03-16-2011 06:46 AM
I've reverse engineered a diode based thermostat which uses a 1N4148 as the temperature sensor.
I've entered the circuit into the schematic editor, but whatever I do it refuses to run the simulation! I get an immediate "Timestep too small" simulation error, and the convergence assistance cannot solve it.
I'm don't *think* I've got any "stupids" that would cause the simulation to fail, but maybe I missed something obvious.
As a "Starter for 10 points" how can I get this simulation to run?
As a "Bonus question for 30 points" how can I change the temperature of the diode to (say) 150C while leaving the rest of the simulation at nominal temperature?
Solved! Go to Solution.
03-16-2011 07:15 AM
I got rid of a singular matrix error by putting a DC power source and lamp on the relay output circuit, but it still fails to simulate.
I attach the updated model
03-16-2011 11:58 AM - edited 03-16-2011 11:58 AM
Hello,
Let's use a better SPICE model for the LM324N:
Run the simulation, this convergence error should dissapear. I've attached your circuit with this modification.
Hope this helps.
Fernando
03-16-2011 12:20 PM
Yes, definitely plenty of Brownie points for that The circuit now works with some tweeks to the relay on current.
I''l open another thread on how to adjust the temperature of JUST the 1n4148 diode (to change VF)?
02-25-2012 09:45 AM
Hi everyone,
I'm having the same convergence issues with a circuit I'm trying to simulate. Can somebody please run it and see if he/she can figure out what I was doing wrong.
I've attached the circuit.
Please explain how you fixed it.
Thanks
02-27-2012 06:31 AM
Wadel,
You have the power supply going to pin 8 (VCC) of the LM358 reversed.
02-27-2012 07:40 AM
Thanks Tom.
02-22-2016 08:43 PM
I working with a solar inverter circuit. I keep getting the convergence error message. Any ideas on what might be wrong?
06-06-2020 01:23 AM
Your AC ground and DC ground are same. Avoid same ground for AC and DC