Hi!
Maybe its not that important, but I would like to hear some opinions on what current sources to use for the various amp stages.
There are many possibilities: One transistor with various reference voltage sources (Diodes, LED, TL431, ...), two transistor, simple FET CCS. I have found some discussions what voltage reference to use in one transistor CCS, but I did not find any "final" agreement on whats best. Two transistor CCS seem to provide a more stable current - what I think would be desirable because I plan to use a unregulated PSU. But apparently this type is rarely used (apart from the examples in Slones book). As far as I can remeber somebody on the forum stated that they would generate higher noise because of the low VBE reference voltage. Further, Slone sais in his book that they should not be used with low Re values, so they do not seem to be useful for higher currents. FET CCS seem to be somewhat unpredictible because of device tolerances (what I have experienced myself already). Then there is further the possibility of cascoding...
Many possibilities and opinions... but what to use now?
Thanks,
stegmaie
Maybe its not that important, but I would like to hear some opinions on what current sources to use for the various amp stages.
There are many possibilities: One transistor with various reference voltage sources (Diodes, LED, TL431, ...), two transistor, simple FET CCS. I have found some discussions what voltage reference to use in one transistor CCS, but I did not find any "final" agreement on whats best. Two transistor CCS seem to provide a more stable current - what I think would be desirable because I plan to use a unregulated PSU. But apparently this type is rarely used (apart from the examples in Slones book). As far as I can remeber somebody on the forum stated that they would generate higher noise because of the low VBE reference voltage. Further, Slone sais in his book that they should not be used with low Re values, so they do not seem to be useful for higher currents. FET CCS seem to be somewhat unpredictible because of device tolerances (what I have experienced myself already). Then there is further the possibility of cascoding...
Many possibilities and opinions... but what to use now?
Thanks,
stegmaie
Some humble opinions of an armchair designer:
For power current sources (in SE class A), I'd prefer the "two transistor" current source, as it will ensure that the standing current drops on incresing temperature, which would be a desirable effect:
http://www.linearaudio.de/scratch/IGBT-3.pdf
The LED has a lot of fans for simplicity I assume.
Elegant, but others will say baroque, is defining your currents completely with mirrors and a single current source.
As the master current source you can use a JFET at the zero tempco point, or a two terminal JFET current source, if you want something more expensive:
http://www.linearaudio.de/scratch/folded-cascode-sym.pdf
Regards,
Peter Jacobi
For power current sources (in SE class A), I'd prefer the "two transistor" current source, as it will ensure that the standing current drops on incresing temperature, which would be a desirable effect:
http://www.linearaudio.de/scratch/IGBT-3.pdf
The LED has a lot of fans for simplicity I assume.
Elegant, but others will say baroque, is defining your currents completely with mirrors and a single current source.
As the master current source you can use a JFET at the zero tempco point, or a two terminal JFET current source, if you want something more expensive:
http://www.linearaudio.de/scratch/folded-cascode-sym.pdf
Regards,
Peter Jacobi
- Status
- Not open for further replies.