Apparently you can set the reltol with the Options menu/Global Settings, unless you want it to change over time to speed up uncritical parts of the simulation. It's a tolerance specification for the simulator.
I also run simulation of the amp circuit with the box checked, however at some input level below 1V there was an error and it stopped.
So I changed Input source amplitude from 11.30mV - 1130mV to 11.30mV - 113mV to skip the problem.
But now strangely the power was only 10mW - 1W.
Before it was 1W to 100W
Don‘t know what‘s wrong here.
Maybe i‘m to tired.
Or a just a glitch in the space-time continuum.
Gain is 28dB, Load 8 ohms
THD of the circuit is now down to -235dB which matches the transient analysis.
So I changed Input source amplitude from 11.30mV - 1130mV to 11.30mV - 113mV to skip the problem.
But now strangely the power was only 10mW - 1W.
Before it was 1W to 100W
Don‘t know what‘s wrong here.
Maybe i‘m to tired.
Or a just a glitch in the space-time continuum.
Gain is 28dB, Load 8 ohms
THD of the circuit is now down to -235dB which matches the transient analysis.
Increase Number of Stabilization Periods, especially if there are any long time constants causing slow settling.
I thought the problem was divergence of the periodic steady state analysis at large signal levels. If that's the case, then increasing the number of stabilization periods may help - at least it usually helps in Spectre RF, a simulator I'm used to.
If there are any DC blocking capacitors or DC bias loop capacitors, you can best replace them with DC voltage sources or shorts to get rid of long settling time constants.
If there are any DC blocking capacitors or DC bias loop capacitors, you can best replace them with DC voltage sources or shorts to get rid of long settling time constants.
Yeah. It's remarkable that a company like AP can't get the units right. Thankfully their gear is better than their marketing.
It also irks me when companies specify time in newton-siemens (NS) instead of nanoseconds (ns).
Tom
Thanks, i will try increasing the number of stabilization periods next.
The error occuring at certain input voltages is „matrix is singular“
PSS sims are hard to get to converge sometimes. Increasing Tstab can help. As can decreasing Gmin. SpectreRF sucked a lot of life out of me with its lack of convergence.
Tom
Tom
Tried to increase the stabilization period but the simulation still stops around 1V input.
So I limited the input voltage span to resulting THD vs Pout peak 500mW - 50W.
Comparison:
Before, the "floor" was at -208dB (black), leaving me with a flat graph, now I can see down to -250dB and the form of the curve is shown.
At low output the THD is down to -235dB (red) which matches the transient analysis.
The simulation is much slower than before, it takes about 90 minutes to complete.
Transient analysis is nearly in real time, fast enough so I can even user a slider and adjust values for optimization
.
So I limited the input voltage span to resulting THD vs Pout peak 500mW - 50W.
Comparison:
Before, the "floor" was at -208dB (black), leaving me with a flat graph, now I can see down to -250dB and the form of the curve is shown.
At low output the THD is down to -235dB (red) which matches the transient analysis.
The simulation is much slower than before, it takes about 90 minutes to complete.
Transient analysis is nearly in real time, fast enough so I can even user a slider and adjust values for optimization
That would be great!
I hope to be able to post the schematic within a year or so. Would be nice to see that becoming a popular diy project here.
To be clear, I know that in the real world it would be somewhere between very hard and impossible to achieve such results because of the necessary precision and stability.
Today I corrected a small detail and at very low power, THD approaches the obvious -247dB limitation of Microcap.
The THD curve should fall further below 1W but it does not.
Black is thd of my amplifier, red is thd of the sine source, -247dB. Pictures are overlayed, Pout is not valid for the source.
.
I hope to be able to post the schematic within a year or so. Would be nice to see that becoming a popular diy project here.
To be clear, I know that in the real world it would be somewhere between very hard and impossible to achieve such results because of the necessary precision and stability.
Today I corrected a small detail and at very low power, THD approaches the obvious -247dB limitation of Microcap.
The THD curve should fall further below 1W but it does not.
Black is thd of my amplifier, red is thd of the sine source, -247dB. Pictures are overlayed, Pout is not valid for the source.
.
Have you tried to build it yet (whatever it is) or still just in the sim?
Would be interesting to see what all these graphs and numbers actually relate to.