Most Greedy Boy, of them all... or (there is no) DEFiSIT of Papa's Koans

Jump to Latest
#441
ZM: I have lost track of which circuit was in the final build for which you conducted listening tests. Can you point me to the post containing the schematic?

What became of the DEFiSIT approach of eliminating the bias servo by using depletion/enhancement mode thermal drift cancellation?
 
#442
I was occupied with daydreaming of 4 similar and still different enough amps:

- Babelfish M25

-SEF/Babelfish M2 , practically two (minimally different) amps on same pcb - SEF being my iteration of M2 with Schade-d lower mosfet ........ of which I really don't know where last iteration of schematic is , so I'm enclosing it here ; pcbs to arrive next week , when I'm going to open dedicated thread at present everything ; Babelfish M2 being - well , plain M2 with few twists of minor importance - 2SK2145BL input buffer cascoded , 3(4) options for input xformer

- will also try soon , on same SEF/BM2 pcbs , iteration with working title SUFI-SIT - with THF51-S up and plain mosfet down ; fingers crossed for that one , too 🙂

-LuDEF amp ....... what else than cascoded LU1014 up , and mosfet down

Babelfish XA25 , dunno when I'm going to squeeze that one on T-Bed 🙂

except BXA25 , all of these are somewhere back in the thread , but lets just repeat SEF amps schm this time

time will tell for other ones

edit:

damn , is it 6 , instead of 4 amps ?

:rofl:

editedit: I didn't pursue elimination of biasing circuit , simply because of difficulties in obtaining proper parts combinations ;

that's viable approach if you're manufacturer (as Pa is) , having semi-finite quantity of parts on hand , to choose from and accomodate in finite number of amps

I'm just eeenyweeeny version of Papa , so ........ I'm backing my hands at least 6C earlier than him

:rofl:
 

Attachments

  • SEF schsshot.jpg
    SEF schsshot.jpg
    159.4 KB · Views: 432
Last edited:
#445
Zen Mod said:
...

editedit: I didn't pursue elimination of biasing circuit , simply because of difficulties in obtaining proper parts combinations ;

that's viable approach if you're manufacturer (as Pa is) , having semi-finite quantity of parts on hand , to choose from and accomodate in finite number of amps
...

:rofl:
Click to expand...

Here is what I was thinking:

Use bias tricks, such as in DEF-Amp post #43 DEF Amp with one or more (IXYS) depletion mode NFETs over a single IXYS enhancement-mode hockey-puck PFET. I do not know how well the temperature drifts would cancel. Possibly also use Shade feedback with the NFET(s). If multiple NFETs are required, the each NFET can have a Vgs offset adjustment as in post #43.
 
#446
damn ....... typed war and peace, then clicked wrong .....

anyway - Mighty ZM unable to find any other way to mimic Wakoo current distribution , parts are having in original Pa's DEFiSIT , and exactly that (current distribution) is most important part of behavior I strive for

what I did show is what I was able to concoct with parts I have (or will have) on hand

chasing another (one or more ) preputium made of parts is outa question
 
#447
I guess I will order some IXYS depletion mode parts and give it a try using an XA25 style FE with feedback around the FE only.

Do you have any SPICE models for Papa-SIT or any of the other SITs? I would like to use them for comparative simulations against Shade FB around various NFETs, such as the IXYS depletion-mode FETs.
 
#449
Actually I am looking for positronic valves. Here is a copy of a message I just sent to Generg:

[FONT=Helvetica, Arial, sans-serif]I am still trying to understanding how the push-pull SIT/PFET combination (DEF-SIT) compares to a depletion-mode NFET/enhancement-mode PFET combination (DEF). It might be like comparing a push-pull stage of a triode follower over a positronic pentode follower to a push-pull stage of a pentode follower over a [/FONT][FONT=Helvetica, Arial, sans-serif]positronic pentode follower. I suppose you might be able to use transformers to flip phases and eliminate the need for the positronic valves. But I do not know how the sounds might compare.[/FONT]
Click to expand...
 
#451
lhquam said:
I guess I will order some IXYS depletion mode parts and give it a try using an XA25 style FE with feedback around the FE only.

Do you have any SPICE models for Papa-SIT or any of the other SITs? I would like to use them for comparative simulations against Shade FB around various NFETs, such as the IXYS depletion-mode FETs.
Click to expand...

just take a look at our Mr. MR's site

he's , like you , number friendly mind and exploiting that .... while I'm more back of napkin scribbling type ....... so I got them from there
 
#457
From post #213 http://www.diyaudio.com/forums/pass-labs/314822-greedy-boy-defisit-papas-koans-post5288307.html:

Zen Mod said:
Dunno Pa , just trying to translate what sim is showing 🙂

I'm pretty happy with DEFiSIT sound :devily: and I'm not so sure when I will have opportunity to steal it from living room (and measure some) - right now whoomping and roaring LOTR 3 Extended

...
-Real McKay DEFiSIT is defisitly etalon of misbehaving
Click to expand...

Did you ever verify that "-Real McKay DEFiSIT is defisitly etalon of misbehavin" with actual measurements?
 
#458
if anything else is behaving well (as defined by model) in simulations , why that should be different?

frankly - I didn't check in vivo ....... simply because DEFiSIT is in use every day and I was lazy to pull it from system

it's easy to check the same with SEF amp ...... few little changes in basic mosfet follower , one doesn't need SIT in circuit

I'm expecting Klunk!! point to be different in vivo vs. sims , but no other difference
 
Last edited:
#460
Zen Mod said:
if anything else is behaving well (as defined by model) in simulations , why that should be different?
... /QUOTE]

"Trust but Verify". I do not trust the Papa-SIT SPICE models without verification.

'In vivo' measurement of the current waveform through the Papa-SIT would be tricky without any sense resistor directly on the Papa-SIT source or drain connections.

A somewhat complicated differential measurement the output voltage V(out) and the voltage on the IRFP9240 source pin Vs(PFET) would give the following expression for I(SIT):

I(SIT) = I(Out)+Is(PFET)
= V(out)/Rload + (V(out)-Vs(PFET))/Rs

where Rs is the source resistor on the IRFP9240.

Such a waveform measurement would impractical.
Click to expand...
 
You must log in or register to reply here.
Top Bottom