Critical error or not ?

[sycomore]

Hello from France,

I have a problem with the (future) pcb hereafter.

ZSE30 is a vacuum sensor.
When i simulate logicstate = 0, no problem, all is OK

But when i use 1, the simulation stop and i have a "surprising" error : "Pin 'n' does not exist in child module." for all the terminal blocks.
Question : why a problem with the whole TB in 1 and not with 0 ?

Second error : what does this error mean : [SPICE] TRAN: Timestep too small; timestep = 4.33681e-019: trouble with node #U1_U1_W1#branch. ?
U1 is using when logicstate is 0. When logicstate is 1, i use U2 and U1 is "sleeping".

The principe of fonctionment is :
- logicstate 0 -> pump on, solenoid valve off
- logicstate 1 -> pump off, solenoid valve on

All seems OK for me but not for Proteus and beetween Proteus and me, i tend to trust Proteus

PS : in the real life, its will be not a lamp for L1 & L3. L1 will be a vacuum pump and L3 a solenoïd valve.

Thank you in advance,
Philippe

pompeV015b_110V.pdsprj

(26.16 KiB) Downloaded 62 times

[Ettore]

1) Edit all connectors J1 to J12 and unselect "Attach hierarchy module" option and enable "Exclude from Simulation" like this:

Immagine 2022-06-11 005110.jpg (40.38 KiB) Viewed 292 times

2) Go to System menu, Set Animation option and finally push SPICE Options. In the dialogue select "Settings for Better Convergence" like this:

Immagine 2022-06-11 005408.jpg (58.6 KiB) Viewed 292 times

and finally push on the Load button then OK.

3) When you raise a simulation question then you have to post it in the Simulation forum and not in Schematic Capture area.
This time I have moved in the correct forum section.

[sycomore]

Thanks a lot, I never would have been able to find this on my own.

No more problems with the TB but always this one "[SPICE] TRAN: Timestep too small; timestep = 4.33681e-019: trouble with node #U1_U1_W1#branch."

By searching on GOOG, it seems i am not the only one to stumble on this "timesstep too small". May be the rookies problem.

I consider that the schematic is OK.

Next step : the routing

A+
Philippe

[Ettore]

Timestep too small error is often because of a schematic error or some sort of critical issue. For example, in your schematic the blue wire (mains NEUTRAL) is connected to ground which makes it connected to the negative pin of the battery B3, which is tied to ground as well. I believe, this is a gross mistake and quite dangerous for your safety if, for some reason, you revert the mains live\phase with the neutral.

So, be aware when making that circuit; Proteus won't kill you if you make mistake, but 110 Vac will !!!

[sycomore]

Thanks for yours remarks !
I have just one life and i don't want to waste it.

When i started my schematic, Proteus absolutely wanted ground 5Vcc and 230Vac otherwise he did not work. So i kept the "2" grounds all along.
If i erase the 230V ground, the simulate lasts longer before stopping (~5s) It's better.

In France, we work with 230V, not 110.
The schematic is with 110 because U1/U2 are the only one zero crossing i found in the library and they are 115Vac max.
I will use MCKSD350D5-W

I will try my schematic in a breadboard with 15Vac and if i haven't flames, i try again with 230Vac.

Merci & A+
Philippe