Circuit Simulation Topics using LTSpice

[Aziz]

Hi friends,

I am totally infected with LTSpice! Analog and digital systems can be modelled down to physical behavior of a circuit or component. I will post my results and experiments here to avoid beeing off-topic to other threads.

To contribute to circuit simulations, please download the free software and documentation from Linear Technology:
LTSpice/SwitcherCAD III
http://www.linear.com/designtools/software/

Following link is also useful (requires membership registration):
http://tech.groups.yahoo.com/group/LTspice/
Lots of spice models for LTSpice out there.

I will discuss some circuit topics based on VLF, PI and other circuits related on metal detection. You are welcome to contribute.

BTW, I found a way, where the target object can be modelled into the active circuit!!!

Let's start.
Aziz

[simonbaker]

Thanks for ref!

[amtech2005]

Circuit Simulation Topics using LTSPICE

AZIZ , Please let us know more on the LTSPICE "free-software" !! Any LIMITS on number of components??.........Thanks.........Eugene

[Tinkerer]

TINA-TI for spice simulation

LT-Spice is nice, but I also like TINA-TI

free download and indefinite use

http://dl-www.ti.com/lit/sw/sloc058d/sloc058d.zip

Tinkerer

[Aziz]

Hi friends,

I have learned much last days. I was beeing confronted with some limitations of LTspice. Particularly, some parts are not available as SPICE models. Anyway, they would not also be available to other SPICE programs.

Sometimes, the solution matrix gets singular or other error messages will occur. By changing some capacitor or resistor values or options solves this.

I like the noise and other analysing tools of the LTspice. Also the possibility of placing a real target (=complex coil circuit) into the circuit simulations is phantastic. This option can be applied to VLF and PI. VLF IB coil, DD coil, coupled coils are not more difficult to simulate accurate. I can use my coil software to get the right coupling factor k.

I have analysed my discrete low-drop-output (LDO) voltage regulator (with 0.2 mV drop-out voltage!). Also some cleaning circuits of power supply noise for very low noise analog parts (pre-amp, LC oscillator, PLL, VCO, etc.).

Now I will sit more hours and days before I turn my soldering iron on. Testing the circuit is a very powerful tool. So I have to investigate the huge analysing possibilities in the coming weeks.

There are some very critical modeling parts:
- power supply (bypassing caps, resistance, load-ripple problems...)
- coil parasitic parameters (Rser, Rpar, Cpar..)
- capacitor parasitic parameters.. (for very critical parts of circuit).
One should also take these options into account to get a more accurate simulation results.

I am sure, I can speed up my hardware development time and be able to optimize some critical circuit parts of MD.

So give me more time to explore the facilities before I put some interesting models.

Aziz

[Aziz]

LT-Spice is nice, but I also like TINA-TI

free download and indefinite use

http://dl-www.ti.com/lit/sw/sloc058d/sloc058d.zip

Tinkerer

Hi Tinkerer,

I tried the Tina from TI. It contains lots of interesting spice part models. But I let the software crashed many times. Maybe I am a stupid or strange user which stresses the software too much.

LTspice is working much reliable and is easy to use.

Aziz

[Aziz]

Be sick and getting tired of missing part model?
Suck this:
http://www.elektronikschule.de/~krau...ice-Models.zip
(59 MB)

and this:
http://www.elektronikschule.de/~krau.../CD_5spice.zip
(68 MB)

The symbols must be probably checked due to port numbers. Missing symbols must be created and in case of missing, they can be done easily in LTspice. Models contain not symbolic definitions of a part. Libraries contain more models.


Aziz

[ivconic]

Aziz, a friend send me some question by mail;

"Skinuo sam simulacioni program SwitcherCAD III,probao da uradim simulaciju oscilatora za MRH detektor i nisam uspeo.Tačno meri jednosmerne napone ali AC vrednosti neće da simulira,AC napon na kalemu naprimer.Molim za pomoć.

Unapred hvala"

Meaning;
I downloaded SwitcherCad III,tried to simulate MRH (Relic Hawk) oscillator without success. It is accurate in measuring DC voltages but AC values wont simulate, AC voltage at coil, fer instance!? Please help!

Many thanks in advance


This question is addressed to you Aziz. Friend do not manage English language well, so asked me to put this question here,for you.

So i guess, it would be nice you to give some hints here, step by step, how to simulate oscillator part with coil connected..
This is also interesting subject for others, me too, although i am not pretty tied to simulators, generally.
Many thanks in advance Aziz!
Best regards!

[Max]

Aziz, a friend send me some question by mail;

"Skinuo sam simulacioni program SwitcherCAD III,probao da uradim simulaciju oscilatora za MRH detektor i nisam uspeo.Tačno meri jednosmerne napone ali AC vrednosti neće da simulira,AC napon na kalemu naprimer.Molim za pomoć.

Unapred hvala"

Meaning;
I downloaded SwitcherCad III,tried to simulate MRH (Relic Hawk) oscillator without success. It is accurate in measuring DC voltages but AC values wont simulate, AC voltage at coil, fer instance!? Please help!

Many thanks in advance


This question is addressed to you Aziz. Friend do not manage English language well, so asked me to put this question here,for you.

So i guess, it would be nice you to give some hints here, step by step, how to simulate oscillator part with coil connected..
This is also interesting subject for others, me too, although i am not pretty tied to simulators, generally.
Many thanks in advance Aziz!
Best regards!

Hi,
I've simulated RH oscillator in LTspice... and works. Maybe your friend need to set "startup" clause in the analisys params... otherwise no oscillations will appear.

In real world some noise make oscillations start in the circuit: you have to simulate a transient in spice too to get initial imbalance that will start oscillations... the trick is starting with raising voltage from 0, the transient will be enough to init the process.

Kind regards,
Max

[ivconic]

Many thanks Max!
I will send him answer.
So it seems he didnt adjusted some param's in "setup"?
Regards!

[Aziz]

Hi ivconic,

I had also problems in simulating oscillator circuits in the beginning. Until I found the following initial settings shown in the attachement. This will work in many cases. The checkbox on "Skip Initial operation point solution" should also be set. As Max mentioned, the "startup" option is also very important.

I will put an another oscillator here (I have it simulated already) for experiments. It is a Hartley-Oscillator. Remove the ".txt" extension and load into LTspice.

Regards,
Aziz

[Max]

Many thanks Max!
I will send him answer.
So it seems he didnt adjusted some param's in "setup"?
Regards!

Like Aziz reported... when he will check the first checkbox , from top, the software will add the label "startup" as parameter ...

One could also add by hand... just right clicking at the analisys tag and adding that param, same stuff.

Once he will set it... the voltage will be supposed starting from 0 volts... thus giving enough imbalance to see oscillations appears. If one start spice considering a steady state initial condition is , instead, likely an ideal equilibrium is already enstablished there, and no perturbations: this prevent the system to oscillate and your friend will see only steady state e.g. voltages of an ideal system really far from reality of built circuits... where noise exist... and you switch on oscillator with voltage then raising from 0v to e.g. 8V for real.

Kind regards,
Max

[ivconic]

Aziz,Max thank you very much!
I apreciate this!
Best regards!

[simonbaker]

Aziz or anyone:

Do you have an LTSpice circuit of the TGS oscillator? If so, would you post it?

I built a stripped-down version of the TGS oscillator, leaving out the JFet gain control path on a breadboard. It oscillates fine (I used a generic PNP small signal transistor).

I made an LTSpice circuit of the same circuit (using the 2N2907 transistor).
See attached file testTGSosc.zip. It will not sustain oscillation in the simulation, slowly dies out over several seconds.

I can make it oscillate by adding another capacitor and resistor to the feedback path. See attached file testTGSoscMod.zip.

I would be interested in seeing your LTSpice circuit of the oscillator to understand why my stripped down oscillator does not sustain in the simulation, but does oscillate fine when I build the actual circuit.

Thanks

-SB

[simonbaker]

Sorry, disregard previous post, dumb connection error on my part. Here is new corrected schematic, it oscillates, although at about 13 kHz. I'll keep checking to fine tune reason frequency different from real life, but not too far off.

Regards

-SB