Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Amplifier oscillates at high freq. but the gain is negative

Status
Not open for further replies.

eladg

Full Member level 3
Full Member level 3
Joined
Jan 13, 2002
Messages
185
Helped
7
Reputation
12
Reaction score
3
Trophy points
1,298
Activity points
1,238
amp getting osi

i builed an amplifiar.
it osilated at high freq, but in the simultion the gain at this freq was negetive.
how can it happend ?
 

1. Some part of the circuit were not correctly modeled.
2. If all parts were correctly modeled, then some stray (capacitive or inductive) coupling between input and output of amplifier was not taken into account.
g579
 

g579 said:
1. Some part of the circuit were not correctly modeled.
2. If all parts were correctly modeled, then some stray (capacitive or inductive) coupling between input and output of amplifier was not taken into account.
g579

what can i do ?
 

ground plane

If you used the part models that came with your simulator this is less of a potential source of the problem than layouts issues.

Your circuit should be made on a card with a ground plane. The component leads should be kept short. The amplifier should be laid out in a straight line across the board instead of looping around. (This makes the distance from the input to the output large) Another potential problem is coupling through the power supply. Make sure that there are bypass capacitors on each power connection to any part of the circuit.
 

Re: ground plane

flatulent said:
If you used the part models that came with your simulator this is less of a potential source of the problem than layouts issues.

Your circuit should be made on a card with a ground plane. The component leads should be kept short. The amplifier should be laid out in a straight line across the board instead of looping around. (This makes the distance from the input to the output large) Another potential problem is coupling through the power supply. Make sure that there are bypass capacitors on each power connection to any part of the circuit.

well i am workink on alumina 15 mill, the amp is die the size of the amp is ablou 0.3 inch on 0.3 inch. and it has good ground
the bais is throw lamda/4 and cap.

??? at this case what can i do ?
 

circuit

Post your circiut diagram and simulation results. What frequncy is it oscillating? Does the frequency change with supply voltage?
 

Re: circuit

flatulent said:
Post your circiut diagram and simulation results. What frequncy is it oscillating? Does the frequency change with supply voltage?

well there is an isoltion from dc to rf.

the imtresting is that with out dc the r.l. is beter then -12 dbr, when i put dc the rl was -1 dbr, and i didnt see any spur at spectrum analyzer, execpt my signal. there is 3 stage at the amp when i disconected the dc from the scend one the rl was -8 dbr.
i tryed to osilate one stage with open stub and slot line (tuner) and i codent (it is uncondition stable)
at -40 deg the unit didnt osilate execp my input signal wicth came back from the oput put to the input (rl -2) when i turn off my signal, the output of my unit is quit.

????
 

Hi eladg,

as flatulent wrote, I think it would be a good idea if you upload a
schematic, or (better) the layout of your amp. At which frequency is
your amp designed to work? Often the problem is the grounding of the
amp. For higher frequencies even short vias (15mil) can have a high
inductance which can cause oscillations.
An amp is changing it's S11 when powered or not. Is your input-matching
network designed for the correct bias?

Bye
 

mr_ghz said:
Hi eladg,

as flatulent wrote, I think it would be a good idea if you upload a
schematic, or (better) the layout of your amp. At which frequency is
your amp designed to work? Often the problem is the grounding of the
amp. For higher frequencies even short vias (15mil) can have a high
inductance which can cause oscillations.
An amp is changing it's S11 when powered or not. Is your input-matching
network designed for the correct bias?

Bye

i cant upload the layout or schem.
it is design for 9-10 ghz. the osi is the same freq as i put inside (that the resen of s11). when i stop dc the s11 is good , it balanced amp
ne321000, there is feedback some how i put on the cover something that stop rf to go ( i dont know how it call in english lung') and it didnt help me
i will try to feed them from diffrent power supply and will see what happend

thank alot
 

injection lock

It looks like a case of injection lock oscillation if it is the same frequency as your drive signal. It also looks lie the parasitics are what are causing the problems. Try laying a piece of lossy ferrite over the circuit. Also try lowering the test signal level to see what results.
 

Re: injection lock

flatulent said:
It looks like a case of injection lock oscillation if it is the same frequency as your drive signal. It also looks lie the parasitics are what are causing the problems. Try laying a piece of lossy ferrite over the circuit. Also try lowering the test signal level to see what results.

i try to put alot of piece of lossy ferrite and it didnt help me.
i get into the amp at -30 dBm under it there is noise at the mesure
also i tried to put 50 ohem at the enter of the circuit and cool it to -35 deg and it didnt osi, there is a feedback some how
1) throw dc
2) throw case

i will try to work with 2 difrent power supplies
and if it will not help me i will put more pice of metel between the stage
and if it will not help me i will pray to g-d

any other idea ?
 

prayer

Prayer is best done at the design stage and not at the test stage.

Try cutting out all but one stage and see if that works properly.
 

Re: prayer

flatulent said:
Prayer is best done at the design stage and not at the test stage.

Try cutting out all but one stage and see if that works properly.

well i think that the best is long pray at the design and shorted at the test :wink:

i put one stage and it didnt osilate at all, i tryed to osilated bye tuner and i coudent , so one stage is uncondition stable.
 

try each stage

Do this test for each stage at a time. If each works it is either a feedback from common power supply or from inductive coupling between stages. It could also be output impedance of a stage being different from 50 ohms and that the source impedance on the next stage causing it to be unstable. Try varying the source impedance on the test of individual stages.
 

Hi eladg,

Problems are here to be solved... So one more question: do you have a
matching network on the gate and drain or do you feed the FET
directly with 50Ohm? Again: How is the source connected to your gnd-
plane? A good way is to use 4 vias at each source-pin (= 8 vias). These
vias has to be as near as possible (20 mils).

Bye
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top