Forums

RF amplifier stability question

Started by amal banerjee April 28, 2020
The standard treatment of amplifier stability is based 
on small signal S parameters, and involves conditional stability, unconditional stability, Rollet's criterion, 
etc., What about steady state amplifier operation, is
there an analytical way to estimate this ? 
On 2020-04-28 00:43, amal banerjee wrote:
> The standard treatment of amplifier stability is based > on small signal S parameters, and involves conditional stability, unconditional stability, Rollet's criterion, > etc., What about steady state amplifier operation, is > there an analytical way to estimate this ?
S parameters work for amplifiers too. Or are you asking about large-signal instability, e.g. snivets or Class C oscillation? Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com
On Tuesday, April 28, 2020 at 12:43:20 AM UTC-4, amal banerjee wrote:
> The standard treatment of amplifier stability is based > on small signal S parameters, and involves conditional stability, unconditional stability, Rollet's criterion, > etc., What about steady state amplifier operation, is > there an analytical way to estimate this ?
If you just need an amplifier go buy a stable amp from minicircuits and don't worry about it. If you are designing an amplifier where you need to know this stuff you are going to have to dive pretty deep. What is negative resistance....go ponder that for a month or two
On Tuesday, April 28, 2020 at 9:21:53 PM UTC+5:30, blo...@columbus.rr.com wrote:
> On Tuesday, April 28, 2020 at 12:43:20 AM UTC-4, amal banerjee wrote: > > The standard treatment of amplifier stability is based > > on small signal S parameters, and involves conditional stability, unconditional stability, Rollet's criterion, > > etc., What about steady state amplifier operation, is > > there an analytical way to estimate this ? > > If you just need an amplifier go buy a stable amp from minicircuits and don't worry about it. If you are designing an amplifier where you need to know this stuff you are going to have to dive pretty deep. > > What is negative resistance....go ponder that for a month or two
There is nothing profound about negative resistance. It occurs for some semiconductor devices(e.g., Gunn diode) when the forward current decreases with increasing forward voltage, and physically means that the device is pumping current into the circuit to which it is attached. I am afraid you have not answered my question-does steady state behavior of an RF amplifier need large signal S parameters ?
On Tuesday, April 28, 2020 at 8:28:34 PM UTC+5:30, Phil Hobbs wrote:
> On 2020-04-28 00:43, amal banerjee wrote: > > The standard treatment of amplifier stability is based > > on small signal S parameters, and involves conditional stability, unconditional stability, Rollet's criterion, > > etc., What about steady state amplifier operation, is > > there an analytical way to estimate this ? > > S parameters work for amplifiers too. > > Or are you asking about large-signal instability, e.g. snivets or Class > C oscillation? > > Cheers > > Phil Hobbs > > -- > Dr Philip C D Hobbs > Principal Consultant > ElectroOptical Innovations LLC / Hobbs ElectroOptics > Optics, Electro-optics, Photonics, Analog Electronics > Briarcliff Manor NY 10510 > > http://electrooptical.net > http://hobbs-eo.com
I am just asking whether analysis of steady state behavior of an RF amplifier needs large signal S parameters. The S parameters supplied by transistor manufacturers are small signal S parameters.
On Wednesday, April 29, 2020 at 6:54:40 AM UTC-4, amal banerjee wrote:
> On Tuesday, April 28, 2020 at 8:28:34 PM UTC+5:30, Phil Hobbs wrote: > > On 2020-04-28 00:43, amal banerjee wrote: > > > The standard treatment of amplifier stability is based > > > on small signal S parameters, and involves conditional stability, unconditional stability, Rollet's criterion, > > > etc., What about steady state amplifier operation, is > > > there an analytical way to estimate this ? > > > > S parameters work for amplifiers too. > > > > Or are you asking about large-signal instability, e.g. snivets or Class > > C oscillation? > > > > Cheers > > > > Phil Hobbs > > > > -- > > Dr Philip C D Hobbs > > Principal Consultant > > ElectroOptical Innovations LLC / Hobbs ElectroOptics > > Optics, Electro-optics, Photonics, Analog Electronics > > Briarcliff Manor NY 10510 > > > > http://electrooptical.net > > http://hobbs-eo.com > > I am just asking whether analysis of steady state > behavior of an RF amplifier needs large signal S parameters. The S parameters supplied by transistor > manufacturers are small signal S parameters.
I think you are likely answering your own question. You probably are not being provided the parameters you want. What frequency? What power? How much into compression? Class A or B or C or AB? Not to be a pain, but how can anyone possibly answer your question when you are so vague? My experience though is to go play with the part and figure it out on the bench unless this is some super laser kilo buck device.
On 2020-04-29 06:54, amal banerjee wrote:
> On Tuesday, April 28, 2020 at 8:28:34 PM UTC+5:30, Phil Hobbs wrote: >> On 2020-04-28 00:43, amal banerjee wrote: >>> The standard treatment of amplifier stability is based >>> on small signal S parameters, and involves conditional stability, unconditional stability, Rollet's criterion, >>> etc., What about steady state amplifier operation, is >>> there an analytical way to estimate this ? >> >> S parameters work for amplifiers too. >> >> Or are you asking about large-signal instability, e.g. snivets or Class >> C oscillation?
> I am just asking whether analysis of steady state > behavior of an RF amplifier needs large signal S parameters. The S parameters supplied by transistor > manufacturers are small signal S parameters.
In general neither is sufficient. The bias conditions change continuously in large-signal operation, so that it's quite possible for an amplifier to break into spontaneous oscillations someplace on the large-signal waveform. (That's called a snivet.) Real RF guys like Gerhard have much fancier software for that stuff. (Joerg's a real RF guy too, but AFAICT generally takes a lower-tech approach.) Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net http://hobbs-eo.com