Low resolution waveforms in exported waveform

Hi hbolandi ,

You'll want to decrease or remove waveform compression using .options. From the help:

mike

MStokowski thanks for your response! ".options plotwinsize 0" decompressed data points when I export data as a text file directly from LTspice. However, still number of data points does not match from what I am getting using the PyLTSpice. Here is how I am getting the waveforms using PyLTSpice:  

raw, log = runner.run_now(net, run_filename='./results.net')

raw = RawRead("./results.raw")
t = raw.get_axis()
v = raw.get_trace('V(mid)').get_wave()
I = raw.get_trace('I(Rlow)').get_wave()

Hi hbolandi ,

I can't speak to PyLTspice, since I don't use it, nor do we support it.

Nevertheless, this may be the confusing thing: note that the raw file is not “raw”, like a photo. It's simply a file extension applied to all waveforms, compressed or not.

You need to have a way for PyLTspice to produce an uncompressed waveform, as you would from LTspice proper using.options plotwinsize 0. Is there a way in PyLTspice to pass the .options to your simulation command?

Or, if you go into LTspice > Settings > Compression, and turn off compression, this should change the .ini file, as this setting is persistent, and I believe this will affect your simulations as run by PyLTspice? (no guarantee there):

mike

Hi MStokowski 

Yes, I added .options plotwinsize 0 to the Spice directive, which increased the resolution of the exported data by around 20%. However, as I mentioned in my previous comment, the number of exported data points still does not match what I am getting in the PyLTSpice simulation. I also unchecked the compression setting, but it had almost no effect on the compression.

Hi hbolandi ,

Guess I was wrong about the persistent setting. I'll have to ask around. Sorry, never used PyLTspice, so no idea what compression options it might have in its raw read algorithm. Just to double-check: you have directly compared the results.raw output of LTspice directly to the results.raw file output of PyLTspice and the file sizes are different?

thanks,

mike

Hi hbolandi ,

From PyLTspice site:

sim_batch.py This is a script to launch Spice Simulations. This is useful because:

• Can overcome the limitation of only stepping 3 parameters

• Different types of simulations .TRAN .AC .NOISE can be run in a single batch

• The RAW Files are smaller and easier to treat

From Generative AI (Take with rock-alt sized grain of salt, make that a distill-pond-sized…)

Hi MStokowski 

Thanks for your detailed response. We're running thousands of simulations using PyLTSpice, and this compression is affecting our waveforms, resulting in much lower resolution compared to LTspice. Unfortunately, it seems there is no workaround for this issue.

Hi hbolandi ,

Hmmm, hmmm, hmmm. I feel like the "more efficient" file size of PyLTspice should not be a lossy compression, like LTspice's. So…

We may be barking up the wrong tree looking at waveform compression. My fault; just read your original post again. Most derivative problems, like dv/dt, are best served by lowering the maximum time step in transient analysis. It is possible that you can do achieve you want in PyLTspice by simply forcing a small time step size in .tran. Just shooting in the dark here. To wit, dTmax:

.TRAN <Tstep> <Tstop> [Tstart [dTmax]] [modifiers]

I can also ask our resident command line expert if it is possible to simply build your own batch app in Powershell as opposed to PyLTspice, but it looks like PyLTspice has a useful command structure.

mike