lhquam. During his tenure at Threshold Corporation, Mr. Pass used 2N3440 [NPN] and its complementary 2N5415 [TO-5 packages] as drivers to STASIS output stages which operated at +/-50V PSU.The ON-Semi BC639/BC640 datasheets rate the power dissipation at 625mW. Fairchild rates the BC63919 at 830mW and the BC640 at 1000mW. The ON-semi BC550C/BC560C datasheets rate the power dissipation at 625mW. So which devices are you proposing for the higher voltage swing design?
...which devices...
I mostly trust to manufacturers datasheets, but in my experience BC639/640 are noticeably sturdier pair than BC550/560 - I use BC639/640 produced by Philips (0.83W declared).
There's nothing wrong with BC550/560 too, if you make an effort to cool them a bit, in interest of long-term reliability .
Of course, you can experiment with other BJT types - I'm sure there are a lot of good BJTs with adequate Pd, hfe, BW...
One more thing - I forgot to post the version with k246/j103. The buffer on the output is obligatory here.
I'd like to build version with k246/j103. I'd like to know the following:
1. Does both R9 and R24 are required?
2. How big heatsinks is recommened for 78/79xx and mosfets?
3. Is it good enough to use one power supply or one per channel is recommended here (not to say obligatory)?
4. I put C6 as film instead of elco. It that ok?
Many thanks.
Attachments
... I'd like to know...
1. Yes if you keep C7 between them because Jfet's gate needs a firm DC reference
2. I didn't use them, but if you think you need them, something like this will be OK
3. One for both channels is fine, but if you are willing to try, one per channel might be marginally better
4. OK

Last edited:
I'd like to build ...
Your sch. is messed up: +/-24V should be PS for the buffer.
Your sch. is messed up: +/-24V should be PS for the buffer.
Oops, my fault! Many thanks Juma for this and previous post. New schematic is in attachment.
Attachments
prasimix,
watch your q6,q8
Yes, they should be of NPN type, BC560C won't work there.
I dont think they are. I looked into getting a large quantity to match and they seemed to indicate they wanted to limit thngs to current distributors. BTW, I will be opening GB thread for the boards. Waiting for final numbers.
Sorry for the OT Juma😱
Did you try this pre together with your K2013/J313 based F5?
Yes, it's my current setup and I enjoy it very much. Few days ago a friend came to visit and his comment was: "it sounds expensive" ... 😀
prasimix,
watch your q6,q8
Huh, "copy&paste" error. Here is new schematic with two channel and single PS. First version of PCB single layer (no jumpers required) if finished. If this layout is correct I can also make version with single channel if someone find that interesting.
Maybe everything is too small and I put very small heatsink on 78/79 to spare some space.
Thanks in advance for comments and suggestions.
Attachments
Huh, "copy&paste" error. Here is new schematic with two channel and single PS. First version of PCB single layer (no jumpers required) if finished. If this layout is correct I can also make version with single channel if someone find that interesting.
Maybe everything is too small and I put very small heatsink on 78/79 to spare some space.
Thanks in advance for comments and suggestions.
Very nicely done.
Today I tested Indra1's proposal (post #76) and finally there's an amp without feedback loop that satisfies me completely (no objections to sound at all). It drives the 4 Ohm, 3-way sealed boxes magnificently.
noted

Very nicely done.
Thanks buzzforb. I'll further modify this one to be capable of accepting additional RC (R14, C3 from post #66) and possibly increase PS section to host bigger heatsinks as Juma suggested.
Hi Juma,
Thanks a lot for the great bjt cascode version implementation of the PASS BAF2013 LSK preamp.
I came up with an idea and hope that you do not mind my using photoshop to merge several of your schematics of LSK Preamp, F5 meets Buzquito and addition of 2x150R at (R23 & R24) for an integrated with no global NFB. Any comments or suggestions for a build with sk/sj pair at idss about 10ma?
Thanks in advance.
indra, are you going test it ?
any reason not to try configuration, but with irfp240/9240 as output fet, in conjunction with higher pre rails and r9,23,24, changed for proper bias and gain?
...though input jfets might not like higher rails that much...
...though input jfets might not like higher rails that much...
Attachments
Last edited:
indra, are you going test it ?
Absolutely, but to be honest, most of my builds can be categorized as the turtle type.
My work require me to frequently be out of town for extended period, was out of town mid of november to end of december. When I got home, I have to remind myself that I am a grandpa already and have to do this patiently and carefully (for me, patience supply is getting leaner with age, lost a 2sk79 last week when I was chasing a barely audible hum on a Baby AKSA amp style - 2sj74gr diff input, 2sk79 vas, darlington out - the second project that started on august after a 25 yr hiatus, had to revert back to 2sc2271 for the vas, and still no joy on the hum)

Just completed collecting required parts this week (still needs apropriate heatsink though), but will be out of town beginning next week. Again. Can still post messages and drool over the forum, but no hot iron. Like I said on post #137, I simply don't have the time I want for this hobby, but I savor whatever time I get and enjoy this fantastic journey as the scenes unfolds.
- Home
- Amplifiers
- Pass Labs
- LSK pre - BAF 2013