Installing and using LTspice IV (now including LTXVII), From beginner to advanced

Jump to Latest
#3,363
If you use parts selected from the original LTspice install, that'll work for anyone.
If you use 3rd party parts you import yourself like the pass transistor and the zeners, than you need to do either of two things if you want anyone to run it.
1 - place those non-LTspice model files on the schematic, or 2 - include their .lib model files with the schematic in the upload. Or zip it all and post that.

Jan
 
#3,364
jan.didden said:
If you use parts selected from the original LTspice install, that'll work for anyone.
If you use 3rd party parts you import yourself like the pass transistor and the zeners, than you need to do either of two things if you want anyone to run it.
1 - place those non-LTspice model files on the schematic, or 2 - include their .lib model files with the schematic in the upload. Or zip it all and post that.

Jan
Click to expand...
Ok, I will upload my files zipped, including txt model files in it. Now the Spice directive can be .inc or .lib.
I am more familiar with the .inc directive, so that's the one I will use. Would that be fine?
 
#3,365
Hi all.


Do someone know where to get ultrafast rectifier diodes? I have some but when using them the LLC smps do not work right, it does only for now with schottky diodes. Ultrafast recovery diodes I have also, smae problem, the output voltage is just half and feedback does not work.

THanks in advance for help finding right diode models.
 

Attachments

  • schottky-works.jpg
    schottky-works.jpg
    343.1 KB · Views: 113
#3,366
I found this in one of my directories, it was datestamped 2019. I have no idea how (in)accurate this simulated diode is, compared to the real thing in the real world. Caveat Emptor.

Code: 
.MODEL UF4007 D ikf=0.15 N=3.97671 IS=3.28772u RS=0.149734
+  EG=1.11 XTI=3 CJO=2.92655E-011 VJ=0.851862 M=0.334552
+  FC=0.5 TT=1.84973E-007 BV=1000 IBV=0.2 Iave=1 Vpk=1000
+  type=silicon

You will note that it is similar to this model found on Vishay's website.
 
Last edited:
#3,367
Mark Johnson said:
I found this in one of my directories, it was datestamped 2019. I have no idea how (in)accurate this simulated diode is, compared to the real thing in the real world. Caveat Emptor.

Code: 
.MODEL UF4007 D ikf=0.15 N=3.97671 IS=3.28772u RS=0.149734
+  EG=1.11 XTI=3 CJO=2.92655E-011 VJ=0.851862 M=0.334552
+  FC=0.5 TT=1.84973E-007 BV=1000 IBV=0.2 Iave=1 Vpk=1000
+  type=silicon

You will note that it is similar to this model found on Vishay's website.
Click to expand...
Thanks for the supply. I have also found a bunch of these diodes in the standart.dio but some do not work well, I get much to low output voltage. Yours do well. Thanks.
 
#3,370
The built-in opamp2.asy symbol can be used for most LM4562 spice models as long as the model matches the usual pin order for op amps -- there really is no need to create a special symbol file for the LM4562.

Attached is the original National Semiconductor model for the LM4562. I've found this one to be better than later ones issued by Texas Instruments, which acquired National Semiconductor back in 2011.
 

Attachments

#3,371
I just realized that I uploaded a version of the National Semiconductor model that I had modified for my own use. The only difference from the original is the renaming of the nodes for the external pins in the model itself, to match the "1 2 3 4 5" order in the model definition. This was purely an aesthetic thing (yeah, I'm like that sometimes).

Attached is the true original (unmodified) National Semiconductor LM4562 model. Sorry for the confusion.
 

Attachments

#3,372
Simplest of circuits, and I can't get the correct answers in simulation.

Domestic mains supply, set at 240 volts rms. Typical resistance of a 100A fuse is around 0.00073 ohms. It all fell apart when I tried adding a similar low value as supply cable resistance.

I can not get the correct voltages across the two low value resistors. They are in series and I get different voltages across each. Have tried running the sim with a correct timestep like we would with audio circuits and I still get crazy results. 4 volts rms across the first 0.00073 ohm is 5.5kA

The second image is set as a DC sim and the results are correct.

What do I need to do to get it to calculate correctly?


Untitled.jpg



Screenshot 2022-12-11 091814.jpg
 

Attachments

#3,379
As originally set up the error for the voltage across R1 is 8411% in error. Using a calculator I get 0.04864 vrms across R1 and the same across the fuse. LTspice gets 4.14

Edit... 4.14 across R1. LTspice gets a totally different result for the fuse which is odd as it is a series circuit.
 
#3,380
Hmm, I calculate V(node1) = 341/sqrt(2) = 241.123V whereas image1 of #3372 says V(node1) = 239.52V ; giving an LTSPICE error of -0.664%

Then I calculate the resistance between node3 and GND is 3.61708 ohms. Which means that V1 drives a load resistance of 3.61854 ohms, so the voltage at node2 is (341/sqrt(2)) * (1 - (0.00073 / 3.61854)) , namely , 241.075V

whereas image1 of #3372 says V(node2) = 235.38V ; giving an LTSPICE error of -2.362% . That's unacceptably high so you probably want to increase the number of datapoints LTSPICE uses to calculate its RMS average. .OPTIONS MAXSTEP is one easy way.
 
  • Thank You
Reactions: Mooly
You must log in or register to reply here.
Top Bottom