0
\$\begingroup\$

enter image description here

I would like to verify the function of this circuit, as well as determine the power usage and heat generated by each component. (I need the latter for doing a thermal analysis)

The circuit has been designed in EasyEDA, and I've tried simulating with same SW (it uses NGspice), but I get the "unknown subcircuit" error for every component but the resistors.

What is the most straightforward, but acceptable, way to simulate/test this? For the calculations of power, I can accept up to 50% error. I'm considering the following options:

  1. Making the missing Spice models, based on the models of similar components and then changing the key values based on datasheet.
  2. Using another simulation SW
  3. Replacing the components with ones that have spice models, for the simulation and possibly also for the final design.
  4. Order a pre-assembled test PCB where I can isolate parts of the circuit with jumpers, for example, and test with this. (The components are all SMD so breadboarding is no option)
  5. Order the equivalent THT components and test on breadboard.
  6. Doing calculations.

And combinations of the above.

\$\endgroup\$
0
\$\begingroup\$

I think just using the loading currents and voltage values for the datasheets of each component, add them all together should give you a power consumption within the error you specified, if nothing is broken just doing that should be a decent approximation. For each resistance just use the current/voltage that the datasheet specifies for the component is connected to.

Now if you have to simulate it and have no models, again for the error you said 50% just making dependent sources(current or voltage) with the values from the datasheet should be enough.

Making new models from the datasheet parameters should be entirely possible, sometimes the datasheet itself includes either a description or a link on how to SPICE model the component so just looking at that might be worthwhile.

Honestly, I don't think you need to go as far as simulating to get a power consumption within 50% of this circuit if you have the datasheet parameters for each IC.

\$\endgroup\$
2
  • \$\begingroup\$ Perhaps. The big unknown as I see it though, are the MOSFET's. Their heat generation would depend greatly on how quickly they switch. And as I've understood from previous posts, figuring out how the switching time is not so easy. \$\endgroup\$ – K0ICHI Oct 3 '19 at 9:12
  • \$\begingroup\$ I think for the MOS-FET they have a typical switching time in the data sheet as Ton or something alike. You can use the effective output capacitance to calculate the energy needed for turn ON and turn OFF and then the ON resistance for losses during conduction. Now, you could either simulate just the MOSFET or one that is similar and see how many losses it generates during switching from a simulation and add it to the rest. Or just build a single unit with the MOSFET and get approximate losses from that if you can afford it, then just assume they all roughly behave the same. \$\endgroup\$ – Juan Oct 3 '19 at 9:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.