Struggling with .meas in LTSpice

[bugbear]

I was setting out to look at PSRR in my Rotel 850 sim.


My initial plan was to simply measure the output with various supply rails.

I was then going to move on to automate this as per

LTspice: Using .MEAS and .STEP Commands to Calculate Efficiency | Analog Devices

But I have fallen at the first hurdle. I added a simple .meas directive to the simulation, via the .meas editor.

But I get no result in the log file.

(I also note a LOT of warnings from the model file)

BugBear

 

[bugbear]

Got it - the second parameter is the type of simulation the "measure" corresponds to.


BugBear

 

[abraxalito]

For classAB amp I have found its important to sim PSRR when signal is present and with the output loaded.

 

[bugbear]

I'm planning to move on to checking what happens when I simulate a real PSU that drops under load (as opposed to a LTSpice voltage source that doesn't), and (finally) what happens to overall THD when there "just happens" to be a high amplitude, low frequency component in the input.

(AKA what happens to the violin sound when the organs kicks in for the finale!?)


BugBear

 

[abraxalito]

With LTS you'll not need to rely on THD, you can do nulling to get more visibility of what's going on with real-world signals (which you can inject via a .wav file). At least that's what I did when working on a headamp sim.


I shall be watching with interest to see how those violins fare....😀

 

[bugbear]

Well, I finally managed to set up what I think is an OK test.


I arranged that I could (optionally) inject 1V of ripple onto either the front end OR output driver stages of the Rotel 850, with or without a signal.


The drivers stage has massive PSRR, so I concentrated on the front end (which is where high grade power amplifiers commonly have some exotic smoothing/regulation).


1V of signal looks like this - some harmonic distortion, but waddya' expect from a mid range amp?

(anything below 120dB and above 100kHz is probably an FFT artefact)

BugBear

 

[bugbear]

Checking for PSRR, I added 1V ripple at 1.3Khz to the front end power rail. Not too bad, down at -50 or so dB.


BugBear

 

[bugbear]

So, if we have a ripple-y front end power rail, what does it do to the signal?

OOps. Wasn't expecting that! Some kind of inter modulation?



BugBear


My struggles with FFT measurement are in this other thread:

FFT mystery?