Simple discrete unity gain buffer

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Founder of XSA-Labs
Joined 2012
Paid Member
If you are using a 220uF cap and 10k resistor per the design, it should take a at least 5 seconds to reach close to working voltage. If you have the diode in backwards, it will come up instantly and not be working as a cap multiplier.

I just ran across this neat looking 0dB buffer from Abraxalito:
https://www.diyaudio.com/forums/dig...e-rbcd-multibit-dac-design-5.html#post5495733

Certainly, a basic MOSFET source follower like what I use for my pocket or desktop Class A amp works quite well too:
xrk971 Desktop Class A (DCA) Headphone Amp

If you want a SE output stage with lower distortion use the output stage of the Aksa Lender preamp (uses a CCS):
AKSA's Lender Preamp with 40Vpp Output

The simplest one is an N-channel MOSFET with a resistive load:
616921d1495124663-xrk971-pocket-class-headamp-gb-pocket-class-schematic-v8-voltages.png
 
Hi,

a hybrid Sziklai pair (or hybrid CFP) is a way to very low distortion and considerable drive capability.
The lower half is used as constant current load (trace ´mo´ open) or as a modulated current source (trace ´mo´shorted), ideally allowing for up to twice the load current.
It can be used as SE circuit with a single supply (adding dc-blocking caps) or dc-coupled with symmetrical supply lines.
The Sziklai pair is by far more linear and can drive heavier loads than a typical single small signal transistor.
Adding JFET cascode transistors to the input JFETs will give You a Calvin-Buffer .... a search will give You a lot to read about that little guy´s qualities ;)

jauu
Calvin
 

Attachments

  • CFP Hybrid Buffer all SMD.PNG
    CFP Hybrid Buffer all SMD.PNG
    75.9 KB · Views: 539
Hi,

the simpler Hybrid CFP variant shown above is not worse compared to the cascoded Calvin Buffer but easier to build.
The cascode just shows very slightly better THD at very low load impedance levels.
Both have to be biased in class-A, because both are singleended in nature.
The output LR-network may be advantageous when driving capacitive loads like long cables, but a resistor alone (with a higher value, say 47Ohms) should suffice in most cases.

jauu
Calvin
 
Hi,

a hybrid Sziklai pair (or hybrid CFP) is a way to very low distortion and considerable drive capability.
The lower half is used as constant current load (trace ´mo´ open) or as a modulated current source (trace ´mo´shorted), ideally allowing for up to twice the load current.
It can be used as SE circuit with a single supply (adding dc-blocking caps) or dc-coupled with symmetrical supply lines.
The Sziklai pair is by far more linear and can drive heavier loads than a typical single small signal transistor.
Adding JFET cascode transistors to the input JFETs will give You a Calvin-Buffer .... a search will give You a lot to read about that little guy´s qualities ;)

jauu
Calvin

Calvin, Have you tried pch on input and of course nch on #2?
I found it sounded better.
 
Hi,

You could of course use a Darlington topology instead of single transistors to achieve considerably higher power capabilities.
I´d suggest a CFP (Sziklai) topology as it dosesn´t change the required bias voltages, is easy to stabilize thermally, it´s very linear, and allows for larger voltage swings.

jauu
Calvin

ps. btw. the parts number count of the buffer in #124 is lower than those in #128 and #129 ;-)
 
Last edited:
I did an ac analysis for the elector buffer in LTSpice and there is a huge peak in the transfer function.how can this be cured?
 

Attachments

  • cascode-bc550-buffer.png
    cascode-bc550-buffer.png
    5.9 KB · Views: 455
  • elektor-discrete-buffer.gif
    elektor-discrete-buffer.gif
    11 KB · Views: 466
  • Untitled.jpg
    Untitled.jpg
    336.9 KB · Views: 446
  • Electorbuffer.asc
    3.9 KB · Views: 38
Last edited:
how strong is this buffer btw? im wanting to see if i can drive my sennheiser 600 with it. its 300ohm impedance.

Theoretically speaking, the distortion will climb as the buffer finds it has to drive a lower impedance, even 300 Ohm is more trouble than if using the buffer as a pre-amp driving the input stage of a power amp. But doesn't mean that it won't sound very good. So did you try this yet ?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.