开云体育

OPA1642 SPICE Models

#dualopaalice

matthewsewell All Messages By This Member #37139 I'm having fun with variations on the OPA Alice circuit and would like to be able to model using SPICE in KiCAD. I found the model on TI's website but it doesn't have the correct subcircuit for a 1642, which has 8 pins. I modified that lib with the following: ? ``` * Added by matthew to model the 8-pin OPA1642AID .SUBCKT OPA1642AID OUT_A IN-_A IN+_A VEE IN+_B IN-_B OUT_B VCC XU1 IN+_A IN-_A VCC VEE OUT_A OPA164x XU2 IN+_B IN-_B VCC VEE OUT_B OPA164x .ENDS OPA1642AID ``` I then made sure that the pins match the symbol in my schematic, but when I tried to run an OP test, I got a bunch of warning and errors. Before I go down the path of trying to trouble-shoot all of this, does anyone have any experience with this who can offer any advice for modeling the OPA Alice circuit?? PS. I'm very new to electronics. I am a software engineer and have worked adjacent to and have even manage electrical engineers but this is not my area of expertise. ? ?
Mark Kahrs All Messages By This Member #37140 Hold on, did you also define the SUBCKT for OPA164x???? I presume you did since this is what TI provides.? But I want to make sure. Basically, what you've done is to "call" two versions of the OPA164x subcircuit.? I used the word "call" in quotes since it's not a true procedure call. Further, do you know that this model works in Gnu SPICE?? (I am presuming that's what KiCAD uses). toggle quoted message Show quoted text On Wed, Feb 5, 2025 at 8:42?PM matthew via <matthew=[email protected]> wrote: I'm having fun with variations on the OPA Alice circuit and would like to be able to model using SPICE in KiCAD. I found the model on TI's website but it doesn't have the correct subcircuit for a 1642, which has 8 pins. I modified that lib with the following: ? ``` * Added by matthew to model the 8-pin OPA1642AID .SUBCKT OPA1642AID OUT_A IN-_A IN+_A VEE IN+_B IN-_B OUT_B VCC XU1 IN+_A IN-_A VCC VEE OUT_A OPA164x XU2 IN+_B IN-_B VCC VEE OUT_B OPA164x .ENDS OPA1642AID ``` I then made sure that the pins match the symbol in my schematic, but when I tried to run an OP test, I got a bunch of warning and errors. Before I go down the path of trying to trouble-shoot all of this, does anyone have any experience with this who can offer any advice for modeling the OPA Alice circuit?? PS. I'm very new to electronics. I am a software engineer and have worked adjacent to and have even manage electrical engineers but this is not my area of expertise. ? ?
matthewsewell All Messages By This Member #37142 On Wed, Feb 5, 2025 at 05:54 PM, Mark Kahrs wrote: Hold on, did you also define the SUBCKT for OPA164x???? I presume you did since this is what TI provides.? But I want to make sure. ? I literally took the .lib file provided by TI, which defines the SUBCKT and appended my SBCKT to it. ? ? Basically, what you've done is to "call" two versions of the OPA164x subcircuit.? I used the word "call" in quotes since it's not a true procedure call. ? ? Yes. My intention was to account for all the pins with the parameters already defined in the .lib file. I am new to this, so I'm not sure that was the right way to go about it. ? Further, do you know that this model works in Gnu SPICE?? (I am presuming that's what KiCAD uses). ? ? I'm running with PSpice and LTSpice compatibility modes and it seems to be able to interpret everything correctly. Again, I'm very new to this. If it helps at all, the output looks like this: Note: Codel model file loading path is /home/matthew/workspace/Mic OPA/ Note: Compatibility modes selected: ps lt a Circuit: KiCad schematic Reducing trtol to 1 for xspice 'A' devices Doing analysis at TEMP = 27.000000 and TNOM = 27.000000 Using SPARSE 1.3 as Direct Linear Solver Note: Starting dynamic gmin stepping Warning: Dynamic gmin stepping failed Note: Starting true gmin stepping Warning: True gmin stepping failed Note: Starting source stepping Warning: source stepping failed Note: Transient op started Warning: singular matrix: ?check nodes xic1.xu1.n3294630 and c1:probe_int_n2#branch Warning: singular matrix: ?check nodes xic1.xu1.n2991579 and xic1.xu2.n2991581 Warning: singular matrix: ?check node xic1.xu1.n2991581 Warning: singular matrix: ?check nodes probe_int_probe_int_12v_c1_c1_2 and xic1.xu2.n2991579 Warning: singular matrix: ?check nodes xic1.xu1.n4619517 and xic1.xu2.n4619517 Warning: singular matrix: ?check nodes xic1.xu2.n3828449 and xic1.xu1.n2991581 Error: -4.91074e+150, -4.23671e+158 out of range for * in line bprobe_int_xic1power Error: -4.91074e+150, -4.23671e+158 out of range for * in line bprobe_int_xic1power Error: -4.91074e+150, -4.23671e+158 out of range for * in line bprobe_int_xic1power Error: -4.91074e+150, -4.23671e+158 out of range for * in line bprobe_int_xic1power Error: -4.91074e+150, -4.23671e+158 out of range for * ... ? From there, it has a bunch of the same kind of errors, then some other lines and then eventually the unhelpful: doAnalyses: OP: ?Timestep too small; cause unrecorded. run simulation(s) aborted
Jerry Lee Marcel All Messages By This Member #37144 开云体育 Usually multiple package opamps are treated as individuals, each section being the typical +in, -in, out, V+ and V-. It does not represent the actual topology, but for simulation purposes, it does the jow fine. Le 06/02/2025 à 02:31, matthew@... a écrit?: toggle quoted message Show quoted text I'm having fun with variations on the OPA Alice circuit and would like to be able to model using SPICE in KiCAD. I found the model on TI's website but it doesn't have the correct subcircuit for a 1642, which has 8 pins. I modified that lib with the following: ? ``` * Added by matthew to model the 8-pin OPA1642AID .SUBCKT OPA1642AID OUT_A IN-_A IN+_A VEE IN+_B IN-_B OUT_B VCC XU1 IN+_A IN-_A VCC VEE OUT_A OPA164x XU2 IN+_B IN-_B VCC VEE OUT_B OPA164x .ENDS OPA1642AID ``` I then made sure that the pins match the symbol in my schematic, but when I tried to run an OP test, I got a bunch of warning and errors. Before I go down the path of trying to trouble-shoot all of this, does anyone have any experience with this who can offer any advice for modeling the OPA Alice circuit?? PS. I'm very new to electronics. I am a software engineer and have worked adjacent to and have even manage electrical engineers but this is not my area of expertise. ? ?
Mark Kahrs All Messages By This Member #37149 You're having convergence issues.? Here's a PSPICE document that covers possible problems. toggle quoted message Show quoted text On Thu, Feb 6, 2025 at 2:48?AM Jerry Lee Marcel via <jerryleemarcel=[email protected]> wrote: Usually multiple package opamps are treated as individuals, each section being the typical +in, -in, out, V+ and V-. It does not represent the actual topology, but for simulation purposes, it does the jow fine. Le 06/02/2025 à 02:31, matthew@... a écrit?: I'm having fun with variations on the OPA Alice circuit and would like to be able to model using SPICE in KiCAD. I found the model on TI's website but it doesn't have the correct subcircuit for a 1642, which has 8 pins. I modified that lib with the following: ? ``` * Added by matthew to model the 8-pin OPA1642AID .SUBCKT OPA1642AID OUT_A IN-_A IN+_A VEE IN+_B IN-_B OUT_B VCC XU1 IN+_A IN-_A VCC VEE OUT_A OPA164x XU2 IN+_B IN-_B VCC VEE OUT_B OPA164x .ENDS OPA1642AID ``` I then made sure that the pins match the symbol in my schematic, but when I tried to run an OP test, I got a bunch of warning and errors. Before I go down the path of trying to trouble-shoot all of this, does anyone have any experience with this who can offer any advice for modeling the OPA Alice circuit?? PS. I'm very new to electronics. I am a software engineer and have worked adjacent to and have even manage electrical engineers but this is not my area of expertise. ? ?
Stephen Tremblay All Messages By This Member #37150 On Thu, Feb 6, 2025, 10:43 AM Mark Kahrs via <mark.kahrs=[email protected]> wrote: You're having convergence issues.? Here's a PSPICE document that covers possible problems. On Thu, Feb 6, 2025 at 2:48?AM Jerry Lee Marcel via <jerryleemarcel=[email protected]> wrote: Usually multiple package opamps are treated as individuals, each section being the typical +in, -in, out, V+ and V-. It does not represent the actual topology, but for simulation purposes, it does the jow fine. Le 06/02/2025 à 02:31, matthew@... a écrit?: I'm having fun with variations on the OPA Alice circuit and would like to be able to model using SPICE in KiCAD. I found the model on TI's website but it doesn't have the correct subcircuit for a 1642, which has 8 pins. I modified that lib with the following: ? ``` * Added by matthew to model the 8-pin OPA1642AID .SUBCKT OPA1642AID OUT_A IN-_A IN+_A VEE IN+_B IN-_B OUT_B VCC XU1 IN+_A IN-_A VCC VEE OUT_A OPA164x XU2 IN+_B IN-_B VCC VEE OUT_B OPA164x .ENDS OPA1642AID ``` I then made sure that the pins match the symbol in my schematic, but when I tried to run an OP test, I got a bunch of warning and errors. Before I go down the path of trying to trouble-shoot all of this, does anyone have any experience with this who can offer any advice for modeling the OPA Alice circuit?? PS. I'm very new to electronics. I am a software engineer and have worked adjacent to and have even manage electrical engineers but this is not my area of expertise. ? ? PSpice-ConvergenceOnly-12.pdf PSpice-ConvergenceOnly-12.pdf
matthewsewell All Messages By This Member #37157 Thanks for this info, everyone.

Previous Topic Next Topic