoying watching your build process as well. And yeah... that ground plane took me a bit to get used to as well...
 
Feb 1, 2010 at 7:43 AM Post #23 of 57
just a few updates, too tired for pics...
i am almost done with all 4 boards except for the installation and matching of the BJT and MOSFETS.

Matching Q1-Q4 was quite an interesting task. i borrowed a Fluke (now i know how inaccurate my harbor freight DMM is), used my opus dac powersupply as a 12v regulated source to test the JFETs. i got them all idss to 0.1mA of each other. i guess that is better than 10%?

it amazes me on how sensitive the JFET are to temperature or load. the readings took about 3 minutes to settle to an exact voltage (usually lower than when i first plugged it in). i let them settle for 3 minutes each before writing down their idss and then matched them up. i find though that inside the case under class A biase they will probably be seeing much higher temps. although matching them like this, they should react similarly (i hope) in the case where theyre all exposed to the same thing. i infer that these components are located on the complete opposite side of the board from the mosfets for this reason. 2 mosfets per block hehe.

i matched the 4392 N JFETs while i was at it. those things have pretty bad consistency....


i neeed to go buy some resistors and alligator clips so i can match the mosfets and BJTs then ill finally be able to fire this thing up and dial it in.

has anyone made a water cooled b22 yet? lol. the only time i listened and saw a b22 at Can jam 09 that thing runs so hooot. this will defnitely blow the budget. probably not worth it?
Newegg.com - Enzotech WMST-66NP Mosfet Waterblock - Water Cooling

i am only getting 29volts out of my o22. i think my zener diode sucks. will it run okay on 29v?
 
Feb 1, 2010 at 8:34 AM Post #24 of 57
Quote:

Originally Posted by jh4db536 /img/forum/go_quote.gif
i am only getting 29volts out of my o22. i think my zener diode sucks. will it run okay on 29v?


That's actually about right. The default R8 and R10 values for "30V" output are 10K and 6.81K, so that makes the regulator's gain = (10K / 6.81K) + 1 = 2.468. Assuming a perfectly accurate zener, then the output voltage would be 12V * 2.468 = 29.62V. If your zener voltage is just 2% lower than spec'ed, then you'd get 29.0V. That's quite good, actually.

Absolute supply voltage accuracy is not required. β22 was designed for nominal rail voltage range of ±24V to ±30V.

Btw, the only transistors that require careful matching are Q1-Q4 for lowest DC offset. You don't have to be pedantic about matching the rest of them, due to β22's cascoded topology in every stage.
 
Feb 2, 2010 at 12:51 AM Post #26 of 57
Quote:

Originally Posted by Magsy /img/forum/go_quote.gif
The transistors really pi$$ me off, I like to keep them all straight and neat but it is so hard to do, plus the holes for the CRDs/zeners are too big. I straightened all my box caps afterward by reheating and moving them.


What I do is solder just the middle pins, then straighten up the transistors before soldering the remaining two pins. Works just as well for any other components that want to end up crooked. Plus it gives you another chance to review all the components you've just placed to make sure they're in the right spots, before you solder them all the way in.



Quote:

Originally Posted by amb /img/forum/go_quote.gif
Btw, the only transistors that require careful matching are Q1-Q4 for lowest DC offset. You don't have to be pedantic about matching the rest of them, due to β22's cascoded topology in every stage.


But where's the fun in that?
tongue.gif
 
Feb 7, 2010 at 8:50 AM Post #28 of 57
IMG_1390.jpg

IMG_1392.jpg

IMG_1394.jpg


finally done that was long but i am sad it was fun stuffing the boards.
i matched everything on the list down to q15,16.

since i only plan to run 1 sigma22, i upgraded the heatsinks on it - this is serving only headphone duty. im still undecided on the chassis...
 

Users who are viewing this thread

Back
Top