[SOLVED] Power Amplifier Classes

[natnoraa]

Hi everyone,

Am questioning the reason behind these discrepancies and have yet to convince myself so here it is:

Linear amplifiers are classified as such according to theory:


I ran a PDK transistor and this is what I'm getting:


Question: Looking at what I'm getting, shouldn't a Class A biasing be at about 2.5V gate voltage if it should be about 50% of the peak current? However, at about 1.1V I'm picking up a Class A transient current waveform at the drain terminal. Could anyone please explain? Any help is appreciated. Thanks



F: 1GHz
Vdd = 1.2V

Natnoraa

 

[johnjoe]

Honestly, for me it doesn't look like class A, especially the waveform at the bottom, it's more class AB, maybe simulate with VCC/2 and post your result with the same simulation setup.

 

[natnoraa]

Hi johnjoe,

Thanks for your reply. By meaning of simulating with VCC/2, i suppose it's simulating Vdd to be 0.6V and everything else the same. Here is the waveform:



P.S: What's the purpose of reducing the VCC by 2 here?

Natnoraa

 

[johnjoe]

Sorry, I expressed myself not correct. You have to increase your biasing to set the operating point correctly. I don't know your schematic, but let's assume your using a emitter or source circuit with resistor in the collector/drain branch. Your dc-operating point should be at Vds=Vce=Vcc/2.

Consider the picture on page 2 of the attached pdf.

 

[natnoraa]

Hi johnjoe,

Now i understand where your vcc/2 comes from. However it's inductor bias as shown:



Natnoraa

 

[johnjoe]

Ok, than Vds=Vcc (neglecting any ohmic resistance of the coil), nevertheless you have to increase your bias. You can also draw a loadline in your output characteristic and set the dc-operating point based on your loadline.

 

[natnoraa]

Hi johnjoe,

I understand your point. Yes it's an ideal inductor still so there's 0 ohmic loss. However, my question is as per my first post. According to theory, half point towards max Ids will be my class A biasing point ie. 2.5V Vg. However, at 1.1V (or safely assume 1.2V) I'm already getting class A waveform. Why is that so?

Natnoraa

 

[johnjoe]

First of all, your first pic of your waveform is not class A, class A means a sinusoidal input signal -> sinusoidal output signal, it's not in your case, it has some distortion. Second (now, I got it with Vg), if you use Vg=1.2V and your input signal is small enough, you still have a linear characteristic. Take your output plot and draw the output signal with large input signal and your output signal with small input signal. Because the transition between AB and A is smooth.

 

[natnoraa]

Hi johnjoe,

Alright I'll look into this first. I've seen too many documents on load-line and output signal = input signal. The example if i were to increase Vg to 1.4V I will be getting it but still why it's << 2.5V that's the question. Thanks for now.

Natnoraa

 

[johnjoe]

Maybe because your input signal is very small and you're using some harmonic filter, aren't you?

 

[natnoraa]

Hi johnjoe,

You're right. My input signal should range between -20dBm to 10dBm typically, base on the IEEE standard and research standard for 60GHz. And yes, there's a harmonic filter. So small input signal could result in the deviation?

Natnoraa

- - - Updated - - -

Hi johnjoe,

It's okay, the problem is solved now. The reason is that for the 2.5V i've been looking at, it should be the AC signal + the DC signal. so input signal does matter, which you gave me the hint. not just looking at the DC signal. Thanks again.

Best Regards,
Natnoraa