开云体育

Hello Helmut,

"This file contains a PowerPoint file(.ppt) with Mike's
presentation. Please open this ppt-file in the presentation
mode. Then you can directly open the hyperlinked examples
by clciking on the arrow-icons in the slides. This feature
only works with MS-PowerPoint, but not with OpenOffice-Impress."

I haven't been able to get the feature to work with 2007 PowerPoint Viewer either.


Regards,

Howard


Howard.Rogers@...

Hello Helmut,
What are you meaning by " LTspice uses the whole
screen. " ?
As you can see in my little benchmarks, there a big difference between mini and maxi ?
And I deleted some curious results.
In fact, I think that a lot of parameters ( graphic card, Windows or Unix, ............) are to be in head.
By the way : simulations, on my PC, seem to be a little bit longer with 4.00c ?
Best regards
Philippe

just kidding

seriously, I mean it, I would pay someone to generate some transistor models for me. SERIOUSLY.

There's no point in me trying to do it myself, it would take a year and I'd never get it right anyway.

I have googled for these for hours, have searched the archives, and I'm just chasing my tail on these.

I would prefer subcircuits, but, trust me, I'd be perfectly happy with models.

Or, I would settle for models/subcircuits for devices that are analogs of these. And, I don't expect (of course) the models/subcircuits to be perfect.

If you are good at generating SS models/subcircuits, please contact me off-list. I will have more in the future. I will also post them for free; this is not for proprietary use, I just need 'em!!

The ones I ESPECIALLY need right now are (models/subcircuits/analogs):

Exicon ECF20N25A (nmos lateral mosfet)
Exicon ECF20P25EA (pmos lateral mosfet)
Fairchild KSC1845U
Sanken 2SC2911
Sanken 2SA1209
Exicon EXC10N20
Exicon EXC10P20

best, charlie

Hi Howard,

You're right. The first Models of 555/556 was from
the LTspice "Masc." folder. The second attempt
was the transistor realization from LTspice's
/examples/Educational NE555.asc, following
Mike's advise. Behavioral and transistor model gave me
the same (wrong) results for small signal inputs.
I investigated on the website and I found a few different
realization of these subcircuits, but they had the same
problems.

Correctly the problem was with low voltage behavior of
Trigger and Threshold inputs. I tried to make PWM circuitry
which was perfect in simulation in range of 0 to Vcc, but on
the bench the physical realization felt under 1V ov input voltage.
I asked for samples from IC vendors (ST, TI). On both these
imputes behaved differently, they lost linearity under
1V at Vcc=5V.

Regards,
Peter

--- In LTspice@..., "philippe b" <basier.philippe@...> wrote:

Hello "????",

I made some benchmarks on my PC.
I made an excel file.
Are you interested ?
How send it on the group ?
Regards,
Philippe

Hello Philippe,

Please upload your file to the Files section.

Files > Temp



Best regards,
Helmut


PS: By the way, I have discoverd that the simulation time of
analog's test circuit is longer if LTspice uses the whole
screen.

--- In LTspice@..., "Ken" <kpotter999@...> wrote:

--- In LTspice@..., "Helmut" <helmutsennewald@> wrote:

--- In LTspice@..., "kpotter999" <kpotter999@> wrote:

Dear Group Members,

I have been attempting to use the following commands
to calculate the output rise and fall times of an inverter
and the propagation delay with the following voltage source:

PULSE(0 3.3 1n 1.5n 1.5n 5n 10n)

.MEAS TRAN fall_delay
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1

.MEAS TRAN rise_delay
+ TRIG V(Vout) VAL=3.3*0.1 TD=0NS RISE=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0NS RISE=1

.MEAS TRAN propagation_fall
+ TRIG V(Vin) VAL=3.3*0.5 TD=0 RISE=1
+ TRIG V(Vout) VAL=3.3*0.5 TD=0 FALL=1

.MEAS TRAN propagation_rise
+ TRIG V(Vin) VAL=3.3*0.5 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.5 TD=0 RISE=1

Unfortunately the results from the .MEAS text do not
agree with manual plotting and reading of the waveforms.

i.e.

From Waveform, Rise time = 118ps, Fall time = 152ps

From text, Rise time = 6.7ns, Fall time = 1.6ns

Clearly, I am doing something fundamentally wrong with
the .MEAS command but I have been unable to resolve
this further. I would really appreciatte some feedback
and can supply the circuit schematic and inverter model
if required.

Many thank, Ken Potter.

Hello Ken,

I have uplaoded an example for you.

Files > Tut > MEASURE > TRAN >RC_measure_rise_fall_delay.asc



I have used the following MEASURE-commands.

.MEAS t1r WHEN V(out)=0.1*5 TD=0 RISE=2
.MEAS t2r WHEN V(out)=0.9*5 TD=0 RISE=2

.MEAS t1f WHEN V(out)=0.9*5 TD=0 FALL=2
.MEAS t2f WHEN V(out)=0.1*5 TD=0 FALL=2

.MEAS tdr1 WHEN V(in)=0.5*5 TD=0 RISE=2
.MEAS tdr2 WHEN V(out)=0.5*5 RISE=2

.MEAS tdf1 WHEN V(in)=0.5*5 TD=0 FALL=2
.MEAS tdf2 WHEN V(out)=0.5*5 TD=0 FALL=2

.MEAS trise PARAM t2r-t1r
.MEAS tfall PARAM t2f-t1f
.MEASURE tdelay_rise PARAM tdr2-tdr1
.MEASURE tdelay_fall PARAM tdf2-tdf1

Best regards,
Helmut

Dear Helmut,

Thank you very much for your reply, I will go and figure this out.

Regards, Ken.

Dear Helmut,

Just to let you know that worked perfectly! Very gratefull, Ken.

Hello "????",

I made some benchmarks on my PC.
I made an excel file.
Are you interested ?
How send it on the group ?
Regards,
Philippe

--- In LTspice@..., "heifetz.7777" <heifetz.7777@> wrote:

2. You have another file LM317.lib in the path
"C:&#92;Program Files&#92;LTC&#92;LTspiceIV&#92;lib&#92;sub&#92;".
This is the path where LTspice firstly look for included
files.

Best regards,
Helmut

Thank you very much! That was the problem. Runs just fine now.

A tip of the hat!
Paul

Our class instructor requires Date and Name on LTSpice screenshots (to prevent cheating, I think). I have not figured out how to add Date And Name to LTSpice plots. Is there a way of automatic generation, say, via spice directive? (I would hate to copy and paste the sucker, manually, to all the plots, or, worse, to use third party-app to do it).

--- In LTspice@..., "Helmut" <helmutsennewald@...> wrote:

--- In LTspice@..., "kpotter999" <kpotter999@> wrote:

Dear Group Members,

I have been attempting to use the following commands
to calculate the output rise and fall times of an inverter
and the propagation delay with the following voltage source:

PULSE(0 3.3 1n 1.5n 1.5n 5n 10n)

.MEAS TRAN fall_delay
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1

.MEAS TRAN rise_delay
+ TRIG V(Vout) VAL=3.3*0.1 TD=0NS RISE=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0NS RISE=1

.MEAS TRAN propagation_fall
+ TRIG V(Vin) VAL=3.3*0.5 TD=0 RISE=1
+ TRIG V(Vout) VAL=3.3*0.5 TD=0 FALL=1

.MEAS TRAN propagation_rise
+ TRIG V(Vin) VAL=3.3*0.5 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.5 TD=0 RISE=1

Unfortunately the results from the .MEAS text do not
agree with manual plotting and reading of the waveforms.

i.e.

From Waveform, Rise time = 118ps, Fall time = 152ps

From text, Rise time = 6.7ns, Fall time = 1.6ns

Clearly, I am doing something fundamentally wrong with
the .MEAS command but I have been unable to resolve
this further. I would really appreciatte some feedback
and can supply the circuit schematic and inverter model
if required.

Many thank, Ken Potter.

Hello Ken,

I have uplaoded an example for you.

Files > Tut > MEASURE > TRAN >RC_measure_rise_fall_delay.asc



I have used the following MEASURE-commands.

.MEAS t1r WHEN V(out)=0.1*5 TD=0 RISE=2
.MEAS t2r WHEN V(out)=0.9*5 TD=0 RISE=2

.MEAS t1f WHEN V(out)=0.9*5 TD=0 FALL=2
.MEAS t2f WHEN V(out)=0.1*5 TD=0 FALL=2

.MEAS tdr1 WHEN V(in)=0.5*5 TD=0 RISE=2
.MEAS tdr2 WHEN V(out)=0.5*5 RISE=2

.MEAS tdf1 WHEN V(in)=0.5*5 TD=0 FALL=2
.MEAS tdf2 WHEN V(out)=0.5*5 TD=0 FALL=2

.MEAS trise PARAM t2r-t1r
.MEAS tfall PARAM t2f-t1f
.MEASURE tdelay_rise PARAM tdr2-tdr1
.MEASURE tdelay_fall PARAM tdf2-tdf1

Best regards,
Helmut

Dear Helmut,

Thank you very much for your reply, I will go and figure this out.

Regards, Ken.

--- In LTspice@..., "kpotter999" <kpotter999@...> wrote:

Dear Group Members,

I have been attempting to use the following commands
to calculate the output rise and fall times of an inverter
and the propagation delay with the following voltage source:

PULSE(0 3.3 1n 1.5n 1.5n 5n 10n)

.MEAS TRAN fall_delay
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1

.MEAS TRAN rise_delay
+ TRIG V(Vout) VAL=3.3*0.1 TD=0NS RISE=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0NS RISE=1

.MEAS TRAN propagation_fall
+ TRIG V(Vin) VAL=3.3*0.5 TD=0 RISE=1
+ TRIG V(Vout) VAL=3.3*0.5 TD=0 FALL=1

.MEAS TRAN propagation_rise
+ TRIG V(Vin) VAL=3.3*0.5 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.5 TD=0 RISE=1

Unfortunately the results from the .MEAS text do not
agree with manual plotting and reading of the waveforms.

i.e.

From Waveform, Rise time = 118ps, Fall time = 152ps

From text, Rise time = 6.7ns, Fall time = 1.6ns

Clearly, I am doing something fundamentally wrong with
the .MEAS command but I have been unable to resolve
this further. I would really appreciatte some feedback
and can supply the circuit schematic and inverter model
if required.

Many thank, Ken Potter.

Hello Ken,

I have uplaoded an example for you.

Files > Tut > MEASURE > TRAN >RC_measure_rise_fall_delay.asc



I have used the following MEASURE-commands.

.MEAS t1r WHEN V(out)=0.1*5 TD=0 RISE=2
.MEAS t2r WHEN V(out)=0.9*5 TD=0 RISE=2

.MEAS t1f WHEN V(out)=0.9*5 TD=0 FALL=2
.MEAS t2f WHEN V(out)=0.1*5 TD=0 FALL=2

.MEAS tdr1 WHEN V(in)=0.5*5 TD=0 RISE=2
.MEAS tdr2 WHEN V(out)=0.5*5 RISE=2

.MEAS tdf1 WHEN V(in)=0.5*5 TD=0 FALL=2
.MEAS tdf2 WHEN V(out)=0.5*5 TD=0 FALL=2

.MEAS trise PARAM t2r-t1r
.MEAS tfall PARAM t2f-t1f
.MEASURE tdelay_rise PARAM tdr2-tdr1
.MEASURE tdelay_fall PARAM tdf2-tdf1

Best regards,
Helmut

--- In LTspice@..., "kpotter999" <kpotter999@...> wrote:

Dear Group Members,

I have been attempting to use the following commands to calculate the output rise and fall times of an inverter and the propagation delay with the following voltage source:

PULSE(0 3.3 1n 1.5n 1.5n 5n 10n)

.MEAS TRAN fall_delay
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1

.MEAS TRAN rise_delay
+ TRIG V(Vout) VAL=3.3*0.1 TD=0NS RISE=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0NS RISE=1

.MEAS TRAN propagation_fall
+ TRIG V(Vin) VAL=3.3*0.5 TD=0 RISE=1
+ TRIG V(Vout) VAL=3.3*0.5 TD=0 FALL=1

.MEAS TRAN propagation_rise
+ TRIG V(Vin) VAL=3.3*0.5 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.5 TD=0 RISE=1

Unfortunately the results from the .MEAS text do not agree with manual plotting and reading of the waveforms.

i.e.

From Waveform, Rise time = 118ps, Fall time = 152ps

From text, Rise time = 6.7ns, Fall time = 1.6ns

Clearly, I am doing something fundamentally wrong with the .MEAS command but I have been unable to resolve this further. I would really appreciatte some feedback and can supply the circuit schematic and inverter model if required.

Many thank, Ken Potter.

PLEASE IGNORE THE TYPING ERROR the 3rd line should say 3.3*0.1

.MEAS TRAN fall_delay

+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1

Dear Group Members,

I have been attempting to use the following commands to calculate the output rise and fall times of an inverter and the propagation delay with the following voltage source:

PULSE(0 3.3 1n 1.5n 1.5n 5n 10n)

.MEAS TRAN fall_delay
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0 FALL=1

.MEAS TRAN rise_delay
+ TRIG V(Vout) VAL=3.3*0.1 TD=0NS RISE=1
+ TRIG V(Vout) VAL=3.3*0.9 TD=0NS RISE=1

.MEAS TRAN propagation_fall
+ TRIG V(Vin) VAL=3.3*0.5 TD=0 RISE=1
+ TRIG V(Vout) VAL=3.3*0.5 TD=0 FALL=1

.MEAS TRAN propagation_rise
+ TRIG V(Vin) VAL=3.3*0.5 TD=0 FALL=1
+ TRIG V(Vout) VAL=3.3*0.5 TD=0 RISE=1

Unfortunately the results from the .MEAS text do not agree with manual plotting and reading of the waveforms.

i.e.

From Waveform, Rise time = 118ps, Fall time = 152ps

From text, Rise time = 6.7ns, Fall time = 1.6ns

Clearly, I am doing something fundamentally wrong with the .MEAS command but I have been unable to resolve this further. I would really appreciatte some feedback and can supply the circuit schematic and inverter model if required.

Many thank, Ken Potter.

--- In LTspice@..., "Helmut" <helmutsennewald@...> wrote:

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

...
I have uploaded a very simple example simulation that
demonstrates this situation, yet is otherwise very
computationally light (it is set up specifically not to
burden the hard drive nor benefit from conversion to Fast
Access format).

The plot file is not strictly necessary, but makes the slow
trace automatically plot as the simulation first runs in
order to fully demonstrate the maximum slowdown effect
(close the plot window to regenerate this first run effect
in order to make "apples-to-apples" run time comparisons).

1) Look at run times in the SPICE Error Log and compare
first run run times with the plot window displayed verses
it being minimized (it's a ratio of almost 20:1 with my
dual core integrated graphics laptop - I am really
interested to know if higher end graphics cards help to
reduce this ratio).
...

Hello analogspiceman,

I checked your test case "1)" on two "older" laptops and a
desktop PC. The first value is with plot window and the
second value is with minmized plot window.

Laptop 1
Core Duo 1.66GHz, ATI Mobility Radeon HD2400
47.7s -> 6.7s

Laptop 2
Core Duo 1.66GHz, Geforce Go 7600
23.4s -> 6.85s

Desktop
Core 2 Duo 2.4GHZ, Geforce 8800GT
14.3s -> 4.4s


I will check it the next days on other laptops if possible.

Best regards,
Helmut

More benchmarks...

Desktop (Linux x64 with WINE)
Phenom X4 3GHz, GeForce 9400GT
12.7s -> 4.2s
Windows tiled: 8.7s
Tiled, .options nomarch: 3.7

Laptop (WinXP 32)
Pentium 4 2.8GHz, Intel 8284G Integrated
37.2s -> 7.5s
Windows tiled: 20.8s
Tiled, .options nomarch: 7.1

Regards,
Tony

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

Does anyone have the correct address for the
POWERPOINT PRESENTATION...the one I have does not appear
to be working:



Can someone send me the correct address

Hello William,

Sorry, it seems I had overlooked your question.

The URL is in our Links-section on the first page.



Link to the world tour zip-file:



This file contains a PowerPoint file(.ppt) with Mike's
presentation. Please open this ppt-file in the presentation
mode. Then you can directly open the hyperlinked examples
by clciking on the arrow-icons in the slides. This feature
only works with MS-PowerPoint, but not with OpenOffice-Impress.

Best regards,
Helmut

--- In LTspice@..., "lsi_ken_stewart" <gordon_ken_stewart@...> wrote:

I'm having trouble implementing the F model for current
dependent current sources. The syntax is listed as
Fxxx n+ n- <Vnam> <gain>. However when I add something
like V1 5 to either of the Value boxes or the Spicelines
(CTRL-right click), I get an error saying

Error on line 2: f1 n001 0 f v1 5
Unknown parameter "v1"
Fatal Error: Missing gain value.

Any examples are appreciated.


Ken Stewart

Hello Ken,

I answered a similar question a week ago.

---

Hello Bill,

It's the F-source. An F-source always require an additional
voltage source for current measurement.

There is an example in our Files section.

Files > Tut > Linear_dependent_sources_E_F_G_H.asc



---


Best regards,
Helmut

--- In LTspice@..., John Woodgate wrote:

I had 8.9 s with the schematic and waveform windows one above the
other, which didn't seem excessive. But I had plotted V1 first, so
I didn't get changes of vertical resolution when plotting V2.

True, but you still would get a lot of screen repainting as the
simulation time progressed. The window tile pattern probably only
makes a small difference, if any.

The most astounding thing is the run time you reported with the
waveform window being actively updated. A shade under nine seconds
is a very fast redraw - much faster than anything reported so far.

What is the platform that produced this time? (Processor and more
importantly the graphics card.)

Regards -- analogspiceman

In message <he9pa0+gm2d@...>, dated Sat, 21 Nov 2009, analogspiceman <analogspiceman@...> writes:

-- In LTspice@..., John Woodgate <jmw@...> wrote:

What magnitude of redraw time is involved?

Just click on the example - you'll see. :)

I did, that's why I asked. See below.

With the waveform window minimized the SPICE Error Log reports a
total elapsed time of 6 or 7 seconds on my current laptop.

I had 8.9 s with the schematic and waveform windows one above the other, which didn't seem excessive. But I had plotted V1 first, so I didn't get changes of vertical resolution when plotting V2.

A first
run with the schematic and waveform windows side-by-side takes 120
to 130 seconds. I assume most of the time difference is due to
redraw time.

I agree that 120 s is unreasonably long. I don't recall ever getting such a long time, except with a schematic or Spice instruction that was incorrect.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK
Help stamp out intolerance!

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

--- In LTspice@..., Helmut Sennewald wrote:

I checked your test case "1)" on two "older" laptops and a desktop
PC. The first value is with plot window and the second value is
with minimized plot window.

Laptop 1
Core Duo 1.66GHz, ATI Mobility Radeon HD2400
47.7s -> 6.7s

Laptop 2
Core Duo 1.66GHz, Geforce Go 7600
23.4s -> 6.85s

Desktop
Core 2 Duo 2.4GHZ, Geforce 8800GT
14.3s -> 4.4s

I will check it the next days on other laptops if possible.

Thanks for that. :)

Those all appear to be various types of integrated graphics ICs,
right?

As I understand it, Integrated Graphics is part of the motherboard
and, because it shares memory with with the main processor, suffers
quite a memory access bottleneck (the data bus speed and bus bit
width can be much higher in a dedicated graphics card).

I just checked on an older desktop Pentium D with Intel integrated
graphics.

170s (First run - ouch!)
6.8s (Minimized Waveform Window)

Regards -- analogspiceman

Hello analogspiceman,

At least the Go 7600 and the Geforce 8800GT have their own
graphic memory. They don't use the CPU memory. I am not sure
about the memory used by the HD2400.

Best regards,
Helmut

--- In LTspice@..., Helmut Sennewald wrote:

I checked your test case "1)" on two "older" laptops and a desktop
PC. The first value is with plot window and the second value is
with minimized plot window.

Laptop 1
Core Duo 1.66GHz, ATI Mobility Radeon HD2400
47.7s -> 6.7s

Laptop 2
Core Duo 1.66GHz, Geforce Go 7600
23.4s -> 6.85s

Desktop
Core 2 Duo 2.4GHZ, Geforce 8800GT
14.3s -> 4.4s

I will check it the next days on other laptops if possible.

Thanks for that. :)

Those all appear to be various types of integrated graphics ICs,
right?

As I understand it, Integrated Graphics is part of the motherboard
and, because it shares memory with with the main processor, suffers
quite a memory access bottleneck (the data bus speed and bus bit
width can be much higher in a dedicated graphics card).

I just checked on an older desktop Pentium D with Intel integrated
graphics.

170s (First run - ouch!)
6.8s (Minimized Waveform Window)

Regards -- analogspiceman