Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
Questions about phase in .AC LTspice Analysis
|
hoa van nguyen
Hello, I have 2 stages Cascode Diff. Amp (Differential Inputs and Outputs) and .AC analysis in LT spice simulations give me different results. All transistors are biased in saturation for high Gain. Vop & Vom are ouputs plus/minus; Vop-Vom = Diff. output Vop is complement of Vom. 1) 2stages_2CascodeDiff_ACsim_test1.asc: schematic?? ??? 2stages_2CascodeDiff_ACsim_test1.jpg : plot ??? all current sources (M9, M_cmfb1, M27 & Mcmfb2) are biased to Vbiasn=0.409V. ??? Phase of Vom starts at DC (1Hz) on +180 degree and goes to + 90 degree at 0dB. ??? Phase of Vop starts at DC (1Hz) on???? 0 degree and goes to - 90 degree at 0dB. ??? Phase of (Vop-Vom) starts at DC (1Hz) on???? 0 degree and goes to - 90 degree at 0dB. ??? Vop & (Vop-Vom) have good phase margin = |-180degree
- (-90degree)| = 90 degree ??? Questions: ??? ??? ??? ??? ??? a) Why Vop and Vom have different phases? ??? ??? ??? ??? ??? b) How
does LTspice calculate the phase of (Vop -Vom)? ??? ??? ??? ??? ??? c) How can I calculate phase margin for
Vom? 2) 2stages_2CascodeDiff_ACsim_test2.asc: schematic?? ??? 2stages_2CascodeDiff_ACsim_test2.jpg : plot ??? all current sources (M9 & M27) are biased to Vbiasn=0.409V. ??? M_cmfb1 is biased to 0.409001V ( ~0.409V). ??? Mcmfb2 is biased to 0.409002V ( ~0.409V). ??? Phase of Vom starts at DC (1Hz) on??? 0 degree and goes to -224 degree at 0dB. ??? Phase of Vop starts at DC (1Hz) on???? 0 degree and goes to - 224 degree at 0dB. ??? Phase of (Vop-Vom) starts at DC (1Hz) on???? 0 degree and goes to - 90 degree at 0dB. ??? (Vop-Vom) has good phase margin = |-180degree - (-90degree)| = 90 degree ??? ??? Questions: with 1uV different in bias of? M_cmfb1 & Mcmfb2 LTspice give dramatically ??? ??? ??? ??? ??? ??? negative phase ??? ??? ??? ??? ??? a) Why Vop and Vom have same phase now? ??? ??? ??? ??? ??? b) How does LTspice calculate the phase of (Vop
-Vom)? ??? ??? ??? ??? ??? c) How can I calculate phase margin for Vom & Vop
now? ??????????????????? d) How is the phase margin calculated When the
phase ??????????????????????? is between -180 degree and -360
degree? 3) Both xxxx.test1/2.asc are simulated in Tran with Vop connected to Vinm (Vcm) and the circuit ???? is stable (not oscillated).?? ??? Question: ??? ??? ??? ??? ??? a) What is the best way to check the OpAmp stability in .Tran? 4)? ? Circuit is diff. input and output:? (Vop-Vom) = Gain*(Vinp-Vim). ??? Question: ??? ??? ??? ??? ??? a) Is it alright to take the phase of (Vop-Vom) to calculate phase margin?I put
all circuits, plot.jpg, models and ReadMe in the folder Files/phase_in_AC. Can someone explain and answer my questions? Thank and regards Hnguyen |
|
I do not think you mean "cascode" and I do not think you mean "saturation".
toggle quoted message
Show quoted text
Upload your entire simulation (not the raw file) to the group's temp folder so that someone else can look at it. Nobody can answer your questions, otherwise. Jim Wagner Oregon Research Electronics. On Dec 4, 2013, at 5:55 PM, hoa van nguyen wrote:
|
|
hoa van nguyen?asked:
Because one is the complement of the other (one is inverted). ?In terms of phase, that means a 180 degree difference from the other, exactly what you saw.
If you plotted -Vom, it should have essentially the same phase as Vop.
Probably the same way it calculates the phase of Vom and of Vop. ?(Vop-Vom) is just another "signal", like Vop is. ?
Could it be that your input signal is a common-mode signal, and what you are measuring at the output is also common-mode?
If the input signal is differential, could it be that the change in biasing causes the amplifier to reach limiting, so that polarities can invert?
Without seeing your schematics, we really can't tell you why.
You should be able to trace out your circuit, and see what the polarities (and therefore the phase) ought to be.
Andy |
|
> ... I do not think you mean "saturation".
For FETs, "saturation" means something different than what some of us analog guys may be accustomed to. ?Normal amplifier operation may be in the "saturation region" where the Id/Vds curves flatten out; but the amplifier is in its linear operating region. ?The amp isn't saturated in the way we tend to think about it.
Andy |
¿ªÔÆÌåÓýWhen I cannot dowload a file from the group, I
refresh The Yahoo¡¯s page.
Then it works
Philippe
From: Andy
Sent: Thursday, December 05, 2013 7:52 AM?
?
To: LTspice@...
Subject: Re: [LTspice] Questions about phase in .AC LTspice
Analysis ? ?
Files
moved from root directory to the "Temp" folder.
?
Trying
to download them now ... Yahoo is not cooperating.? Their servers seem to
be overloaded.
?
Andy
?
? ? |
|
Philippe Basier wrote:
It doesn't work today. ?Yahoo seems to be having major server problems, at least in my area of the world. ?I can't even get to the Files area at all now. ?It only seems to be affecting this group.
I wonder if we are still below the 100MB limit? ?(Yes they say the limit is greater, but in reality it isn't.)
Andy |
|
Has anyone considered moving this forum to Google Groups?
This ¡° Use rich-text editing to customize your posts with fonts, colors, and images.¡± would be very beneficial. w1qg ________________________________________ From: LTspice@... [mailto:LTspice@...] On Behalf Of Andy Sent: Thursday, December 05, 2013 7:01 AM To: LTspice@... Subject: Re: [LTspice] Questions about phase in .AC LTspice Analysis Philippe Basier wrote: When I cannot dowload a file from the group, I refresh The Yahoo¡¯s page. Then it works It doesn't work today. ?Yahoo seems to be having major server problems, at least in my area of the world. ?I can't even get to the Files area at all now. ?It only seems to be affecting this group. I wonder if we are still below the 100MB limit? ?(Yes they say the limit is greater, but in reality it isn't.) Andy |
|
Dick Benson wrote: > Has anyone considered moving this forum to Google Groups? I would not wish that on my least favorite group. Google Groups are just as bad, if not worse. ?Google randomly drops key features, including the "Files" area that they used to have. ?This caused some forums to scramble back to YahooGroups. Our group depends on a reliable and structured file storage space. ?It probably wouldn't exist without it. Imagine if Yahoo decided to drop the "Files" areas for everyone. ?Well that's what Google did. Google is even worse than Yahoo as far as following "fads" is concerned. ?Google will happily drop features and whole programs that people use and love, if they simply decide not to do them anymore. ?As far as I am concerned, Google Groups itself could be at risk of disappearing at any moment. Google is not without its temporary glitches too (like the one affecting my access to the Files area today, here on Yahoo). ?Just this morning I read about another one
like that affecting Gmail. Regards, Andy |
|
John Woodgate
In message <CALBs-TiN24wLua8R=dPjdr=scmfnAMOEMoeXRu0L7Jf4B52iWg@...>, dated Thu, 5 Dec 2013, Andy <Andrew.Ingraham@...> writes:
Dick Benson wrote:Is there any other provider of mailing list services?Has anyone considered moving this forum to Google Groups?I would not wish that on my least favorite group. -- OOO - Own Opinions Only. With best wishes. See www.jmwa.demon.co.uk Nondum ex silvis sumus John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK |
|
John Woodgate asked: Is there any other provider of mailing list services? There are many. ?Unfortunately, most are mailing lists only. The nice thing about YahooGroups is that it comes with a Files area, and a Photos area, and a Links area. ?The combination of all that, plus the easy to access message archives (well, it used to be), and the fact that you can use the group either via the web or your email program, are what made YahooGroups so nice.
I am not aware of another service that has similar features.
Andy |
¿ªÔÆÌåÓýYes, lets do it please!Helmut? On 12/05/2013 10:27 AM, Dick Benson
wrote:
Has anyone considered moving this forum to Google Groups? -- AC2CL I do not think there is any thrill that can go through the human heart like that felt by the inventor as he sees some creation of the brain unfolding to success... Such emotions make a man forget food, sleep, friends, love, everything. - Nikola Tesla |
|
hoa van nguyen
Hello Andy, The schematics are under Files > Temp > phase_in_AC (someone moves the folder there). Could you take a look on that? Thank. Andy, The MOS in saturation region when Ids is flat with Vds-variation. Every books call it MOS in saturation when Ids/Vds is
flat. Thank Hnguyen On Thursday, December 5, 2013 11:14 AM, alzie wrote:
?
Yes, lets do it please! Helmut? On 12/05/2013 10:27 AM, Dick Benson
wrote:
Has anyone considered moving this forum to Google Groups? -- AC2CL I do not think there is any thrill that can go through the human heart like that felt by the inventor as he sees some creation of the brain unfolding to success... Such emotions make a man forget food, sleep, friends, love, everything. - Nikola Tesla |
|
Is there a reason that your so called bias sources, Vcmfb1_2c and Vcmfb2_2c, have an AC component that is hidden on the schematic?
toggle quoted message
Show quoted text
--- In LTspice@..., hoa van nguyen <hnguyen4280@...> wrote:
|
¿ªÔÆÌåÓýI see that the input MOS M7 is not properly biased. The source and drain voltages are both < than the gate voltage. As a result, the simulation indicates that both the drain and source signals are out of phase with the input. As a consequence the source of M8 recives an out of phase signal at its source, which results in an out of phase output on the drain. Both outputs vodp and vodm have the same polarity, when they should be out of phase.The gates of M1 and M6 should be biased at a higher voltage. Le 05/12/2013 21:20, hoa van nguyen a
¨¦crit?:
? |
|
Hello,
1. I am sure the AC 1 in the two other AC sources is an operator mistake. They make no sense for the open loop gain AC analysis. 2. It's much more convenient to drive the input source with AC 1 and not with AC 0.1 as I see in this example.? 3. Vop-Vom has 0¡ã at lower frequency because it's referenced to the input source and thus Vop is in phase. Best regards, Helmut ? |
|
John Woodgate
In message <1386274833.29310.YahooMailNeo@...>, dated Thu, 5 Dec 2013, hoa van nguyen <hnguyen4280@...> writes:
The MOS in saturation region when Ids is flat with Vds-variation. Every books call itYes, but it's very confusing; bipolars are 'saturated' when Vce is as low as it gets, with Ic > 0. I suppose that's more politically acceptable than the original term, 'bottomed'. (;-) -- OOO - Own Opinions Only. With best wishes. See www.jmwa.demon.co.uk Nondum ex silvis sumus John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK |
|
Hnguyen,
I finally was able to download the files and run them. My results do not look like yours! ?Both of my runs (test1 and test2) look essentially the same as one another, and like your test2 graph but not like your test1 graph.
Did you ever wonder why the two outputs are not complementary? ?Why there is a 3 dB difference between them? ?And why the differential output, V(vop,vom) is so much smaller than either output alone? ?The output signal has a strong common-mode component. ?See the earlier notes about your two other voltage sources which contain an AC component.
I don't know why my test1 plot looks so much different than your test1. ?Are you sure it was from the same schematic as the one you uploaded? ?Either they are not, or perhaps you had different parameters/tolerances set in the Control Panel.
Andy |
|
hoa van nguyen
Hello Jerry, I understood in this way: M7/8 are biased to Vinp=Vcm=750mV Drain voltages of M7/8 are ncr/ncl=589.5mV=590mV Source voltages of M7/8 are n_src1_2c=345.856mV=346mV MOS-Threshold VTn=0.2V (with Vbs=0V) in Model; VTn=0.25V (guessing) (with n_src1_2c=346mV). Transistor in saturation when Vds_M7/8 >= Vgs - VTn. Vds_M7/8 = 590mV - 346mV = 244mV Vgs - VTn = (750mV - 346mV) - 250mV = 154mV Vds_M7/8 (244mV) >= Vgs - VTn (154mV): M7/8 should be in saturation and this Cascode Diff. Amplifier should gives high gain which I saw in TRAN-Analysis. All Books call this amplifier topology as Cascode Diff. amplifier, because the transistors are stacked on each other and Diff. is differential inputs/outputs. Your Guys as Analog Guru understand Cascode and Diff. in different way. My English and German are not my native language (I studied in Germany long time ago). Regards Hnguyen On Thursday, December 5, 2013 2:58 PM, Andy wrote:
?
Hnguyen,
I finally was able to download the files and run them. My results do not look like yours! ?Both of my runs (test1 and test2) look essentially the same as one another, and like your test2 graph but not like your test1 graph.
Did you ever wonder why the two outputs are not complementary? ?Why there is a 3 dB difference between them? ?And why the differential output, V(vop,vom) is so much smaller than either output alone? ?The output signal has a strong common-mode component. ?See the earlier notes about your two other voltage sources which contain an AC component.
I don't know why my test1 plot looks so much different than your test1. ?Are you sure it was from the same schematic as the one you uploaded? ?Either they are not, or perhaps you had different parameters/tolerances set in the Control Panel.
Andy |
|
hoa van nguyen
Hello Andy, Vop is complementary from Vom. I ran TRAN-Analysis in this circuit and confirmed that the outputs are complementary. Andy's Q: Did you ever wonder why the two outputs are not complementary? Why
there is a 3 dB difference between them? ?And why the differential
output, V(vop,vom) is so much smaller than either output alone? A: I do not know simulator software good enough to answer your question. But ???? I understood in this way: ??? ??? a) Gain equation:? A=20log(Vo/Vi); Vo=output, Vi=input. ??? ??? b) Vop=53.7dB > Ap=4840 ( with Vinp: AC .1); Ap = gain of Vop ???????????? Vom=50.7db > Am=3420???????????????????????????? ; Am = gain of Vom ???????????? It shows that that Vom is complementary of Vop ( Ap > Am). ??? ??? c) Gain of (Vop - Vom) = 42.7dB > Apm=1364 ??? ??? d) Ap - Am = 1420 is closed to Apm=1364 which is what I saw in ??????????? TRAN-Analysis ( actually in Tran = 1230) and
DC-Analysis. The schematics which you have are the same what I have and no other parameters. My most interest is how to see the stability for this circuit in .AC Analysis Regards Hnguyen On Thursday, December 5, 2013 4:44 PM, hoa van nguyen <hnguyen4280@...> wrote:
?
Hello Jerry, I understood in this way: M7/8 are biased to Vinp=Vcm=750mV Drain voltages of M7/8 are ncr/ncl=589.5mV=590mV Source voltages of M7/8 are n_src1_2c=345.856mV=346mV MOS-Threshold VTn=0.2V (with Vbs=0V) in Model; VTn=0.25V (guessing) (with n_src1_2c=346mV). Transistor in saturation when Vds_M7/8 >= Vgs - VTn. Vds_M7/8 = 590mV - 346mV = 244mV Vgs - VTn = (750mV - 346mV) - 250mV = 154mV Vds_M7/8 (244mV) >= Vgs - VTn (154mV): M7/8 should be in saturation and this Cascode Diff. Amplifier should gives high gain which I saw in TRAN-Analysis. All Books call this amplifier topology as Cascode Diff. amplifier, because the transistors are stacked on each other and Diff. is differential inputs/outputs. Your Guys as Analog Guru understand Cascode and Diff. in different way. My English and German are not my native language (I studied in Germany long time ago). Regards Hnguyen On Thursday, December 5, 2013 2:58 PM, Andy wrote:
?
Hnguyen,
I finally was able to download the files and run them. My results do not look like yours! ?Both of my runs (test1 and test2) look essentially the same as one another, and like your test2 graph but not like your test1 graph.
Did you ever wonder why the two outputs are not complementary? ?Why there is a 3 dB difference between them? ?And why the differential output, V(vop,vom) is so much smaller than either output alone? ?The output signal has a strong common-mode component. ?See the earlier notes about your two other voltage sources which contain an AC component.
I don't know why my test1 plot looks so much different than your test1. ?Are you sure it was from the same schematic as the one you uploaded? ?Either they are not, or perhaps you had different parameters/tolerances set in the Control Panel.
Andy |
Dick Benson
to navigate to use esc to dismiss