Maker Pro
Maker Pro

LTspice, Warning node is floating

Vinasse

Oct 14, 2015
11
Joined
Oct 14, 2015
Messages
11
The VIPER16 model is encrypted. You will have to talk to the model supplier to work out why its not working, unless they will provide you an unencrypted model file. Then we might be able to help.

eT

Yes, the encrypted file, I was hoping that we could have solved this without this document.

I really don't think they will give it to me and even if they agree to send it to me, I don't think that I would be allowed to post it on the web...

No way we could solve this without the encrypted file ?
 
Last edited:

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,747
Joined
Nov 17, 2011
Messages
13,747
Ok thanks for the prefix, I will change them, and yes the V104 has no value, I am still waiting for a model from ST Micro. I can't find it on the web (STTH108).
Before you start worrying about the right models, try to get the simulation up and running using standard models from the default library. Thus you can eliminate any wiring errors and later refine the simulation with the correct models.
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Got it working but it takes ages to simulate. You had a label for 12 Volts but didn't have a voltage source :) Delete the 12V label from the output and add a 12 Volts source to your switching regulator.
Thanks
Adam
 

Vinasse

Oct 14, 2015
11
Joined
Oct 14, 2015
Messages
11

Good morning Adam, and thanks a lot.

But I still have questions, This 12V was supposed to be generated by the secondary. The Viper16 has an internal start-up using the drain current and when the output 12V is up and running, the Viper16 should switch to this "secondary supply". So my question is, putting this 12V power source connected to the Viper16l, isn't that cheating a little bit ?

Another one, I am trying to simulate the design using your solution, but I think I did something wrong when I imported the diode models. So I know I should start with standard models as Harald said but anyway I would like to know how to import properly some components for the future. So when I try to simulate the design an error appears "Unknown subcircuit called in" for the first diode S1M V101. Except that the model exists in the "standard.dio" file present in the lib\cmp directory. And the attribute is X.

Do you have an idea, did i forgot to do something during the import ?

Thanks a lot for every thing you are doing !

[Harald Kapp] Attachments removed upon request by op.
 
Last edited by a moderator:

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178

eetech00

Nov 17, 2014
95
Joined
Nov 17, 2014
Messages
95
Hi Vinasse

The Viper needs feedback from the output to provide regulation so the 12v voltage source doesn't really belong there.
There is something not right with the Viper spice model.

Also, If you want the AC supply voltage to output 230v AC (RMS) set the value to 325v.
To verify setting, Control-Left-clk the label in the waveform viewer to see RMS value.
 

eetech00

Nov 17, 2014
95
Joined
Nov 17, 2014
Messages
95
Good morning Adam, and thanks a lot.

Another one, I am trying to simulate the design using your solution, but I think I did something wrong when I imported the diode models. So I know I should start with standard models as Harald said but anyway I would like to know how to import properly some components for the future. So when I try to simulate the design an error appears "Unknown subcircuit called in" for the first diode S1M V101. Except that the model exists in the "standard.dio" file present in the lib\cmp directory. And the attribute is X.

Thanks a lot for every thing you are doing !

Like I said in my previous post, you don't need to change the prefix attribute if you've used "model" statements to define the device. Model statements are contained in the standard.dio file, so you should be able to just select a diode from the component menu,place it on the schematic, change the "value" field to the device you want, then simulate. If you've used "subckt" statements to define the device, then you need to change the symbol prefix attribute to "X". In either case, if the device is not in the default LTspice supplied libraries (and in their default folder locations), you need to place an include statement on the schematic to tell LTspice where to find the device definition.

Hope that helps...
 

Vinasse

Oct 14, 2015
11
Joined
Oct 14, 2015
Messages
11
Hello Arouse1973 and eetech00,

Thanks a lot for your answers, so yes eetech00 you are right, the attribute should be D and not X when using "model" statements in the standard.dio.

So I tried to simulate with the 12V voltage source and I stil get an error "Time step too small" and the warnings "Node is floating".

So I think that there is definitivly something wrong with the Viper16 spice model, I will try to ask ST but they are not really responsive...

Now I also tried without the external source 12V. It looks like the simulation is not converging, And I think it is linked to the Viper spice model. Because I also get the warning "Node is floating".

So now I guess I am waiting for ST answers...

Thanks a lot to all of you. I didn't know that so many people were willing to help a total stranger like this, make me smile.

Have a nice day !
 
Last edited by a moderator:
Top