home.t.-online.de/home/weberconnect/adlab2.htm

hi

i am designing a class A power amplifier
my question is

the K factor for the whole amplifier is less that 1 in some bands

while S11 and s22 is less than 1 in all the band
should i care about the K factor
also why should i care about low frequency stability
should i care about the K factor

thanx

E

The K factor is part of unconditional stability. If your load is well behaved, you have no problems. You can examine the loads and sources which cause unstability with the vierpol.exe program from adlab plus at http://home.t-online.de/home/weberconnect/adlab2.htm for about 20 free uses.

i ve modified the msg as the previous one was wrong. was confused myself, haven't build a amplifier for many years. anyway, just a tips for you, it's much easier to confirm ur amplifier stability by drawing the source/load termination together with the stability circle (for each frequency), and to make sure that the source/load stay as far as possible from the unstable region. sometimes, u get S11/S22 less than 1, although it is stable, but it is not robust enough. because in real life, transistor input/output impedance varies with temperature and power amplifier works in large signal condition, meaning, S-parameter varies with ur driving signal.


cray

Kaif al-7al.
better to use mu-factor instead of K-factor, b/c K-factor is necessary condition but not sufficient (you MAY need to calculate another variable called 'delta' in conjunction with K-factor).

Mu-factor is the modern way of doing stability analysis, b/c it is sufficient to have only one variable (that is: Mu), and also unlike K-factor, it gives you 'degree of stability'; that is to say if mu is close to one then it is about to oscillate. You cannot determine degree of stability using K-factor only.

Cheers...

i think s11 and s22 less than 1 mean that my input and output impedance are not negative so my whole circuit is stable

The values for s11 and s22 must be one port measurements or calculations for the entire circuit including the source or load hanging on the other end. You must examine them over all frequencies.

that is wht iam saying my S11 and S22 for the whole circuit are less than 1 ,does this make my circuit satble for 50 ohm system

thanx

Yes. But make sure to examine them over all frequencies, especially if you have LC or transmission line networks hung on each end. You also have to use the actual source and load circuits during the measurements. Be aware that if you use a transmission line to connect the load or source to the circuit, the line length will change the load or source impedance. If you use another length of line the amplifier may not be stable.

no one answered me why i should check stability at low frequency for example at 100 MHz while my operation frequency is 2 GHz

thanx

The amplifier will oscillate at any frequency where the input or output impedance has a negative real part.

well, if your amplifier oscillates at 100MHz, then:
1- It will modulate and mix with your carrier which is at 2GHz, so it will corrupt your modulation spectrum of your real signal.
2- Oscillations are totally unpredictable, and the amplifier can draw a lot of current due to oscillations, thus it may cause the amplifier to blow after some time.
3- most probably this oscialltion will affect other parts of the circuit board. Example: 100MHz oscillations may affect your DC regulators, thus it may change the DC levels of the entire board in a very unpredicatble way, and it may get worse over time.
4- Your product could radiate this 100MHz outside the board, and it may affect other stuff, and you can easily fail RF-interference specifications.

In short: unstable amplifiers are totally unpredictable, so you never never want to pass-on an unstable design... It mixes with other frequencies ..etc, and other circuits, and it radiates. It is a mess. :)
In fact, stable but almost near oscillatory designs are not good either, because it can run into oscilltions if temp changes, or humidity, or just by aging.

(PS: I replied to the PM, but maybe it did not reach you yet.)

Cheers....

for some reason, my PM reply does not want to leave my outbox, though everything si emplty (inbox, sendbox, savebox)...

The mail will leave your out box when the receiver of it reads it and it will then go to your sent box.