So how do I import a model to LTSpice? I have followed the explanations on the web and they fail to work.
Using this file, can you guide me through it?
I had a quick look at your AD745 file and couldn't get it run although it was 'accepted' as a valid model file.
First try following this (post #146):
Installing and using LTspice IV (now including LTXVII). From beginner to advanced.
You will note your .txt file has strange pinouts allocated, change them to match the one in my example which is 1, 2, 3, 4 and 5 and which match the generic opamp model.
After all that, it just wouldn't simulate
This was in LTXVII
First try following this (post #146):
Installing and using LTspice IV (now including LTXVII). From beginner to advanced.
You will note your .txt file has strange pinouts allocated, change them to match the one in my example which is 1, 2, 3, 4 and 5 and which match the generic opamp model.
After all that, it just wouldn't simulate
This was in LTXVIIAttachments
You will note your .txt file has strange pinouts allocated, change them to match the one in my example which is 1, 2, 3, 4 and 5 and which match the generic opamp model.
After all that, it just wouldn't simulate
You can't do that, spice netlist is not valid any more. To make it work you will have to change all 3->1, 2 stays as it is, 99->3, 50->4 and 37->5 in model netlist...I would say not practical and useless.
Steps to ad new .subckt "model you are interested in" to sub file:
- ran Ltspice as administrator
- open lib->sub (Files of type: All files (*.*)
- copy the first file in the list (whatever it is) with .sub extension
- up one level to lib files (you will see cmp, sub, sym listed)
- paste copied file to sub and open sub
- you will see "whatever it is" - Copy.sub on top of the list
- rightclick on that file and rename it to AD745J.sub for example
- place opamp symbol on schematic and rename ti to AD745J
- place .inc AD745J.sub as spice directive on schematic
- you are good to go
- ran Ltspice as administrator
- open lib->sub (Files of type: All files (*.*)
- copy the first file in the list (whatever it is) with .sub extension
- up one level to lib files (you will see cmp, sub, sym listed)
- paste copied file to sub and open sub
- you will see "whatever it is" - Copy.sub on top of the list
- rightclick on that file and rename it to AD745J.sub for example
- place opamp symbol on schematic and rename ti to AD745J
- place .inc AD745J.sub as spice directive on schematic
- you are good to go
Followed the steps. Made a new AD745J.sub to the sub directory.
Then placed an opamp on the circuit, which I renamed AD745J, with the .inc AD745J.sub loaded.
What I get is this message.
Now another question. How I get to load and use the model without using the .inc command, like I do with transistors and diodes?
Then placed an opamp on the circuit, which I renamed AD745J, with the .inc AD745J.sub loaded.
What I get is this message.
Now another question. How I get to load and use the model without using the .inc command, like I do with transistors and diodes?
Attachments
I missed one step, after you renamed a file to AD745J.sub, open it and replace existing spice model with AD745J spice model and save it in sub folder. It should work now, my bed...Now I have to take a nape...yes I am old enough to do that
I had done that, with AD745J 1 2 3 4 5 for the pins. Got that message.
Now I changed that to:
* Node assignments
* non-inverting input
* | inverting input
* | | positive supply
* | | | negative supply
* | | | | output
* | | | | |
.SUBCKT AD745J 3 2 99 50 37
And got an error mismatch between AD745J and instance xu7. What is that instance xu7?
I may have missed stuff but I do it way, way simpler with no directory and file moving.
Example AD745. When I download it from the ADI website, it is automatically opened by LTspice. The lines have colors depending on their function.
I highlight this line:
.SUBCKT AD745 3 2 99 50 37
Then I click right mouse and select 'create symbol', answer yes.
You will see a rectangle with the model node pins: 3, 2, 99, 50, 37.
I right-click on each pin and rename them according to the model file: 3=+in, 2=-in, 99=Vc etc.
If you want to go fancy, move the pins around so that the output is at the right, the inputs on the left, Vc top right, Ve bot right, but that's just cosmetics.
Save the symbol, save the model file.
You can now call the symbol in the normal way, from the 'autogenerated'
set. Depending on where you saved the model file you may need to include .inc {model file name you used to save it under} which in this case would be AD745.cir so that LTspice can find the model the symbol points to.
Done.
Jan
Example AD745. When I download it from the ADI website, it is automatically opened by LTspice. The lines have colors depending on their function.
I highlight this line:
.SUBCKT AD745 3 2 99 50 37
Then I click right mouse and select 'create symbol', answer yes.
You will see a rectangle with the model node pins: 3, 2, 99, 50, 37.
I right-click on each pin and rename them according to the model file: 3=+in, 2=-in, 99=Vc etc.
If you want to go fancy, move the pins around so that the output is at the right, the inputs on the left, Vc top right, Ve bot right, but that's just cosmetics.
Save the symbol, save the model file.
You can now call the symbol in the normal way, from the 'autogenerated'
set. Depending on where you saved the model file you may need to include .inc {model file name you used to save it under} which in this case would be AD745.cir so that LTspice can find the model the symbol points to.
Done.
Jan
BTW, as you are certainly familiar with Gary Galo's mod of Adcom's RIAA preamp, changing the original first stage chip for an AD745, let me ask you this. I would ask Gary Galo or Walt Jung, but I'm not sure they are around.
As I was simulating the original LT schematic used by Adcom in LTSpice, with superb THD results, I wanted to see how things changed replacing the chip.
And the THD results worstened considerably. So I wondered if something was wrong with my AD745J model, or if my simulation was correct.
As not rarely a worst THD does not say it all audibly speaking, maybe this was just one of those cases.
As I was simulating the original LT schematic used by Adcom in LTSpice, with superb THD results, I wanted to see how things changed replacing the chip.
And the THD results worstened considerably. So I wondered if something was wrong with my AD745J model, or if my simulation was correct.
As not rarely a worst THD does not say it all audibly speaking, maybe this was just one of those cases.
Last edited:
I saved the .asc and the opamp model configuration, closed the file and then reopened it and it wouldn't run
You really can't conclude anything from distortion measurements in LTspice. The models provided are generally so-called macro-models. They do a reasonable job of modeling the ac and tran behaviour but for distortion all bets are off unless you are absolutely sure that the non-linear behaviour is accurately modeled, which normally it isn't.
Discrete devices are better in this respect and the simulated distortion of a discrete circuit can be pretty accurate.
Jan
Thanks, Jan. It sounds quite logical. I will assemble several preamps, including that one, and do a serious listening.
Ther's one discrete "affordable" preamp thread that I did some simulations for, that I would appreciate your input.
Same for everybody, of course.
https://www.diyaudio.com/forums/analogue-source/91497-bc550-bc560-low-noise-riaa-21.html#post5679913
I got very promising THD results by increasing the voltage to +/-30v and trying different KSA/KSC complementaries. I would like to simulate other things besides THD and FR.
All these preamps, when built. I plan to try with superegulators.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.