How to stablize this amplifier ? :(

Hello all,
I designed this amplifier, and when i simulated for stability, i found it much lower than unity :(

What's the problem that make it unstable?
And how can i stablize it?

Thanks,
Ahmad,

A few suggestions:

Put a small resistor in series with the emitter resistor bypass capacitor.

Put a small resistor in series with the transistor base.

Put a series RLC network from collector to base.

Some advices:

You should decouple Bias circuit of amplifier for High frequency. Then use DC feed/inductor series/capacitor to ground.


You have to take in account quality factor&ressonance frequency to choice components (in amplifier and DC feed) at work frequency.


And there is another problem this caracteristics influence stability out of work frequency.

I added a series resistance to the base and tuned it till became 100 Ohms, and k became 1.35 (Stable)
but i want to understand the theory of it, can someone explains it?

Thanks,
Ahmad,

This base resistor method is very common and Agilent shows it in some of their documents. One result of instability is negative input resistance. Putting a positive resistance in series cancels this out. The drawback of this is increased noise.

The advantage of the feedback method (collector to base or emitter resistor) is that it also makes the gain more constant with frequency.

I tried to put series RLC from base to collector, but gain degraded too much!
what are typical valuse of that RLC?

In the series RLC, the C is just a DC block. The R controls the gain and the L is to get the R out of the circuit at higher frequencies.

The procedure is to adjust R to get the gain you want at lower frequencies and then adjust L to get stability at higher frequencies.

There are programs that will make this easier. Ad Lab plus is a cheap set of programs. The two port program will draw the stability circles that are immediately updated each time you change R L or C.

Thanks alot flatulent for these valuable help
You really helped me, but i want to ask about something you said, 'stability circles', i heard about them before here, what are them? and what's their importance? How to draw them by ADS?

Regards,
Ahmad,

Back in the ancient past when engineers used mechanical analog computers (slide rules) someone discovered that you could predict what loads and sources on a two port network would produce oscillations or not. The boundary between the stable and unstable source or load impedance values was a circle on the smith chart. The center and radius of the circle were found algebraically from the four two port parameters. So if the circle was contained in the smith chart or intersected with it, you had to make sure the load or source network never had the impedance values that would cause oscillation.

I am not familiar with ADS enough to tell you how to get these diagrams.

That's great indead!
But another problem arised after stablizing the amplifier by adding series resistance to the base of the BJT, that's minimum NF got increased from 0.5dB to 3dB..how to maintain noise as small as possible even with stablization exist?!

this is the new schematic with K factor=1.37

Thanks alot
Ahmad

Try the feedback method next. Another idea is to do the stability circles on the first amplifier and see if the source and load are in the stable region. The K method gives the worst case conditions. The circles must be drawn at all frequencies, not just your operating range.

Another option is to use a JFET. They are a good low noise choice from DC to 400 MHz or so.

Another option is to use a different BJT or bias point. Since you are only interested in below 200 MHz there is no need for transistors which have high gain at higher frequencies.

First u need to ensure stability, i see u have some matching at the input prolly for minimun NF, but u need first to tackle stability then design for min NF. As others said Feedback will do the trick for you. ofcourse depents how much gain you want your amplifier to have cause feedback will reduce the gain of your amplifier, also will afect noise figure. After u aply the feedback and u have a stable circuit , match it for NF. Also dont forget that your Bias resistors affect your design as well, do u need the potential divider to present an impedance on your input? maybe put an inductor between them and the base same applies for your collector and output. Also some decoupling on your dc paths needs to placed.

I will give u also the basic equations how feedback is related to your gain. For more details reading Gonzalez will help much.

RF=Zo(1-√S21)

RE=Zo2/RF -1/gm






Under Simulation-S-Parameters on ADS u will find different boxes with measurement expresions , there are the Source and load stability circles as well, named as LSTBcir and SSTBcir (show a small smitch chart picture on them), place them on your schematic, simulate and plot them seperatly on a smitch chart, after you plot them edit the smith chart and change its radius to be 1, cause ADS automatically feets the radius to show u the whole circles. If S11 and S22 are negative dB or less than 1 in magniture, then where the circle covers the smith chart is the unstable , and outside the circle is stable region.