no mercy simulator killer circuit

[Bernhard]

I was thinking in the first post with less than -200dB numbers for distortion.
From the APx datasheet:

SYSTEM PERFORMANCE
Residual THD+N (22kHz BW)
-117 DB + 1.0 μV
Typical < -120 DB (1 KHZ, 2.0 V

< -200dB is THD, not THD+N

 

[Bernhard]

Boo-hoo! My stimulator will only do -170dB!
View attachment 1137336

In your picture it is -107.9588 ? That would be pretty useless.
Microcap is a free program, you can download

 

[MarcelvdG]

What happens when you make the steady state tolerance 10 times smaller? (Apparently it's some kind of steady state analysis, I never knew there were free simulators supporting that.)

 

[Bernhard]

Tried set steady state tolerance to zero, does not help.

 

[Bernhard]

Burr Brown OPA627, noninverting buffer, gain of 2, Vin max. 2V

 

[MarcelvdG]

A zero tolerance makes no sense, maybe it uses the default tolerance whenever the tolerance is set to a value that's <= 0, instead of giving a normal error message.

 

[Bernhard]

Microcap includes a few amplifiers, for example the Leach amp
Same settings

 

[Bernhard]

A zero tolerance makes no sense, maybe it uses the default tolerance whenever the tolerance is set to a value that's <= 0, instead of giving a normal error message.

It is suggested in the help menue to set it zero, I also tried just to reduce it.

 

[Bernhard]

My circuit THD+N

 

[NickKUK]

What are you feeding this with? Bit rate?

If you have a ultra clean amp, great. However it will simply show up the other components around it.. (IIRC 24bit DAC has a noise floor of -144dBS).

Yes the sim will help, but in the end you need to start making prototypes to really find the noise.

 

[MarcelvdG]

It is suggested in the help menue to set it zero, I also tried just to reduce it. View attachment 1137451

I see. Are there any other tolerance settings hidden anywhere? In Spice-like simulators, reltol is usually the most important one.

 

[MarcelvdG]

Does it help to set the "Number of time points" to a power of 2 and "Time step ratio" to its reciprocal?

http://www.spectrum-soft.com/faq/help/faq195.shtm

 

[Bernhard]

time points 256 ratio 0.00390625 is even worse, flat -147dB

 

[Bernhard]

Hiraga amp:

 

[MarcelvdG]

How about 16384 and 0.00006103515625?

 

[Bernhard]

With every doubling, it gets better, however at 4096 it seems hitting the wall again -208dB...
Now I try 8182, will take some time.

 

[MarcelvdG]

8192, I hope.

 

[Bernhard]

16384, THD of the sine source, autoscale, -207,75dB

I now try time step ratio 0.01m
Takes very long time…

 

[MarcelvdG]

Page 941 of http://www.spectrum-soft.com/download/rm12.pdf shows that there is a reltol and it can even be changed over time, if needed.

My guess is that you will need something of the order of 1E-6 or less.

The PSS option could also be interesting to try.

 

[Bernhard]

Small success
Checked the box „periodic steady state“
Simulation of the sine source now gives -248dB.

Independent from those changes, transient analysis of the sine source shows flat harmonics at -250dB, so maybe this is the real limit.

Transient analysis of the circuit shows -235dB flat harmonics.