What the documentation actually says is;
So perhaps a good compromise?
No problem, happy to help...
I couldn't see one either but I guess this will transpire in his thread, along with solving the puzzle of suitable devices.
Last edited:
The SEN/CEN IV requires an octet of matched 2SK170V or 2SK170V/2SK74V.
I'll let you find out how much they cost these days.
(Yes, you can replace 1x 2SK170V by 4x matched 2SK209GR, or 2x matched 2SK2145GR.
Does not make it any simpler.)
On top of that, you will need 4x independent floating power supply.
That in turn means 4 independent transformers or batteries.
And if you are not careful, or do not know how, you will end up with oscillations.
So for people those who are not totally skilled in electronics, the current mirror IV is a better choice.
Don't take my word; read Pedja's excellent article :
https://www.diyaudio.com/archive/bl...-out-production-pedja-rogic-ddnf-iv-stage.pdf
Cheers,
Patrick
I'll let you find out how much they cost these days.
(Yes, you can replace 1x 2SK170V by 4x matched 2SK209GR, or 2x matched 2SK2145GR.
Does not make it any simpler.)
On top of that, you will need 4x independent floating power supply.
That in turn means 4 independent transformers or batteries.
And if you are not careful, or do not know how, you will end up with oscillations.
So for people those who are not totally skilled in electronics, the current mirror IV is a better choice.
Don't take my word; read Pedja's excellent article :
https://www.diyaudio.com/archive/bl...-out-production-pedja-rogic-ddnf-iv-stage.pdf
Cheers,
Patrick
Before you jump to any PCBs, you can easily build a prototype using through hole components on a Vero board.
https://www.diyaudio.com/forums/dig...mirror-iv-converter-la-ad844.html#post6342006
BC327/337 is fine for a start.
When you are convinced this is the right solution for you, then you can do it again with selected devices.
Patrick
I personnaly use the bc 550/560c in the Pedja arrengment and likes it... I'm not against a little of H3 because the good H2. My ears like it, the power supply and decoupling caps are more changing the sound than H2/H3 level. (ah Pedja's discrete power supply : great).
The 560 from Faitchild are hard to source since it is over and I don't know about the quality of the new smd sibblings ???
The 560 from Faitchild are hard to source since it is over and I don't know about the quality of the new smd sibblings ???
Well if you use balanced, then you cancel H2.
So I also look for low H3, assuming my matching can cancel H2 by 10~20dB.
If you do it right (good thermal tracking, etc.) you do not need the output cap as in Pedja's.
Also it is easy enough to add a output follower that adds very little to the distortion.
But of course all a matter of taste.
Both circuits (mine and Pedja's) have poor PSRR. Which is why he uses output cap.
So a stable supply is a must for DC coupling.
I do not use discrete here, because they are not necessarily stable over time.
LT3045/3094 is a better choice here.
Patrick
PS : Of course I (and I am sure also Pedja) know how to do this with good PSRR.
Just adds too much complication.
So I also look for low H3, assuming my matching can cancel H2 by 10~20dB.
If you do it right (good thermal tracking, etc.) you do not need the output cap as in Pedja's.
Also it is easy enough to add a output follower that adds very little to the distortion.
But of course all a matter of taste.
Both circuits (mine and Pedja's) have poor PSRR. Which is why he uses output cap.
So a stable supply is a must for DC coupling.
I do not use discrete here, because they are not necessarily stable over time.
LT3045/3094 is a better choice here.
Patrick
PS : Of course I (and I am sure also Pedja) know how to do this with good PSRR.
Just adds too much complication.
I like the darlington reg in the emitter follower conf without local feedback at its outputt, it's subjectivly sounding fast to me and let me focus to snake oil caps after the diodes bridge. It's enough for my ears with such pcm chips and low speeds. Indeed it works better for me with C-Core traffos and the RC Pedja is using as filter with the main smoothing cap. It's subjectivly allows according the circuit layout and grounding to work with very low value after the darlington and just decouple load with small values if the ears don't see what you can look at scope... But must say Pedja is doing very good layouts and pcbs, despite looking simple it's certainly not and full of experience... very good designer !
@pistollero
try no cap
... 1000pF still be in the limit for any VFA opamp, but can cause an oscillation for a CFA opamp..
the higher the value, the more audibly affected the sound (by a capacitor type)
anyway, read article:
transimpedance amplifier - opamp I/V
try no cap
... 1000pF still be in the limit for any VFA opamp, but can cause an oscillation for a CFA opamp..
the higher the value, the more audibly affected the sound (by a capacitor type)
anyway, read article:
transimpedance amplifier - opamp I/V
@diyiggy
1nF could be the maximum, but there will be an audible sound effect caused by this high capacitor value, it is often negative - colored by the diaelectric and similar things
... go with the lowest value as possible, 47p is stable and the sound coloration is low ... and best of all, if the capacitor can be omitted
... did you even try THS3121? (or a near family)
1nF could be the maximum, but there will be an audible sound effect caused by this high capacitor value, it is often negative - colored by the diaelectric and similar things
... go with the lowest value as possible, 47p is stable and the sound coloration is low ... and best of all, if the capacitor can be omitted
... did you even try THS3121? (or a near family)
yes yes yes, I was not talking about this capacitor, my mistake ! As said i didn't purchase the 47 pF at all for the feedback of the oaps...
No never tried that oap !
The only oap I have in my system is the opa 861, the pre stage is 100% discrete in my pre and the amp...
On the chief menu we have here the 861 again, the opa1612, the opa1656, the ada I advised and the opa 810 ! Let see what happen but because the reg stage it will take 2 months imo before I can edit about the ones I prefer in my whole setup ! many I/V resistor to try as well and half a ton of caps ! That's you have to compete with the Painkiller Ad1862 discrete and the AYA II tda1541A dac the last being still my Ref... so I have to make some great effort to try to optomize while staying in the spirit of the board, KISS but good parts well chosed !
No never tried that oap !
The only oap I have in my system is the opa 861, the pre stage is 100% discrete in my pre and the amp...
On the chief menu we have here the 861 again, the opa1612, the opa1656, the ada I advised and the opa 810 ! Let see what happen but because the reg stage it will take 2 months imo before I can edit about the ones I prefer in my whole setup ! many I/V resistor to try as well and half a ton of caps ! That's you have to compete with the Painkiller Ad1862 discrete and the AYA II tda1541A dac the last being still my Ref... so I have to make some great effort to try to optomize while staying in the spirit of the board, KISS but good parts well chosed !