NAND output doesn't show Low on the output

Hi there!
I was doing some custom logic gates and ran out into this problem. Maybe I am doing something wrong with CMOS config/amperimeter config, but this NAND gate doesn't give me 0 on the output.

screenshot


And the circuit is attached

Thanks, I will have a look.

This issue shoulde solved in 0.4.14 Rev 163 and trunk rev 254.

I previously moved this topic to "Development" thinking it was related to version 0.5.
But I see this applies to 0.4 as well, so it's back into "Report Bugs".

arcachofo wrote:This issue shoulde solved in 0.4.14 Rev 163 and trunk rev 254.

I previously moved this topic to "Development" thinking it was related to version 0.5.
But I see this applies to 0.4 as well, so it's back into "Report Bugs".

thanks! I can see the amperimiter is now working, however the transistor is still being open despite having voltage on the gate above the threshold. This is why the ouput on the NAND gate is not 0 , but should be 0.

Here is the screenshot:

thanks! I can see the amperimiter is now working, however the transistor is still being open despite having voltage on the gate above the threshold. This is why the ouput on the NAND gate is not 0 , but should be 0.

Threshold refers to Gate to Source voltage.
And in Sinulide you set it as an absolute value (no sign).
In this case you need a voltage below: 3V - 0.5V = 2.5V

It starts conducing a little bit below threshold then is when it becomes yellow.

The circuit is working for me.
Did you change some of the characteristics?

I found why it is not working for you:

You have the simulation step ay 1 second and maximum non-linear iterations to 3.
If you set non-linear iterations to the default 100000 it will work ok.

arcachofo wrote:I found why it is not working for you:

You have the simulation step ay 1 second and maximum non-linear iterations to 3.
If you set non-linear iterations to the default 100000 it will work ok.

indeed, the simulation steps are the problem, I remember having set this config for debugging. Setting to 40 iterations gives 0.11 Volts which is good enough. But still the left PNP transistor is shown as turned ON, isn't this a problem? Also the wire is blue and not red.

But still the left PNP transistor is shown as turned ON, isn't this a problem?

The transition fron "Off" to "On" does not happen at an exact point, it is a curve.
And most datasheets give a threshold range, not a single value.

The threshold voltage in simulide is a simplification, it starts to conduce before this point.
The change of color occurs just when it starts to conduce, even if it's only a few uA.
This can be misleading. Perhaps I should use Vthreshold or use 2 colors.


In this case that transistor is exactly at the point when it starts conducing.
And that is an issue in your circuit:

The voltage drop in the multiplexer is equal to Gate_to_Source Voltage in "middle" and "Bottom" cases, where N channels are used.

So the voltage that the P channel after the AmpMeter is receiving is equal to that voltage, so it will never be totally turned Off.
It is conducing just a little bit, much less that the transistors below that are totally On, so the circuit works fine.

If you want that transistor to be totally turned Off yo should use a threshold greater than the threshold used for the multiplexer N channels. For example 0.7 Volts.
Or find some other solution.

Also the wire is blue and not red.

Wires threshold is 2.5 V and voltage is 1.16 V.
I think it would be good if this threshold was also configurable.

arcachofo wrote:

But still the left PNP transistor is shown as turned ON, isn't this a problem?

The transition fron "Off" to "On" does not happen at an exact point, it is a curve.
And most datasheets give a threshold range, not a single value.

The threshold voltage in simulide is a simplification, it starts to conduce before this point.
The change of color occurs just when it starts to conduce, even if it's only a few uA.
This can be misleading. Perhaps I should use Vthreshold or use 2 colors.


In this case that transistor is exactly at the point when it starts conducing.
And that is an issue in your circuit:

The voltage drop in the multiplexer is equal to Gate_to_Source Voltage in "middle" and "Bottom" cases, where N channels are used.

So the voltage that the P channel after the AmpMeter is receiving is equal to that voltage, so it will never be totally turned Off.
It is conducing just a little bit, much less that the transistors below that are totally On, so the circuit works fine.

If you want that transistor to be totally turned Off yo should use a threshold greater than the threshold used for the multiplexer N channels. For example 0.7 Volts.
Or find some other solution.

Also the wire is blue and not red.

Wires threshold is 2.5 V and voltage is 1.16 V.
I think it would be good if this threshold was also configurable.

thanks! I think it is fine, as long as the Probe shows 0 on the output is good enough!
maybe in the future you could add "leakage" field so the transistors would be more idealistically simulated