A little info on current feedback bias for transistors (Cream Pie)

[Gilles C]

While preparing to work on the my Cream Pie circuit that is not working the way I expected, I found this info on that type of circuit.

http://www.interq.or.jp/japan/se-inoue/e_dance26.htm

I thought it would be interesting for some of you.

The original thread from Dragonfly is here

http://www.diystompboxes.com/smfforum/index.php?topic=50747.0

Gilles

[Gus]

look close
not the same kind of bias
Makes a difference

[Sir H C]

The cream pie looks to use the negative feedback of the base/collector with a single resistor for the biasing.  This technique is seen a lot, you can get pretty high gain.  The base-collector resistor value is divided by the gain much like a miller cap on the same node and so can affect the gain more than it would initially look but since the emitter is grounded you don't get any loss of gain on that side or any issues with emitter degen.

The link you show is an odd biasing, if there were no emitter resistor it would be one of the oldest biasing schemes, one that really sucks as there is no feedback and it can vary wildly for hfe and Vbe variances.  The emitter resistor does the regulation of the biasing but that is not that good as you get the gain loss associated with an emitter resistor, and for the addition of one more base resistor you can get a great biasing scheme.

[Gilles C]

You're right guys 

I also don't remember seeing a single resistor to Vcc like in the link used somewhere. That must be why I didn't notice the difference. At first sight, I was sure it was the same.

The reason why it's not good is explained there.

http://www.zen22142.zen.co.uk/Design/bjtbias.htm

I was trying to know the circuit a bit more...

Gilles