Audio: Pre-amplifer 2011

when building a receiver, for fun and toy,  we often  need a  audio pre-amplifier or power amplifier.  the first of thing is design the bias for a amplifier, this might take you some time to set up a correct bias. And next time you use a different transistor, it will changed so much.  if you google a simple amplifier, use  a simple bias method,  but you don't have a same transistor, or the beta is so much different.  take you much time to bias them.

the possible solution, built a tester, make your favourite bias method in it, integrated resistor box/or POT with them,  then, you will found your bias so quickly.

or, use a DC coupled DC negative feed back amplifier, all of them had amazing part, suitable different supply voltage , don't need change any part of it.  change the transistor to any one, only very minor bias voltage different.

and NFB amplifier always have high gain or fixed gain, and good frequency response.

reference:

most of these circuit introduce by: http://www.talkingelectronics.com/te_interactive_index.html

The right picture show  the tricks to analysis the bias of such circuit.  the idea is :

* DC negative feed back,  if first transistor is much open, it's collector will drop , then make secondary transistor emitter also drop.

* So, first transistor must be fixed at a half open mode, make it's base around 0.7V [high beta, lower voltage, 9014 will around 0.55~0.65V ]

* for estimate the bias, it's enough accuracy to assume first transistor base is 0.7V [or 0.6V]. and 220K bias flow very little current, drop about 0.1~0.2V and even less, then first collector should around 1.4V~1.6V.

* the DC bias is so stable! change VCC make very little impact to the bias.

* NOTE: if transistor bias at high current , for example 2mA, the voltage drop on 220k is not omit-able,  or by another way, use low value resistor replace the 220k is a good solution.

this amplifier dos not introduce the AC feed back, just use the maximum  gain it can achieve. the output point around 0.7V, if you want bias it toward to middle of supply? use the following circuit.

I built one circuit to analysis the bias.  the left image show the estimation and result.  this amplifier add 1k to first transistor's emitter does  this impact the bias, and how much?

*1k resistor add to secondary transistor's emitter, double the bias voltage to about 1.2V, now we temporarily do not count the 1k resistor in the first transistor's emitter.  then first transistor collect is around 2.2V.

*now count the 1k resistor, voltage drop on 22k is 5V-2.2V=2.8V, so 1k resistor will be 2.8/(22) ~=0.12V, add this to our calculator result

first transistor base: 0.6+0.12~=0.72, out point 2*0.72 = 1.4V.   this estimation is actually enough for us, the real voltage read be  multimeter  is , the base, 0.85V, the out point:1.73V.  this error cause we don't count the 220k resistor voltage drop, that's will be around 0.1V~0.2V

the estimation error about 0.3V, it's truly use-able to design and investigate. when we do estimation assume the Vbe from 0.55 to 0.7V we don't need much accuracy, indeed.

bias:

* 1M/220k fix the BC547@1V

* 10K voltage limited by BC558 Vbe=0.6V, given current 0.6V/10k

*  3k3 given current: 0.4V/1k - 0.6V/10K

*  voltage drop on 3k3: (0.4/1-0.6/10)*3.3 = 1.02V

*  BC558 collector voltage:  0.4+1.02 =1.42V

real Test result: [bias use 100k:22k not 1M/220k, it's different, cause 1M can not give enough bias current ]

*BC547 base        1.08V@6V supply    1.81V@10V supply

*BC547 emitter     0.51V@6V supply    1.25V@10V supply

*BC558 collector   2V@6V supply       5V@10V supply

and simulation for 1M:220k vs. 100k:22k

Audio Pre-amplifier: NPN/PNP DC-AC NFB

@7/18 2012

8R speak booster

@7/18 2012

* bias estimation: 

1) DC negative feed back circuit, so the bias will stable somewhere

2) the 8R spk compare to 10K+2.2k load of bottom transistor is omit-able, so the base will stable on 0.55 to 0.65V

3) 4.7k boost this voltage to ~3.4V(omit the base current), make bottom transistor's collector "fix" on  ~4V

4) replace 4.7k with 1k make this circuit suitable from VCC 3V to 20V.

* bootstrap : top 22u is bootstrap, the positive feedback make so much gain

* bottom is common base, low input impedance, match to 8R spk

the following test is replace the 4k7 with 1k

bottom bc547 base:  0.65V@10Vsupply, 0.59V@3V supply

OUT:  1.5V@10V supply  1.22V@3V supply

bottom BC547 collector:  2.12@10Vsupply  1.86V@3V supply

What if 8R spk replaced with 4.7k resistor?

now the 'spk' resistance can't be omit assume voltage drop is x,so

(x+0.6)*2+0.6= [VCC-(10k/4.7)*x]     ## base current is omit, top 2.2k is removed

so x= 2.05V @Vcc=10V, test result  show the x=2.07V