Hi there...
for driving pair not so usual amplifier for my open baffle speakers I need some transparent unity gain buffer stage that will supply audio signal to my amps. Buffer stage must have low output impedance in order 50…200 ohms. My preference is all discrete technology, without IC. My amplification setup have 2x24Vdc and 30Vac sources (and any derivate from this voltage is possible). So, any recomendations for that type of buffer stage?
Thank you all for helping me on this and given ideas..Dubravko from Croatia
for driving pair not so usual amplifier for my open baffle speakers I need some transparent unity gain buffer stage that will supply audio signal to my amps. Buffer stage must have low output impedance in order 50…200 ohms. My preference is all discrete technology, without IC. My amplification setup have 2x24Vdc and 30Vac sources (and any derivate from this voltage is possible). So, any recomendations for that type of buffer stage?
Thank you all for helping me on this and given ideas..Dubravko from Croatia
A Nelson Pass's B-1 (or another inspired/variation like Salas's DCB1 -dc coupled, use no capacitors in signal path.)?
Are you definitely against 'ic'? Because Lme49600 is impressive component, simple to implement which sound 'transparent' to my ears ( it doesn't have sound at all to me).
With a superegulator as psu i'm sure it would left no wishes unsatistied ( or a shunt regulator... it's true for other buffers too imo).
And for smd it still have some easy to handle size/soldering ( even with an iron and a small tip - but better use heated air station imho).
Figure 29, page 11 of TI datasheet show the typical implementation needed. You can't do simpler than that imho :
https://www.ti.com/lit/gpn/lme49600?HQS=TI-null-null-alldatasheets-df-pf-SEP-wwe
Are you definitely against 'ic'? Because Lme49600 is impressive component, simple to implement which sound 'transparent' to my ears ( it doesn't have sound at all to me).
With a superegulator as psu i'm sure it would left no wishes unsatistied ( or a shunt regulator... it's true for other buffers too imo).
And for smd it still have some easy to handle size/soldering ( even with an iron and a small tip - but better use heated air station imho).
Figure 29, page 11 of TI datasheet show the typical implementation needed. You can't do simpler than that imho :
https://www.ti.com/lit/gpn/lme49600?HQS=TI-null-null-alldatasheets-df-pf-SEP-wwe
Last edited:
Interesting idea, good to know that. I was un avare of such IC existence. Thanks for tip. My PSU is adequate (with propper regulation, filtering and decoupling).
I entirely agree on that.
Despite severe components shortage, it is available from the EU supplier: https://www.tme.eu/hr/en/details/lme49600ts_nopb/smd-operational-amplifiers/texas-instruments/
@Tombo,
You don't about regulators? 😋
Yes i was surprised to see they are in stock @ different places ( mouser as them currently in stock).
I've used them as buffers ( open loop as fig 29) and nested into global feedback loop as a 'super opamp'.
They performed great both way. Tombo how did you used them?
You don't about regulators? 😋
Yes i was surprised to see they are in stock @ different places ( mouser as them currently in stock).
I've used them as buffers ( open loop as fig 29) and nested into global feedback loop as a 'super opamp'.
They performed great both way. Tombo how did you used them?
Hah, yes, I like them seasoned with super regulators. 🙂
This is my use-case: https://www.diyaudio.com/community/...ludef-sissysit-and-similar-amplifiers.382345/
This is my use-case: https://www.diyaudio.com/community/...ludef-sissysit-and-similar-amplifiers.382345/
Very nice circuits Tombo56!
I like both your buffer and servo. 👍
I will have to dig deeper with your buffer though... i just eyballed it and devil is in the detail.
And about psu i see we are on same camp. I did some comparison ten years ago and at that time Salas shunt reg were a definitive subjective improvement over lm317/337 in the circuit i tested them.
Now i've got a Salas Bib and the LM 3X7 psu on my bench for test purpose between circuits i play with. The Bib wins always.
I implemented a superreg for my cousin and i think i prefered it to the Bib. I should do another comparison session with both and some ldovr i have used in other circuits ( tube mic with direct bias ( direct dc voltage into the cathode of common ground amplifier... lm317 is way too noisy for that kind of thing! Ldovr was surprising) and liked too.
Kmet,
Member OPC had a serie of circuits using that generation of IC like lme49600 called 'the wire'.
Worth a read:
https://www.diyaudio.com/community/threads/the-wire-all-boards-and-kits-explained-here.272621/
I like both your buffer and servo. 👍
I will have to dig deeper with your buffer though... i just eyballed it and devil is in the detail.
And about psu i see we are on same camp. I did some comparison ten years ago and at that time Salas shunt reg were a definitive subjective improvement over lm317/337 in the circuit i tested them.
Now i've got a Salas Bib and the LM 3X7 psu on my bench for test purpose between circuits i play with. The Bib wins always.
I implemented a superreg for my cousin and i think i prefered it to the Bib. I should do another comparison session with both and some ldovr i have used in other circuits ( tube mic with direct bias ( direct dc voltage into the cathode of common ground amplifier... lm317 is way too noisy for that kind of thing! Ldovr was surprising) and liked too.
Kmet,
Member OPC had a serie of circuits using that generation of IC like lme49600 called 'the wire'.
Worth a read:
https://www.diyaudio.com/community/threads/the-wire-all-boards-and-kits-explained-here.272621/
M2x input stage is great idea. If I correctly make recalculation of current on output stage, this is in magnitude of 20mA. This high enough to be in class A. What is output impedance of this stage?
Since it's a transconductance amplifier operated at unity gain, output impedance is (1 / gm). If you feel like calculating gm by hand, it may help to know that measured beta of ZTX558 is 220, measured beta of KSA1220AY is 250. Or just run a quick spice simulation and skip the hand calculation.
If you're supremely inquisitive, simulate and plot Zout for several different values of C5, to see how much of an effect C5 has (if any) upon closed loop output impedance.
If you're supremely inquisitive, simulate and plot Zout for several different values of C5, to see how much of an effect C5 has (if any) upon closed loop output impedance.