Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronic 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.

how to calculate the phase margin

Status
Not open for further replies.

michaelhust

Junior Member level 1
Joined
Sep 30, 2011
Messages
15
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,404
hello,everyone
the plot is the open loop gain and phase versus frequency and the close loop gain is 18.however, i have many confusions about it.
1:what is the phase margin when the feedback factor is 1/18?
2:Does the overshoot cause close loop instability?
3:Is it common to see the phase never achieve 180 degree?
thank you
clip_image002.jpg
 

LvW

Advanced Member level 5
Joined
May 7, 2008
Messages
5,841
Helped
1,742
Reputation
3,488
Reaction score
1,342
Trophy points
1,393
Location
Germany
Activity points
39,300
Hi Michael,

it would be interesting to see the corresponding circuit.
Question: Does the diagram really show the loop gain (the gain of the open loop)?
I doubt because the phase response indicates instability of the simulated circuit (phase jump with positive slope).
 

michaelhust

Junior Member level 1
Joined
Sep 30, 2011
Messages
15
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,404
thanks very much for your reply.
that reminds me I may have the wrong test bench.
I will check the simulation and reply as soon as possible.
 

michaelhust

Junior Member level 1
Joined
Sep 30, 2011
Messages
15
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,404
hello all,i have just got back from the national day.
i tried other kinds of test bench, however the simulation kept unchanged.
here i post the main opamp circuit and hope someone can give me some suggestion.
thanks

View attachment ???.bmp
 

LvW

Advanced Member level 5
Joined
May 7, 2008
Messages
5,841
Helped
1,742
Reputation
3,488
Reaction score
1,342
Trophy points
1,393
Location
Germany
Activity points
39,300
Hi Michael, as mentioned in posting #2 the phase response indicates already instability of the simulated circuit (that means: not just after closing the loop).
 

FvM

Super Moderator
Staff member
Joined
Jan 22, 2008
Messages
48,170
Helped
14,197
Reputation
28,653
Reaction score
12,887
Trophy points
1,393
Location
Bochum, Germany
Activity points
279,108
I'm not sure, if it's actually unstable. The rising phase may be due to feedforward effects as well. But the gain peak shows, that the amplifiers internal compensation is at least near to instability. I'm not familiar with the compensation schemes for telescopic amplifiers, but guess series resistors to Cc can defuse the situation.
 

biff44

Advanced Member level 5
Joined
Dec 24, 2004
Messages
4,872
Helped
1,357
Reputation
2,710
Reaction score
1,038
Trophy points
1,393
Location
New England, USA
Activity points
36,704
The peak in gain near 0dB shows that you need to shift one of your zeros a little lower in frequency. The peaking will cause ringing for any transient waveform, which may be undesireable.
 

michaelhust

Junior Member level 1
Joined
Sep 30, 2011
Messages
15
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,404
hi, i don't understand why the circuit could be unstable when the phase jump is positive.
Doesn't instability always happen in the close loop circuit?

---------- Post added at 10:23 ---------- Previous post was at 10:13 ----------

there are two RHP plural poles in the DM transfer function.
I think it is the main reason of the positive phase jump.

---------- Post added at 10:33 ---------- Previous post was at 10:23 ----------

there are two RHP plural poles in the open loop transfer function.
i think it's the main reason of the positive phase jump.
poles ( hertz)
real imag
-801.599 0.
47.0743x 303.468x
47.0743x -303.468x
-530.111x 0.
-3.84233g 0.

zeros ( hertz)
real imag
-255.267x -331.115x
-255.267x 331.115x
360.289x 0.
27.5830g 0.
this is poles and zeroes position when the Cc is 10pf.
all the poles and zeroes that could be canceled have been neglected.
however, it's very confused the RHP poles turn to be LHP poles when the Cc is 1pf.
 
Last edited:

FvM

Super Moderator
Staff member
Joined
Jan 22, 2008
Messages
48,170
Helped
14,197
Reputation
28,653
Reaction score
12,887
Trophy points
1,393
Location
Bochum, Germany
Activity points
279,108
Doesn't instability always happen in the close loop circuit?
The internal feedback forms also a closed loop, and is well able to become unstable in an usuitable design. The open loop characteristic suggests that it's at least near to instability.
 

michaelhust

Junior Member level 1
Joined
Sep 30, 2011
Messages
15
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,404
I never knew that.
If there is a internal feedback, where is the close loop?
Does the compensation capacitor form a close loop?
 

FvM

Super Moderator
Staff member
Joined
Jan 22, 2008
Messages
48,170
Helped
14,197
Reputation
28,653
Reaction score
12,887
Trophy points
1,393
Location
Bochum, Germany
Activity points
279,108
It obviously forms a feedback structure. A risk of instability is particularly brought up, if the compensation crosses multiple stages, involving additional poles. This is the case in your design.
 

LvW

Advanced Member level 5
Joined
May 7, 2008
Messages
5,841
Helped
1,742
Reputation
3,488
Reaction score
1,342
Trophy points
1,393
Location
Germany
Activity points
39,300
Michael, two RHP poles always indicate instability. Often there are internal hidden loops that cannot be identified by visual inspection.
(Example: A simple common collector stage can be unstable due to a feedback lop that is not visible as such). As a result, there is a peak in the magnitude response that is connected with a sharp phase increase. This phase jump is caused by an RHP pole pair. It can be justified by a method that uses some geometrical rules to construct the magnitude and phase response from the pole location. This procedure is described in several textbooks on control theory.
 

michaelhust

Junior Member level 1
Joined
Sep 30, 2011
Messages
15
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,404
thanks,maybe i should try to cancel the intermediate voltage shift stage and find what will happen.

---------- Post added at 17:18 ---------- Previous post was at 17:08 ----------

It seems I have a lot of complicated work to do.
I have learned so much from all of you.
thanks very much.
 

michaelhust

Junior Member level 1
Joined
Sep 30, 2011
Messages
15
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,404
hello,everyone. I have figured this problem out recently after acquiring the basic knowledge of control theory. I used two port method to compute the loop gain transfer function and find there are four low frequency poles and one zero locating at origin due to feedforward path and feedback path respectively. Utilizing root locus method , I found that one pole goes to origin, one pole goes to infinite, and two poles bend to RHP after the interval loop is closed. So the pole generated by intermediate stage makes the circuit unstable. After I cancelled the intermediate stage,the circuit just behave like regular Ahuja' compensation way.
So thank you all for your suggestions.
 
Last edited:

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top