Beginner symasym circuit analysis

[serj.t.s]

Hi, I am trying to understand the popular symasym 5.3 audio amplifier because I think it is a good example to start with..

I'm currently analyzing the input stage and have a few doubts, please help if you can:

Ok so this is for the INPUT STAGE:

- Calculating the current of the BJT current sink give a value of about 3mA. This current then separates in half to bias the MPSA18 BJTs, but here comes the part i don't understand, shouldn't the collectors DC voltage be set near 0v to allow maximum possible swing? With this small current the collectors sit on vdd and have no swing at all. The collector resistors should be of a higher resistance?

- The differential gain i calculated is Avd=-gm*Rc=-0.0592*(680+22*2)=-43 But in multisim(with the weird dc value) give a gain of about -14. I also note that the output is a distorted sinusoid, because of the swing difficulties?

- What is the 10 Ohm resistor for? Why doesn't it connect directly to ground?

- What are all those resistors in the BJTs bases for? Couldn't one connect the input directly to the transistors bases because is would help CMRR? By my calculations :
Left BJT:
- The 4.7uF capacitor implements a high pass filter with fc=1.41Hz
- 100pF low pass 868Khz

Right BJT:
- 470uF low pass 0.34Hz!
- 100n low pass but when in comes into consideration the 470uF capacitor is shorting it, 470uF is more like a resistor in high frequencies?
- The 22kOhm implements a feedback loop?

I hope my English is understandable enough, sorry about that.

Thanks.

[serj.t.s]

These pictures are from the symasym5 project page http://www.lf-pro.net/mbittner/Sym5_...ymasym5_3.html

[serj.t.s]

Also what is the 22Ohm series with supply resistor for?

[ttan98]

I think you should direct these questions to M. Bittner himself, he would be able to answer them promptly.

I am able to answer one question, ie the 10ohm resistor, you will note that the o/p gnd is connected directly to 0 V and the i/p gnd is via 10ohm, this because the you want to isolate the "noisy" o/p gnd from the "quiet" i/p gnd via a small resistor, anything close to 10ohm is adequate. This technique was used many years ago that I can remember, say >30yrs. I built an amplifier which used this technique, built 30 yrs ago!

[Andy L. Francis]

Let's read the primal Symasym thread regularly, and study Randy Slone's and/or Douglas Self's book(s)!

[Photonicube]

The reason the collectors of your input pair stay close to the rail voltage is because they are clamped to that voltage via the Vbe drop (the diode between the base and the emitter) of the next stages PNP transistors whos emitters are tied to the plus rail through the 68 ohm resistor. The Q2N5401 transistors get the base current sourced through them and amplify it (Ic = Ib * beta). there are many reasons for adding input resistors to transistors, for the input pair it is more than likely it was to increase the input resistance as well as making a low pass filter.

[sawreyrw]

serj,

The differential gain of a differential stage is approximately

Gain=gm*Rc

where gm=40*Ic at 25 degrees C.

So you equation is somewhat correct, but where did you get the 2*22? In addition, the loading of the second stage must be considered. The input resistance of the second stage is approximately Rin=2*Hfe/gm. This resistance must be included in the value for Rc. 1/2 of Rin must be put in parallel with each collector resistor.

I don't know what the 10 ohm resistor is used for, but it won't have much effect at audio frequencies.

The 4.7 uF cap is just for DC blocking. The 100 pF is a noise filter.

The 470 uF cap is just for DC coupling. It could be smaller, but you don't want it to effect the low frequency performance. The 100 nF would not be necessary except for high frequencies. The ESR of the 470 uF cap could be fairly large.

The 22K feedback resistor along with the 499 ohm resistor sets the gain.

Rick

[jaycee]

The 10R resistor places some isolation between power ground and signal ground, to avoid hum issues.

The 22R resistors in the supply rails form an RC filter with the 100n/100u capacitors near the input stage. This is to filter out noise for the input stage and VAS stage.

[serj.t.s]

Originally posted by Andy L. Francis
Let's read the primal Symasym thread regularly, and study Randy Slone's and/or Douglas Self's book(s)!

O consulted Douglas Self's book but he doesn't go into much circuit analysis, i think it's a little more advanced. But i will look for Randy Slone's book if i can find it thank you.

Originally posted by sawreyrw
serj,

The differential gain of a differential stage is approximately

Gain=gm*Rc

where gm=40*Ic at 25 degrees C.

So you equation is somewhat correct, but where did you get the 2*22? In addition, the loading of the second stage must be considered. The input resistance of the second stage is approximately Rin=2*Hfe/gm. This resistance must be included in the value for Rc. 1/2 of Rin must be put in parallel with each collector resistor.

I don't know what the 10 ohm resistor is used for, but it won't have much effect at audio frequencies.

The 4.7 uF cap is just for DC blocking. The 100 pF is a noise filter.

The 470 uF cap is just for DC coupling. It could be smaller, but you don't want it to effect the low frequency performance. The 100 nF would not be necessary except for high frequencies. The ESR of the 470 uF cap could be fairly large.

The 22K feedback resistor along with the 499 ohm resistor sets the gain.

Rick

Hi sawreyrw. I got the 2*22 because i think since the currents from the collectors join in the 22Ohm they see efectively double the resistance, at least this is how it works for the resistor in the emitters.

I think you can ignore the input resistance of the second stage because it must be much larger than 680Ohm.

Thanks all I'll try to wrap things up and analyse he rest of the circuit soon.

[AndrewT]

download Leach low tim.
read and re-read and if necessary read again.
Look at Pass articles. Again worth re-reading.