Entire Forum This board This topic Members Entire Site
 Pages: [1]
 Author Topic: How to measure frequency vs voltage  (Read 5715 times) 0 Members and 1 Guest are viewing this topic.
dotm
Active Member

Offline

Posts: 162

Thank You
-Given: 65

C15H15NO2S

 « on: April 16, 2012, 10:15:23 10:15 »

how do i do that:
I got myself a couple of VCOs designed.
Now i want to plot frequency vs voltage in Spice.
I got Proteus and Multisim.
I tried to make a VI in Labview but this is way to complex for this simple task. But how can i achieve such a measurement?
As far as i see it can only be done with multiple transient analysis. And it should wait for the VCO to start oscillating. In multisim there is no way to get hold of the frequency value of the probe and the there is not xspice directive like .meas (could define a user xspice command simulation in multisim)
.meas is available in ltspice but not the bb112 as a model (could be programmed) but ltspice lacks of mcu cosimulation and is ugly and i hate it
this gives me sleepless nights , please relieve me!
 Logged
pickit2
Moderator
Hero Member

Offline

Posts: 4087

Thank You
-Given: 646

There is no evidence that I muted SoNsIvRi

 « Reply #1 on: April 16, 2012, 12:24:33 12:24 »

I seen that question at http://denethor.wlu.ca/PSpice/pspice_tutorial.html
 Logged

Note: A forum without posters is not a forum. posts of no value to topic maybe deleted
borberk
Senior Member

Offline

Posts: 298

Thank You
-Given: 309

 « Reply #2 on: April 16, 2012, 01:30:38 13:30 »

I don't think it is possible. In time domain you can use HARM fuction only to find frequency from spectrum of output signal for specified voltage. No possibility to get f(v) graph.
You can use Mathcad and write needed equiation. For varicap BB112 you will need to find suitable function C(v) (for example by means of CurveExpert and Datasheet).
 Logged
dotm
Active Member

Offline

Posts: 162

Thank You
-Given: 65

C15H15NO2S

 « Reply #3 on: April 16, 2012, 03:03:02 15:03 »

in LTspice i could measure after a transient analysis
i should decouple the dc part or subtract it then i do

.MEAS TRAN time when V(out)=0 cross 100
wich means give me the time to the 100th zero-crossing.

after finding the end of a period:
.MEAS TRAN time2 when V(out)=0 cross 102

i can determine the frequency
.MEAS TRAN freq PARAM 1/(time2‐time)

the main problem is that i can't find a equivalent command neither in xspice nor in spice3f5.
even if proteus implementation of spice could do it , i have no idea how to implement a measuring chain here.
there is a spice scripting block i can insert , but i don't think it has control over the simulation start itself.
however, by deconstructing the graph itself , it seems to me that even the isis graph is made in the proteus VSM which , i assume, has full control over the simulation flow.
but implementing that is even more a pain in the ass than doing the labview-multisim combination.
how about some rf tools? shouldn't they have such functions to test out UHF-VCOs?
 Logged
borberk
Senior Member

Offline

Posts: 298

Thank You
-Given: 309

 « Reply #4 on: April 16, 2012, 03:15:00 15:15 »

Try to find out what are the possibilities of ADS. I am not used to it.
 Logged
dotm
Active Member

Offline

Posts: 162

Thank You
-Given: 65

C15H15NO2S

 « Reply #5 on: April 16, 2012, 03:29:00 15:29 »

Try to find out what are the possibilities of ADS. I am not used to it.

 Logged
dotm
Active Member

Offline

Posts: 162

Thank You
-Given: 65

C15H15NO2S

 « Reply #6 on: April 23, 2012, 11:16:46 23:16 »

hi.
Finally , i realised this in labview using the multisim connectivity toolkit.
this is a quick and dirty implementation but does the trick.
i had to fetch the data as XY instead of waveform because the toolkit seems to be a little bit incompatible to newer versions of labview.
this is for circuits with single inputs (sources) and outputs (probes) only.
i modified a given colpitts vco with anti-serial bb112s to have a simple vco.
you can load it , test the offset parameters and derive f(v) with "go".
 « Last Edit: April 24, 2012, 05:55:24 05:55 by dotm » Logged
 Pages: [1]