Searching the "best" CCS

[FdW]

Resistor on top or bottom does not make a difference. It is important to use a pair of transistors with a high enough Vce. Using 2SA970 and 2SC2240 will be fitting for amplifiers running on a V+ V- differential up to 120Vdc. Also the 'mid'-resistors (as proposed by me, but may be overkill [kenroken]) are at the least just as good Selecting witch resistor(s) to use is mainly a layout decision (for the PCB or other).

As I have not been able to simulate the 'thyristor' behavior (assumed to be caused by [presumably] Vce breakdown under high Dv/Dt circumstances) it is not posible to say to much about that (it has never been seen in my applications of this circuit). As (my hypothesis) Dv/Dt plays a role in the phenomenon it may be useful to check if base-stopper resistors are of any benefit.

Note: It is known that the simple 2 BJT CCS may oscillate and needs to be compensate by using a small capacitor or/and base-stopper resistor, discussions (here in the forum [I do remember that Bob Cordell attended ]) are here (at DiyAudio.com) to find (but I do not recall where/when exactly)

 

[Hearinspace]

Post #95 edited so links go to indicated posts:

Some references and observations about the 4 BJT CCS as used in the 'Nobrainer' amplifier thread.

The historical origin of the circuit is not known to me (FdW)

First showing [in the Nobrainer thread] (circuit on left side)
JG´s Nobrainer and Nobrainer Discrete
Post# 312

Using: 2SA970 and 2SC2240
JG´s Nobrainer and Nobrainer Discrete
Post# 323

JG´s Nobrainer and Nobrainer Discrete
Post# 328

Problems in paradise
JG´s Nobrainer and Nobrainer Discrete
Post# 356

Investigations
JG´s Nobrainer and Nobrainer Discrete
Post# 362 and up

Solved (circuit on right side)
JG´s Nobrainer and Nobrainer Discrete
Post# 366 The circuit shows a strange mix of transistors, this is no problem but unintended (not noted at the time)

The circuit works problem free in most, if not all, situations and configurations. Using transistors with a higher hfe will produce a better current regulation. Note that Vce(max) must at the least equal the power supply differential (from V+ to V- of the circuit, if not the circuit may have startup problems due to some form of thyristor behavior of the circuit.

Footnote
JG´s Nobrainer and Nobrainer Discrete
Post# 370
Attached Thumbnails
Searching the "best" CCS-4bjt-ccs-png

 

[FdW]

The links (in the original message) seem to work for me but if these work better for you 'make it so'

 

[RCruz]

The links (in the original message) seem to work for me but if these work better for you 'make it so'

The links in your original post did not work here....

 

[RCruz]

Just remenber that I was using your ccs in my amp and it did not work on a breadboard so I added the two 2k2 resistors and everything worked.....

The original idea for the two resistors came from me.....
https://www.diyaudio.com/forums/ana...rainer-nobrainer-discrete-19.html#post3840898

 

[FdW]

And I like[d] it

I know it's just some illogical mental forse that likes the total symmetrical schematic (maybe it's worth a resistor) if not, your right

 

[RCruz]

Nice

BDW I still use that amp now with a slightly higher voltage and new drivers.......

Nothing like your beast but it sounds really good.

 

[Hearinspace]

One question about this circuit. Does difference in Hfe (or not) between the NPN and PNP transistors have much effect on performance? Anybody looked at this?