Other Parts Discussed in Thread: TINA-TI,
Tool/software: TINA-TI or Spice Models
It seems, the model for LMG5200 works only for the example circuit with specific load and with specific driving signal timings.
When trying
1)Implement test circuit from the manual(SNOSCY4D–MARCH 2015–REVISED MARCH 2017), page 8 Fig. 5, TINA-TI fails to converge at OP step 2. The computation becomes infinite without any useful result.
2) The original demonstration schematics fail to simulate when the dead time is reduced by the trimming resistor at the AND-gate input. Setting 100Ohm makes simulation unstable. with lower values it will fail
That is not right.
3) Since I got also a non-encrypted model for PSPICE, (* Date: 16SEP2015 Version: Final 3.00), I tried it in LTspice with negative result: The simulation was either very(hours) slow, especially at input pulse edges, or failing with "time step to small". The load current was fundamentally wrong, indicating some short circuit leaking inside the model, with random amplitude shots.
Here is 3 Ohm load under 24V source. The current is totally wrong, if you divide the latter by the former.
I.e., the model can be used for demonstration, but have no practical use.
Dear TI, do you have another, working model for LMG5200?
Unfortunately, we do not have a prototyping machine here to examine the operation in hardware. And the development got very slow at the point of picking predictable high frequency components with models available.
