Strange. Not LTspice related but I would always recommend that anyone always has a recent system image backup to revert to .
I think what I would do here is first backup and move all your own LT files to a flash drive for safety. Uninstall any versions of LTspice via the uninstall option in 'Programs and Features' in control panel. Then delete all entries related to LTspice in the registry.
@Hans Polak also had strange issues (a few posts above this one) and I can only assume there can be some conflicts between older installations that have been updated over months/years.
See post #1 here:
I think what I would do here is first backup and move all your own LT files to a flash drive for safety. Uninstall any versions of LTspice via the uninstall option in 'Programs and Features' in control panel. Then delete all entries related to LTspice in the registry.
@Hans Polak also had strange issues (a few posts above this one) and I can only assume there can be some conflicts between older installations that have been updated over months/years.
See post #1 here:
The good news seems to be that although it uninstalls any existing installation of LTSpice XVII (or 17.1 or later) it seems to leave a lot of stuff behind, it also installs to a different directory and then aborts.. I was able to install 17.1 and it was able to access my modified libraries. Have not tested extensively but seems to be working.
Where it installs depends on the option you choose in the Wizard. Install for Me or Install for Everyone uses the AppData and Program files areas respectively.
I am running a linux system: PCLinuxOS KDE5 Plasma. Have both versions installed XVII and 24 and running them under wine.
When I installed 24 it did not delete my XVII version, thank God for that.
After fiddling around for a while with 24, I now switch back to XVII for more convenience until now.
When I installed 24 it did not delete my XVII version, thank God for that.
After fiddling around for a while with 24, I now switch back to XVII for more convenience until now.
@Ray Waters I was just setting up a differential filter with your OPA1632 model but it doesn't seem to work correctly. The output is just a fraction of a mV and stuck near the pos rail. I know you said you didn't check it but that was a year and a half ago. Any ideas?
Jan
Nothing wrong with the model. Try instead e.g. the filter of figure 6 from the attached report.
One look at that circuit and the light went on - I had no DC path from out to in.
A couple of 1Meg resistors fixed it.
Thanks!
Jan
A couple of 1Meg resistors fixed it.
Thanks!
Jan
I also checked the model and the symbol and they both looked OK, but I'm glad that you got this fixed.
Some time back I posted some templates I'd made for making distortion measurements using LTspice. Last time I looked, seven people had downloaded them and presumably looked at them. That's almost eight!
So, here is a more complete and refined collection. In addition to the usual harmonic distortion test from a single sine wave source, pretty much all of the dual tone, triple tone, and multi-tone tests available in REW's signal generator are included. There's both annotated versions with instructions and notes and streamlined versions that take up less space on an LTspice schematic.
I'd originally intended to include a more wordy notes file, but I realized that much of it had been said before by others and it wouldn't be of much interest anyway. So, you all are saved from that.
So, here is a more complete and refined collection. In addition to the usual harmonic distortion test from a single sine wave source, pretty much all of the dual tone, triple tone, and multi-tone tests available in REW's signal generator are included. There's both annotated versions with instructions and notes and streamlined versions that take up less space on an LTspice schematic.
I'd originally intended to include a more wordy notes file, but I realized that much of it had been said before by others and it wouldn't be of much interest anyway. So, you all are saved from that.
Hello,
Last night I updated to the latest, version 24.1.0 and now every sim files that I'm working with including the shared sim files that I downloaded from the forum no longer works. LTSpice keeps on popping with numerous errors, unknown syntax, no such parameters defined etc. etc. What is going on..😤🤨
Last night I updated to the latest, version 24.1.0 and now every sim files that I'm working with including the shared sim files that I downloaded from the forum no longer works. LTSpice keeps on popping with numerous errors, unknown syntax, no such parameters defined etc. etc. What is going on..😤🤨
OK I solved some issues, it seems the .option directive is no longer recognized, changed it to .param and it works but the Piece Wise Linear Squarewave test (with the value PWL repeat forever ....) cannot be recognized. Is there an update or changes made to the syntax?
Thanks!
Albert
Thanks!
Albert
There is my trick which allows to change up to 8 signal sources with one parameter. This is hack with multiplexer:
Set only .param S=1 to 7 and switch selects inputs from S1 to S7 acordingly.
Works as a charm, switch pass both positive and negative voltage without any amplitude limits.
Parameter 'S' may also be provided as .step
NOTE: switch introduces 4 Ohm series resistance.
Set only .param S=1 to 7 and switch selects inputs from S1 to S7 acordingly.
Works as a charm, switch pass both positive and negative voltage without any amplitude limits.
Parameter 'S' may also be provided as .step
NOTE: switch introduces 4 Ohm series resistance.
Last edited:
The lack of response is certainly not due to lack of interest.
I think it is because many of us (like me) don't understand what you are doing.
Can you explain?
Jan
I think it is because many of us (like me) don't understand what you are doing.
Can you explain?
Jan
I'm using the boolean compare operator '==' all the time which works like in C, equal gives 1 and not equal gives 0. It can be used in any expression, but in the simplest case it is equivalent to the IF operator.
Zbig,
I just really looked at your first idea using an ADGS1408. That works, but has some limitations.
These CMOS switches aren't perfectly linear switches. Depending on what you are doing with your simulation, that could cause a problem. For example, if you want to perform really low level harmonic distortion or IMD simulations, the switch linearity is a limitation. That all depends on the load impedance for the switch, your test frequencies, and so on.
So, you may be better off with your second approach using resistors as switches. The ideal resistors in LTspice are very linear. 🙂
You could also use the ideal switch component available in LTspice to build your own multiplexer.
Voltage Controlled Switches
Or, you could just connect all the voltage sources in series and use KSTR's suggestion. When a voltage source is set for zero Volts output and runs at zero frequency, it just acts as a short.
I was thinking about "building" a kind of universal test signal generator that could be a subassembly or part of a hierarchal arrangement that would allow for all sorts of different tests. Your approach would allow for that. But, then laziness set in and I stuck with different "templates" for each test. Takes less schematic space.
I just really looked at your first idea using an ADGS1408. That works, but has some limitations.
These CMOS switches aren't perfectly linear switches. Depending on what you are doing with your simulation, that could cause a problem. For example, if you want to perform really low level harmonic distortion or IMD simulations, the switch linearity is a limitation. That all depends on the load impedance for the switch, your test frequencies, and so on.
So, you may be better off with your second approach using resistors as switches. The ideal resistors in LTspice are very linear. 🙂
You could also use the ideal switch component available in LTspice to build your own multiplexer.
Voltage Controlled Switches
Or, you could just connect all the voltage sources in series and use KSTR's suggestion. When a voltage source is set for zero Volts output and runs at zero frequency, it just acts as a short.
I was thinking about "building" a kind of universal test signal generator that could be a subassembly or part of a hierarchal arrangement that would allow for all sorts of different tests. Your approach would allow for that. But, then laziness set in and I stuck with different "templates" for each test. Takes less schematic space.