[Ktechlab-devel] Use case analysis, 2A3 amp. =P

[Alan Grimes]

A few posts ago, I think I sketched out a 2A3 amp I was working on.

First the reference material:

http://www.circuitcellar.com/microchip2007/winners/MT2209.html
(design is VERY different from mine, the key items are the SMPS and the
file formats for source code and schematics...)

http://www.jensen-transformers.com/as/as088.pdf

http://cinemag.biz/line_input/line_input.html
http://cinemag.biz/line_input/CMLI-15-15B.pdf

http://edcorusa.com/products/transformers/cxpp/cxpp50-8-3_4k.html
(I might custom order a pair with a smaller core to keep the size of the
amp down...)

http://scottbecker.net/tube/sheets/049/2/2A3.pdf
http://scottbecker.net/tube/sheets/093/6/6SL7GT.pdf
http://scottbecker.net/tube/sheets/137/6/6SN7GTA.pdf

First issue is the tubes, but then nobody cares about tubes so we'll
simply mention it and move on...

Second issue are the transformers and the complex AF features of them
such as bandwidth and frequency dependent phase delay....

This brings up a critical issue with the simulator.

Right now it attempts to do an instant-by instant approimation. This
gives it a frequency ceiling of 1khz... =\ To go further, it needs to be
able to switch to static analysis techniques to approximate filter
responses clear out to 200khz or so. Furthermore, it needs to be able to
compute general parameters of frequencies all the way into RF. -- this
is critical to the SMPS design...

Morgan Jones (no on-line reference, I'm aware of), uses voltage
regulators all over the place to drop as much as 150 volts off the B+
rail for the voltage gain and driver of his amp... =P

The power supply for the amp I'm thinking of needs a 300V B+ rail, a
secondary B+ rail at around 150v, a B- rail of -300v (maybe even
-350v!), three separate 6.3v supplies, and four monster-current 2.5v
supplies... So that's like 10 supplies, or maybe 9 and and absurdly tall
regulator, not to mention the startup ordering required to safely bring
up the very expensive output tubes... (they need to be warmed up and the
B+ should be brought up slowly...)

Conventional SMPS design software only allows you to specify voltages up
to a pathetic 100v! =(      sad.

So a major benchmark for the Ktechlab simulator will be the ability to
simulate any of the major SMPS topologies and come with reference
designs for PFC input and DC-DC converters with output voltages all the
way up to 1.5 KV.

First, however, we need to design and implement the new parts library. =(

-- 
New president: Here we go again...
Chemistry.com: A total rip-off.
Powers are not rights.