LT SPICE model for MCP602-I/SN ..#2

Bordodynov

Joined May 20, 2015
3,430
.opt cshunt=1p very bad!
The Spice model of the microcircuit contains many resistors. Including high-resistance 100 megaohms. By connecting a 1 pF capacity to them, you will disrupt the normal mode - greatly reduce the performance!
1753420279374.png
 
Last edited:

Thread Starter

OldElectron

Joined Jul 24, 2025
3
I see where that might affect the simulation. But the problem is that LTspice 24.1.9 under Win10 or Linux/wine fails to accept the MCP602 model as valid. It generates several errors when processing the model. It never simulates the circuit. See the error log I posted yesterday (MCP602.txt)

I tried the model and original circuit on an old copy of LTspice 4.22y running under WinXP. It ran correctly. So I think something has changed in LTspice which might cause it to be more strict about interpreting model syntax.
 

Jony130

Joined Feb 17, 2009
5,593
So I think something has changed in LTspice which might cause it to be more strict about interpreting model syntax.
Yes, this is the case.

The recent versions of LTspice for Windows (version 24.1.*) are a significant change from previous versions. Any major program revision such as that can be subject problems, and this was no exception. Analog Devices has worked hard to fix any new bugs. Also, some of LTspice's behavior fundamentally changed, which may cause a few older simulations to work differently or not at all.
Version 24.0.12 was the last version before this major change to LTspice.
From here:
https://groups.io/g/LTspice

This is why I have two LTspices on my computer. The old one LTspiceXVII ver. 17.1.15 and the newest one LTspice ver. 24.1.9.
 

Thread Starter

OldElectron

Joined Jul 24, 2025
3
I just managed to get it working in 24.1.9 by:
1) deleting the spurious parentheses on lines 126 and 126 for G35 and G36 and
2) changing the final line to .ENDS MCP602 in your file MCP602.lib.

Yes, the model parser is more strict now. Thanks for your comments!
 
Top