开云体育

On 04/18/2013 03:19 PM, Andy wrote:

Figure 21 applies to the output common-mode voltage, not input
common-mode voltage.

You are claiming that a plot of common mode input voltage versus output
voltage says nothing about common mode input voltage?

Surely you must be joking.


Or perhaps looking at a different revision of the datasheet than I am. I
just grabbed one off the Analog Devices web site and that figure is now
Figure 22. "Maximum Output Voltage vs. Common-Mode Input, G=1, Vs=5V, Rl
= 100K"



--
David W. Schultz

Returned for Regrooving

I think Helmut asked a very good question and I'm sure he knows what IBIS is.

I am well aware Helmut knows what IBIS is. He has helped answer other
people's questions about IBIS models. But I got the impression he
either didn't catch that this was an IBIS model, or didn't catch that
the user had an IBIS model and now wanted to know what to do with it
in LTspice.

I thought it was a legitimate question, how do I simulate in LTspice
using this IBIS model. I didn't think Helmut's answer was on-track
with that question, even though it may have been a very good question
to ask.

Like I say, I may have simply misunderstood Helmut's answer.

Andy

--- In LTspice@..., Andy <Andrew.Ingraham@...> wrote:

David W. Schultz <david.schultz@...> wrote:

The AD623 is an instrumentation amplifier and the gain is set with a
resistor between the two Rg pins. While leaving them open should get you
a gain of 1, it makes me nervous. ...

The AD623 data sheet allows (recommends) leaving Rg open. (I
questioned that too, until I read the datasheet.)

The AD623 amplifies the difference between its input voltages. One of
the details that you will find in the data sheet is that it doesn't care
for common mode voltages close to the rails. Your circuit has a common
mode input voltage of zero.

The data sheet allows this. The "input range extends 150 mV below
ground (single supply)." The applied input signal is within that
range.

If you look at Figure 21 in the AD623 data
sheet you will see that it has a sharply reduced output voltage range
under this condition. About 0.5V maximum.

Figure 21 applies to the output common-mode voltage, not input
common-mode voltage.

The output common-mode voltage in this circuit, set by Vref, is
exactly in the middle between +Vs and -Vs.

Me, I am not happy with this simulation. Something seems very wrong.
I downloaded Helmut's suggested fix and it seems to work, EXCEPT that
there is still a strange offset from the desired Vref voltage. I
don't get it. Vout should be centered on +2.5V but it isn't. I tried
implementing Helmut's suggested fix (1.0 ohm Rser) and it didn't work
at all; the voltages "blew up". I downloaded Helmut's other suggested
file and the offset from Vref was still there but in the opposite
direction! Something very strange seems to be going on here. (Or
maybe I am doing something wrong?)

Andy

Hello Andy,

please reset your SPICE settings.

Control Panel -> SPICE -> Reset to default

Best regards,
Helmut

David W. Schultz <david.schultz@...> wrote:

The AD623 is an instrumentation amplifier and the gain is set with a
resistor between the two Rg pins. While leaving them open should get you
a gain of 1, it makes me nervous. ...

The AD623 data sheet allows (recommends) leaving Rg open. (I
questioned that too, until I read the datasheet.)

The AD623 amplifies the difference between its input voltages. One of
the details that you will find in the data sheet is that it doesn't care
for common mode voltages close to the rails. Your circuit has a common
mode input voltage of zero.

The data sheet allows this. The "input range extends 150 mV below
ground (single supply)." The applied input signal is within that
range.

If you look at Figure 21 in the AD623 data
sheet you will see that it has a sharply reduced output voltage range
under this condition. About 0.5V maximum.

Figure 21 applies to the output common-mode voltage, not input
common-mode voltage.

The output common-mode voltage in this circuit, set by Vref, is
exactly in the middle between +Vs and -Vs.

Me, I am not happy with this simulation. Something seems very wrong.
I downloaded Helmut's suggested fix and it seems to work, EXCEPT that
there is still a strange offset from the desired Vref voltage. I
don't get it. Vout should be centered on +2.5V but it isn't. I tried
implementing Helmut's suggested fix (1.0 ohm Rser) and it didn't work
at all; the voltages "blew up". I downloaded Helmut's other suggested
file and the offset from Vref was still there but in the opposite
direction! Something very strange seems to be going on here. (Or
maybe I am doing something wrong?)

Andy

Jean Pierre Daviau <daviaujp@...> wrote:

No file

See Helmut's first reply. He moved it.

--- In LTspice@..., "sawreyrw" <sawreyrw@...> wrote:

--- In LTspice@..., Andy <Andrew.Ingraham@> wrote:

What do you expect from an ADC model in SPICE?

Maybe you misunderstood the question, Helmut. Or maybe I
misunderstood the answer.

Linear Tech. has only an IBIS file for this part, no SPICE model
(apparently). LTspice doesn't take IBIS model files. Some other
SPICE simulators do. There is a program to convert IBIS model files
to SPICE models, but it does not always give usable results. Even so,
you need to be aware of what IBIS models are for ... they represent
the electrical characteristics of pins and their attached circuitry,
but not of the entire IC.

Andy

Andy,

I think Helmut asked a very good question and I'm sure
he knows what IBIS is.

Rick

Hello Rick,

Yes I well know IBIS files. But Elena asked like not be
interested in IBIS files.

Best regards,
Helmut

--- In LTspice@..., Andy <Andrew.Ingraham@...> wrote:

What do you expect from an ADC model in SPICE?

Maybe you misunderstood the question, Helmut. Or maybe I
misunderstood the answer.

Linear Tech. has only an IBIS file for this part, no SPICE model
(apparently). LTspice doesn't take IBIS model files. Some other
SPICE simulators do. There is a program to convert IBIS model files
to SPICE models, but it does not always give usable results. Even so,
you need to be aware of what IBIS models are for ... they represent
the electrical characteristics of pins and their attached circuitry,
but not of the entire IC.

Andy

Andy,

I think Helmut asked a very good question and I'm sure he knows what IBIS is.

Rick

--- In LTspice@..., Jean Pierre Daviau <daviaujp@...> wrote:

No file

----- Original Message -----
.

Jean,

Links like that in the original message don't don't work, but you can "figure out" that it is in files/files sorted by message number etc.

Rick

What do you expect from an ADC model in SPICE?

Maybe you misunderstood the question, Helmut. Or maybe I
misunderstood the answer.

Linear Tech. has only an IBIS file for this part, no SPICE model
(apparently). LTspice doesn't take IBIS model files. Some other
SPICE simulators do. There is a program to convert IBIS model files
to SPICE models, but it does not always give usable results. Even so,
you need to be aware of what IBIS models are for ... they represent
the electrical characteristics of pins and their attached circuitry,
but not of the entire IC.

Andy

--- In LTspice@..., Jean Pierre Daviau <daviaujp@...> wrote:

No file

----- Original Message -----
From: free2rhymedd
To: LTspice@...
Sent: Thursday, April 18, 2013 2:37 PM
Subject: [LTspice] AD623



I am new to the LTSPICE. I was trying to shift the input voltage up by 2.5V. I am using AD6232 to shift up the voltage, but it does not work properly. I have uploaded the circuit on the forum. Here is the link:
.

Hope someone can help me. Thank you.

Hello,
I can't see your uploaded file. Please check that you have
really uploaded it. Your mentioned directory doesn't exist
anymore.

Normally users should upload circuits for discussion to
Files > Temp > your_files.zip

Best regards,
Helmut

On 04/18/2013 01:37 PM, free2rhymedd wrote:

I am new to the LTSPICE. I was trying to shift the input voltage up
by 2.5V. I am using AD6232 to shift up the voltage, but it does not
work properly. I have uploaded the circuit on the forum. Here is the
link:
.

Hope someone can help me. Thank you.

The AD623 is an instrumentation amplifier and the gain is set with a
resistor between the two Rg pins. While leaving them open should get you
a gain of 1, it makes me nervous. At the very least you have to be vary
careful about leakage currents between the two open pins.

The AD623 amplifies the difference between its input voltages. One of
the details that you will find in the data sheet is that it doesn't care
for common mode voltages close to the rails. Your circuit has a common
mode input voltage of zero. If you look at Figure 21 in the AD623 data
sheet you will see that it has a sharply reduced output voltage range
under this condition. About 0.5V maximum.

If I replace the ground connection on the input with a 2.5V DC source to
provide an offset, the output of the AD623 follows the input as
expected. (Or replace it with a connection to your 2.5V reference circuit.)


Despite its common mode voltage limitations, the AD623 is one of my
favorite instrumentation amps. I used it in this project:



--
David W. Schultz

Returned for Regrooving

No file

Hi, Helmut

Thank you very much. It works now. I appreciate it very much.

Xinjun

--- In LTspice@..., "free2rhymedd" <xinjundong@...> wrote:

I am new to the LTSPICE. I was trying to shift the input voltage up by 2.5V. I am using AD6232 to shift up the voltage, but it does not work properly. I have uploaded the circuit on the forum. Here is the link:
.

Hope someone can help me. Thank you.

Hello,

I agree that this circuit converged to a wrong DC operating point.

I tried different things. Finally I decided to use the solution
with only one +5V supply. This has the advantage that it's
useful for the .AC simulation too.

I have added data labels to see the DC-operating point voltages.
This enables you to check that ..AC-simulation has been done
in the correct DC-operating point.

I just discovered that it has been sufficient to set a 1Ohm
series resistance in V1. See L-28_.asc.

I have moved your files and added my solutions.

Files > Files sorted by message number > msg_66295


Best regards,
Helmut

I am new to the LTSPICE. I was trying to shift the input voltage up by 2.5V. I am using AD6232 to shift up the voltage, but it does not work properly. I have uploaded the circuit on the forum. Here is the link:
.

Hope someone can help me. Thank you.

Hi amaktoom,

Thank you very much for that explanation about the time invariant system. Like you I learn something every time.
By the way I posted a simplified version of the op Amp response, see in the Temp Files 'Re: Inverting Op Amp'
Please note:- the first block in the chain represents the transfer function of the proportional block (unity gain, for convinience); the next block represents the transfer function of a single exponential lag(unity gain for convinience, you can select any gain to suit); the next block represents the R2/R1 of the op amplifier. Vxa represents the overall transfer function of the preceding blocks and Vxb represents the corresponding output signal response after multiplyng the overall transfer function by the input signal, w
[Vo =(Vo/Vin)*Vin].

Best regards,

Michael P Kiwanuka




To: LTspice@...
From: amaktoomamu@...
Date: Thu, 18 Apr 2013 17:21:40 +0000
Subject: [LTspice] Re: inverting opamp simulation: rapid component variation






Dear Michael,
I will be able to see your solution closely after completing some work that I am doing currently. To use your solution, I will have learn many new concepts of control system (and perhaps including that of LTspice! what an irony that I am writing FREE beginners book for LTspice: !).
I am not good in Signals System (or any other subject- still a lot of concepts to learn) but to quote wiki-
"Time invariance means that whether we apply an input to the system now or T seconds from now, the output will be identical except for a time delay of the T seconds. That is, if the output due to input x(t) is y(t), then the output due to input x(t-T) is y(t-T). Hence, the system is time invariant because the output does not depend on the particular time the input is applied".
Using this definition, the system with output Vo=-(R2/R1)Vi will be linear for some R2(t) and Non-linear for other R2(t)- it depends what kind of function R2 is that of time.

I actually avoid those discussion here that don't relate to LTspice specifically. Although I very much like to listen to the views of experienced members that we have here, If you wish we could take this discussion to edaboard. I would love to learn new ideas.

Laplace is not valid for such time varying systems:

--- In LTspice@..., Michael Peter Kiwanuka <michael883575@...> wrote:

Hi amaktoom,

Thank you for your message. The 'w' is supposed to be the same signal as that at the input of the control process. It was not meant to be replaced by any value although that may correct the LTspice error. To correct the LTspice error issued, delete the source V2 and label the subsquently remaining input line to the proportional block. Label it 'w' . The LTspice error will go and things will run as intended.

Could you please explain to me what a time varying signal/circuit is. I thought Laplace is intended for time varying functions and signals. Please enlighten me about the prof, advice.

Best regards,

Michael




To: LTspice@...
From: amaktoomamu@...
Date: Thu, 18 Apr 2013 13:58:42 +0000
Subject: [LTspice] Re: inverting opamp simulation: rapid component variation







--- In LTspice@..., Michael Peter Kiwanuka <michael883575@> wrote:

Hello Michael
Thanks for your file. I saw that file and when it reported error msg for 'w' i assigned it some value and it ran smoothly.
But, I don't have much idea of control theory so i couldn't follow at all what you have tried to do. However, Thank you for taking my attention Helmut's control lib. Its a new thing for me to do control theory using LTspice.
A prof (Phd form UCB in 1960) told me that my circuit is time varying so i cant think of 'transfer function, Laplace etc. That's why I decided to manage things with time-domain plots as suggested by Andy and others.
Thank you,again.

Hi amaktoom,

I posted a solution in the temp files of this group

and I assumed that the source signal approximate a steady voltage over an appropriate period of time. For example, the source frequency could be 1mHz and in modelling this, an AC source would definetely be adequate and correct and would approximate to a steady dc soorce for a relatively short period of time. The resulting response of gain Vs frequency would not be far too wrong. Please take a look at my solution.

Best regards

Michael P Kiwanuka




To: LTspice@...
From: amaktoomamu@
Date: Wed, 17 Apr 2013 10:28:43 +0000
Subject: [LTspice] Re: inverting opamp simulation: rapid component variation





Hi All,
Thanks for your comments. I got it why hypothesis doesn't make sense at high frequency. I will rely on transient, as suggested.

--- In LTspice@..., Andy <Andrew.Ingraham@> wrote:

MOHAMMAD A MAKTOOMI <amaktoomamu@> wrote:

Thank you, Andy for your hints.
But, in '.AC' we need to have an AC voltage or current source. Here, I don't
have any such thing (as I wish to vary the frequency of R2, NOT that of any
source), so how do I proceed?

Yes indeed, and I think that is why you must do this as a .TRAN
analysis. .AC analysis just doesn't work for your case.

Then .STEP the frequency.

With your 741 op-amp model, you do have an interesting situation
because of the severe distortion at higher frequencies (which by the
way only a .TRAN analysis would show, if there was a way to do this as
a .AC analysis). Given that, you need to consider what parameter of
output amplitude you measure: peak-to-peak vs. RMS vs. fundamental
amplitude?

Andy

[Non-text portions of this message have been removed]

[Non-text portions of this message have been removed]

[Non-text portions of this message have been removed]

--- In LTspice@..., "lainthales" <miboe24238@...> wrote:

Hello Guys,

im looking for a model of the ltc 1666 / 1667 / 1668
its an ADC and i cant find any ADC models at the standart lib of LTSpice

someone got this model? or got an idea what i should use as equivalent?

thx :)!

Hello,

The LTC1666 is not an ADC. It's a DAC.
That's a very fundamental difference.
Please read the datasheet more carefully. :-)


Best regards,
Helmut

--- In LTspice@..., "elena.ruorui" <elena.ruorui@...> wrote:

HI all,
i'm really new with LTspice...
I want to simulate an LTC2377-18 SAR ADC, but in the library there isnt any LTC2377 model.
I've seen that the LT gives an .ibis file. How can i add or use this with LTspice?
Or maybe there is another library with this component in it?

Thanks in advance for help!

Hello,

May be you misunderstood the target applications/simulations
of SPICE programs.

What do you expect from an ADC model in SPICE?

Best regards,
Helmut

In message <kkp9jt+euqg@...>, dated Thu, 18 Apr 2013, hisotaso <flirm777@...> writes:

In the header of the .lib file from which I posted my example, mu professor states "Models are written in LTSpice syntax." But where is this syntax detailed thoroughly!?

Are you not allowed to ask your professor? OK, some give silly answers, but not all.

Also, go to the list's web site and click on 'Links' in the left column. That leads you to enough documentation to last at least a year.
--
OOO - Own Opinions Only. See www.jmwa.demon.co.uk
They took me to a specialist burns unit - and made me learn 'To a haggis'.

John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK

Dear Michael,
I will be able to see your solution closely after completing some work that I am doing currently. To use your solution, I will have learn many new concepts of control system (and perhaps including that of LTspice! what an irony that I am writing FREE beginners book for LTspice: !).
I am not good in Signals System (or any other subject- still a lot of concepts to learn) but to quote wiki-
"Time invariance means that whether we apply an input to the system now or T seconds from now, the output will be identical except for a time delay of the T seconds. That is, if the output due to input x(t) is y(t), then the output due to input x(t-T) is y(t-T). Hence, the system is time invariant because the output does not depend on the particular time the input is applied".
Using this definition, the system with output Vo=-(R2/R1)Vi will be linear for some R2(t) and Non-linear for other R2(t)- it depends what kind of function R2 is that of time.

I actually avoid those discussion here that don't relate to LTspice specifically. Although I very much like to listen to the views of experienced members that we have here, If you wish we could take this discussion to edaboard. I would love to learn new ideas.

Laplace is not valid for such time varying systems: