well, not really
in IP, upper JFet trio is signal half of buffer, lower JFet trio is source CCS load for upper one
so, buffer, no gain
here, we have one stage comprised of single 2SK2145, drain loaded with simple resistor; second stage is BD140, colector loaded with CCS (in lower rail)
so, apples and oranges
anyhow, one brain was the same for both
even if I often feel as Zaphod
edit: in fact, Zaphod is a little kitten for Mighty ZM; in all of my musings, I'm often switching between Infinite Improbability and Futile Attempts Method; that's probably then Zaphod^Zaphod
in IP, upper JFet trio is signal half of buffer, lower JFet trio is source CCS load for upper one
so, buffer, no gain
here, we have one stage comprised of single 2SK2145, drain loaded with simple resistor; second stage is BD140, colector loaded with CCS (in lower rail)
so, apples and oranges
anyhow, one brain was the same for both
even if I often feel as Zaphod
edit: in fact, Zaphod is a little kitten for Mighty ZM; in all of my musings, I'm often switching between Infinite Improbability and Futile Attempts Method; that's probably then Zaphod^Zaphod
Proof is in Da Pudding!
Got pcbs ystrday and, instead of being useful - packing some kits I ought to send ....... here's I, Mighty ZM, simply incapable to resist tiniest amount of Guilty Pleasure
So, here's I, Mighty ZM, assembled final pcbs of Scryer
Today's version with THF51, tomorrow's one (more likely day after tomorrow) with 2SC2087C
so, final schm looks as:
Got pcbs ystrday and, instead of being useful - packing some kits I ought to send ....... here's I, Mighty ZM, simply incapable to resist tiniest amount of Guilty Pleasure
So, here's I, Mighty ZM, assembled final pcbs of Scryer
Today's version with THF51, tomorrow's one (more likely day after tomorrow) with 2SC2087C
so, final schm looks as:
Attachments
then some GigglyWiggly;
rails aroundish 23Vdc (yeah, 24vdc nominal)
Iq 1A8
again - today - THF51
initial DC Offset nothing to worry about ( less than 100mV in any case)
same during Power Off
DC Offset stability and Iq stability - in time, temperature and rails change domain - not even need to think about
now, if you compare today's graphs with those on first page of thread - clipping behavior is starting practically step before; that's due to fact that today my crummy PSU of T-Bed was loaded with both channels ( even if youst one being under test, second one still pulling steady juice) and that's sole difference - in initial measurements ( page one) I had just one channel connected to PSU
OK, graphs for 8R Load:
rails aroundish 23Vdc (yeah, 24vdc nominal)
Iq 1A8
again - today - THF51
initial DC Offset nothing to worry about ( less than 100mV in any case)
same during Power Off
DC Offset stability and Iq stability - in time, temperature and rails change domain - not even need to think about
now, if you compare today's graphs with those on first page of thread - clipping behavior is starting practically step before; that's due to fact that today my crummy PSU of T-Bed was loaded with both channels ( even if youst one being under test, second one still pulling steady juice) and that's sole difference - in initial measurements ( page one) I had just one channel connected to PSU
OK, graphs for 8R Load:
Attachments
Sound ......
my permanent amp in WShop is Papa's Koan; fed either with Iron Pumpkin or Black Beauty based passive pre
when I switched to Scryer - difference is immense - even if I know that bloody THD levels are sorta comparable
go figure
anyway - Scryer is Spooky; fun, foot is tapping ; and another foot is tapping
first difference to Papa's Koan - much more details, scene much closer in the face, immediate
anyway, 'nuff of audiophile vocabulary - or I'm going to mention thundering bass, silky palpable mids and sparkling tinkling highs
@Plott ....... good question - I even put Wasser in Coffee machine, then I realized that I don't have more paper filters ......... who's gonna go to Store to buy them, when amp is in the making
my permanent amp in WShop is Papa's Koan; fed either with Iron Pumpkin or Black Beauty based passive pre
when I switched to Scryer - difference is immense - even if I know that bloody THD levels are sorta comparable
go figure
anyway - Scryer is Spooky; fun, foot is tapping ; and another foot is tapping
first difference to Papa's Koan - much more details, scene much closer in the face, immediate
anyway, 'nuff of audiophile vocabulary - or I'm going to mention thundering bass, silky palpable mids and sparkling tinkling highs
@Plott ....... good question - I even put Wasser in Coffee machine, then I realized that I don't have more paper filters ......... who's gonna go to Store to buy them, when amp is in the making
got back from Schooling trip in time, so I can do what I did planned for today
So, Scryer with 2SK2087C ...... or - Best and Cheapest 300B amp you'll ever find and ever need and ever have
So, Scryer with 2SK2087C ...... or - Best and Cheapest 300B amp you'll ever find and ever need and ever have
Attachments
same rails and Iq as with THF51, as shown in posts of yesterday (#44 .....)
compare graphs with THF51 iteration ...... sole difference is slightly lesser power with 2SK2087C, behavior already seen in https://www.diyaudio.com/community/threads/zms-2sk2087c-musings-phase-one-sissysit-as-cradle.386654/ and https://www.diyaudio.com/community/...ngs-phase-two-sissysit-42-as-a-cradle.387767/
as in Scryer overall NFB is doing its work, performance up to clipping with small one is same as with big one
time for GigglyWiggly pics
Scryer with 2SK2087C, 8R load
compare graphs with THF51 iteration ...... sole difference is slightly lesser power with 2SK2087C, behavior already seen in https://www.diyaudio.com/community/threads/zms-2sk2087c-musings-phase-one-sissysit-as-cradle.386654/ and https://www.diyaudio.com/community/...ngs-phase-two-sissysit-42-as-a-cradle.387767/
as in Scryer overall NFB is doing its work, performance up to clipping with small one is same as with big one
time for GigglyWiggly pics
Scryer with 2SK2087C, 8R load
Attachments
Now, before someone ask me this simple question, to explain simple thing - why difference in output power between (say) SissySIT (42) and Scryer, whichever SIT used?
Nature of autoformer based FE (in SissySIT and similar amps) is that output stage modulation can easily hop above positive rail and below negative rail; for instance - see clamping diodes in F4 (no autoformer inside, but Pa needed to protect little amp from excessive modulation on input) then clamping diodes in M2 ......... I did use same approach in my first musings with M2 but later decided that gate clamping zeners of output mosfets are capable of doing same duty (preventing modulation overvoltage)
so, autoformer FE based amps are capable to trash output devices all the way to rails, leaving only internal losses of devices itself to play a role in decreasing voltage window
now - Scryer - having active FE, output node is nested in collectors of BD140 (Q102) and adjacent CCS BD139 (Q103);
of course that said stage is working in A Class, so gigglywiggly voltage across R110 (approx. 1V55) is really not that much wiggly........ consider it more or less constant
then you have comparative voltage stolen from lower rail, across group of resistors in emiter of Q103
add to that few hundreds of mV across both Q102 and Q13, there lies reason for smaller output swing, thus lower output power of active Fe vs. autoformer one ......
edit: Silly ZM - simple question, simple thing ............. and even if I tried, I couldn't explain it more complicated
well, maybe some another day
Nature of autoformer based FE (in SissySIT and similar amps) is that output stage modulation can easily hop above positive rail and below negative rail; for instance - see clamping diodes in F4 (no autoformer inside, but Pa needed to protect little amp from excessive modulation on input) then clamping diodes in M2 ......... I did use same approach in my first musings with M2 but later decided that gate clamping zeners of output mosfets are capable of doing same duty (preventing modulation overvoltage)
so, autoformer FE based amps are capable to trash output devices all the way to rails, leaving only internal losses of devices itself to play a role in decreasing voltage window
now - Scryer - having active FE, output node is nested in collectors of BD140 (Q102) and adjacent CCS BD139 (Q103);
of course that said stage is working in A Class, so gigglywiggly voltage across R110 (approx. 1V55) is really not that much wiggly........ consider it more or less constant
then you have comparative voltage stolen from lower rail, across group of resistors in emiter of Q103
add to that few hundreds of mV across both Q102 and Q13, there lies reason for smaller output swing, thus lower output power of active Fe vs. autoformer one ......
edit: Silly ZM - simple question, simple thing ............. and even if I tried, I couldn't explain it more complicated
well, maybe some another day
Last edited:
it can be only in 2SK2087C clipping region ...... as far as I did hear, and as far as I can see from graphs
practically - if you have speaker goodenough for purpose ( read - sensitive and efficient), there is no difference between small one and big one
same applies to application in SissySIT (42) ........... there is slight difference seen on graphs, but frankly I couldn't hear it ....... it was day after day comparison ( not direct one) so based on auditive memory
in fact, I don't care for auditive memory; goal of my comparisons when judging between similar things is - how much fun and joy I have - and if there is enough of both, it's close
so, SissySIT (42), no loop NFB, some measuring difference between THF51 and 2SK2087C (in linear region of smaller one), which I did not hear
Scryer, having loop NFB, no measuring difference between THF51 and 2SK2087C (in linear region of smaller one), and I did not hear any
Last edited: