request for opinion about "design"

[Antey]

Hello Everybody

I would like to ask You for expressing your opinions about this so-called design It's simple amplifier, which should have 45 dB gain @ 30 Hz - 100 kHz bandwidth. There's "little" problem with characteristics - phase crosses 0 at highest frequencies (circut is unstable?). I enclose schema & SIMetrix plotted characeristics for reference.

Any comments, sugestions about possible flaws, improvements etc are highly desired and welcomed

[capacitors' values are...well ... experimantal at this stage]


Best Regards,
Peter

 

[Nigel Goodwin]

Hello Everybody

I would like to ask You for expressing your opinions about this so-called design It's simple amplifier, which should have 45 dB gain @ 30 Hz - 100 kHz bandwidth. There's "little" problem with characteristics - phase crosses 0 at highest frequencies (circut is unstable?). I enclose schema & SIMetrix plotted characeristics for reference.

Any comments, sugestions about possible flaws, improvements etc are highly desired and welcomed

[capacitors' values are...well ... experimantal at this stage]

My feelings on simulators are well known round here :lol:

But the design is VERY likely to be unstable, you don't have any bandwidth reducing components - in order to keep amplifiers stable it's usual to 'roll off' the HF response. You should also have decoupling capacitors across the HT rail.

 

[Antey]

But the design is VERY likely to be unstable, you don't have any bandwidth reducing components - in order to keep amplifiers stable it's usual to 'roll off' the HF response.

Thank You for showing interest in my problem

Another words... the Gain curve should fall dawn and the end of bandwidth,my selected bandwidth - because as of now, it actually fall down - but @ about 1 MHz .
The problem with limiting the bandwidth is beeing adressed - but as yet i can't find (or calculate ) all [h] matrix for transistors, so i am unable to perform 'not simulated' AC analysis so capacitors are just an experiment.
What bothers me most, is phase - it goes to fast to zero (an beyond..)...DCOP is wrong ?

You should also have decoupling capacitors across the HT rail.

What do you mean by 'ht rail' ? My Yet Another Lack of Skill includes also specific english terminology
Do you mean a mesh with R_f resistor, which the feedback signal goes through ?

Best Regards,
Peter

I'm aware that simulators have their limitations, but at the moment i don't have anthing better

 

[Antey]

What do you mean by 'ht rail'

OK, i got than one Mistype, right ? - (H)igh (F)requency Rail - a path that amplified signal goes through.... but where exactly ?

 

[Nigel Goodwin]

Not a mistake, I meant "HT rail", HT means High Tension, it's an old term for the power supply rail - back from the valve days :lol:

It means a capacitor (or more than one) from the +ve rail to the ground rail - the power rails need to be at the same AC potential, for almost all circuits to work.

 

[Roff]

What bothers me most, is phase - it goes to fast to zero (an beyond..)...DCOP is wrong ?

The phase should be zero in the passband. It will go to -180 (and beyond) above the passband, but this is not a problem.
I believe you are thinking of the open loop phase of a feedback circuit, where excess phase shift (180 degrees lag) with greater than open loop unity gain at that frequency will guarantee that you will have oscillations when you close the loop.

What does DCOP mean?

 

[Antey]

What does DCOP mean?

Well, according to my sources (which i don't claim to be 100% accurate )
it's an acronym for [DC] [O]perating [P]oint

Thank You for other explanations
So my passband now is about.... 50 kHz.


Best Regards,
Peter

 

[Roff]

What does DCOP mean?

Well, according to my sources (which i don't claim to be 100% accurate )
it's an acronym for [DC] [O]perating [P]oint

Thank You for other explanations
So my passband now is about.... 50 kHz.


Best Regards,
Peter

Well, the passband is traditionally indicated by the amplitude response, and the corner frequencies are usually measured at 3dB below the mid-band gain. At the corners, the phase shift is a function of how many RC networks in the circuit are contributing to the rolloff.

 

[Antey]

Not a mistake, I meant "HT rail", HT means High Tension

It means a capacitor (or more than one) from the +ve rail to the ground rail - the power rails need to be at the same AC potential, for almost all circuits to work.

I understand, that without calculations (which i'm currently tring to perform) i can not select proper capacitors' values -> contain a bandwidth in desired shape -> it's all guessing for a moment.

But maybe it's fatigue or again my lack of expirience - but why aren't C_e1 , C_2 (which also guides signal to output) enough for this purpose ?
I guess that only needed capacitors should cooperate with T2, because for T1 C_e1 does the job and signal provided by R-F and C_e2 is necessary for feedback loop to work.

Best Regards,
Peter

Thank You all for Your patience - i hope i'm not straining it.

 

[Antey]

Well, the passband is traditionally indicated by the amplitude response [..]
At the corners, the phase shift is a function of how many RC networks in the circuit are contributing to the rolloff.

It's hard to admit, but i forgot about this simple definition
Anyway - it means, that Gain curve should look like somewhat symetric "hill", while gain values at 30 Hz & 100 KHz should be 0,707 (3 dB lower) than maximum between them. So for the moment it means... that (while stability is still uncertain ) my circut/device (well...amplifier) has much bigger, and "shifted" bandwidth then desired... ?

Best Regards,
Peter

 

[Roff]

I can pretty much guarantee you that a 2-transistor amplifier with over 40dB of closed loop gain will not have a stability problem. Besides, if stability were a problem, you would see a big gain peak at the high end of the passband.

 

[audioguru]

The circuit just has the 1st transistor inside the feedback loop, the 2nd transistor is just an emitter follower without much high frequency phase shift. Therefore the 1st transistor will rolloff naturally at a frequency determined by its collector resistor and capacitance, but the feedback will extend the rolloff frequency, and will be stable because only a single pole is present.
The low frequency rolloff and phase shift is also caused by a single pole inside the feedback loop and will also not cause instability.

 

[Antey]

The circuit just has the 1st transistor inside the feedback loop

Tranistor in feedback loop ? Well.... i thought that feedback loop includes only
R_e21, R_F, and perhaps R_b1.... anyway, thanks for correcting my mistake and for other opinions


Best Regards,
Peter