# Sziklai pair design

#### emosms

Joined May 13, 2020
8
Hi, is that a Sziklai pair configuration?

I works as a low shelving filter (exaclty what I need).
How can I adjust the gain? If possible, to go bellow gain 1 for the filter attenuated frequencies.
Example range - gain 1 for the attenuated frequencies to gain 1 for the boosted frequencies.

I am also trying to set it for minimum consuption (total of 0,2 mA if possible).
On the simulator it seems it needs to be at least 0.4mA, to drive 50k load.
(3.2 v p-p, 50 hz signal)
But, I am just wiggling with the simulator, not sure what I do. Here it is the simulated circuit:

Is there any way have a headroom for more than 3,2 v p-p input signal?
- apart from lowering the total gain of the circuit

#### crutschow

Joined Mar 14, 2008
33,310
is that a Sziklai pair configuration?
Only if you removed the 2k2 resistor and 330nF capacitor in the first circuit.

What, exactly, is that circuit supposed to do?
What is the response of the filter supposed to be?

#### Ian0

Joined Aug 7, 2020
8,927
Looks like a two-transistor gain-stage to me, with a gain of 1+(2.2k/240) at low frequencies, and a roll off that is 3dB down at 220Hz and a gain of unity at high frequencies.
EDIT: Corrected gain equation

Last edited:

#### emosms

Joined May 13, 2020
8
Looks like a two-transistor gain-stage to me, with a gain of 1+(2.2k/240) at low frequencies, and a roll off that is 3dB down at 220Hz and a gain of unity at high frequencies.
EDIT: Corrected gain equation
Which 2k2, the one in parallel with the cap?
And also, which resistance forms the filter with the cap.. I guess it is a high pass filter (negative feedback)

#### emosms

Joined May 13, 2020
8
Only if you removed the 2k2 resistor and 330nF capacitor in the first circuit.

What, exactly, is that circuit supposed to do?
What is the response of the filter supposed to be?
Shelving bass boost, with gain 1 for the high frequencies.
I dont know how to set the transistors in proper mode, with a minimum possible current consumption.
(it will drive a 50 k volume pot).
Just playing with it in sumulator...

#### emosms

Joined May 13, 2020
8
When playing in simulator, seems it can work with R17 (my schematic) - the collector resitor of first transistor, 2k2, in pretty wide range..
f.ex. 22k-470k.
No change in gain or current draw.
Don't know what to put there

#### Ian0

Joined Aug 7, 2020
8,927
Depends on how much current you want to run the input transistor at.
That can be anywhere between 40uA and 5mA depending on whether you want best noise performance with a high impedance or low impedance source.
And if you don‘t care about noise performance, pick somewhere in the middle like 500uA, and choose R so that the next stage is biased up properly . I.e. if R<1.2k then there won’t be enough voltage on the base of the next transistor to turn it on.

#### crutschow

Joined Mar 14, 2008
33,310
Which 2k2, the one in parallel with the cap?
Yes.
That's the problem with schematics that have no component reference values.
You have to do a long explanation of which component you are referring to.

#### emosms

Joined May 13, 2020
8
Depends on how much current you want to run the input transistor at.
That can be anywhere between 40uA and 5mA depending on whether you want best noise performance with a high impedance or low impedance source.
And if you don‘t care about noise performance, pick somewhere in the middle like 500uA, and choose R so that the next stage is biased up properly . I.e. if R<1.2k then there won’t be enough voltage on the base of the next transistor to turn it on.
I got it how to calculate:
- gain: 1 + R19/ RV2
F.ex. 1 + 10/3,3 = 4 = +12dB (~11dB in simulator)
-3db rollof freq (I guess high pass, negative feedback) - R19 and C15
- max current draw: (Vcc/2) / ((RV2 + R19) || R Load))
F.ex. 9/2 = 4,5v/ 3k3+10k || 50k = 4,5/ 10,5 = 0,428 mA
It all matches pretty close to the simulator results.

Dont know how to calculate the base current (R17) based on the specs.
There is no base current vs anything graph.
I do care for noise

#### Attachments

• 138.1 KB Views: 3

#### Ian0

Joined Aug 7, 2020
8,927
Base current is collector current divided by beta (Hfe).
BC549 datasheet is a bit vague about its noise characteristics, only specifying a maximum of 4dB at one source impedance.
If you look at a datasheet like this one
https://www.mouser.com/datasheet/2/408/2SC3324_datasheet_en_20140301-1150780.pdf
You get a contours of noise figure graph on page 4. Find your source impedance on the y-axis ad go across until you get to the lowest noise, then down to get your collector current,

#### emosms

Joined May 13, 2020
8
Actually, the second transistor is a PNP transistor. Current flows from emiter to base and there is no base (or emitter) resistor....
So, R17 is part of the compound pair topology ??
- https://sound-au.com/articles/cmpd-vs-darl.htm

#### Attachments

• 6.3 KB Views: 5

#### Ian0

Joined Aug 7, 2020
8,927
As @crutschow said earlier, this isn't a Sziklai pair, it's just a two-transistor amplifier.
It would only be a Sziklai pair if there were no components between the two transistors.
There is no series resistor for the second transistor, because the first transistor is connected as a constant current source.
There is a shunt resistor, so that extra current that is not required by the base of Q6 has somewhere to go, and doesn't just drive Q6 into saturation.
If you want a good discussion of simple transistor amplifiers, have a look a Linsley-Hood's books.
https://www.amazon.co.uk/Linear-Ele...p-0750637463/dp/0750637463/ref=dp_ob_title_bk

#### emosms

Joined May 13, 2020
8
I will put 100k for R17, dont know how to caluclate it

#### emosms

Joined May 13, 2020
8
As @crutschow said earlier, this isn't a Sziklai pair, it's just a two-transistor amplifier.
It would only be a Sziklai pair if there were no components between the two transistors.
There is no series resistor for the second transistor, because the first transistor is connected as a constant current source.
There is a shunt resistor, so that extra current that is not required by the base of Q6 has somewhere to go, and doesn't just drive Q6 into saturation.
If you want a good discussion of simple transistor amplifiers, have a look a Linsley-Hood's books.
https://www.amazon.co.uk/Linear-Ele...p-0750637463/dp/0750637463/ref=dp_ob_title_bk
Isnt the base current of the second resistor allready limited by the first transistor emitter resistor?
For emiter I put smth between 3k3-13k3 (with a trimpot) to dial the boost gain.

So, max possible current: (9v-0.65v)/ 3k3 = 2.5 mA (base current)
100k for the shunt resistor (R17) should be 0,65v/ 100k = 6.5 microamps excess current to go away?

It does not really matter, if 2.5 mA base current is ok for the second transistor.
Mby thats why it works in simulator whatever I put (if the shunt resistor is not too low).