Hello all
The attached simulation was working without problem a few months ago. Now it crashes and gives the following message
Fatal error: Timestep too small(1.25006e-19) at t=4.44851e-11
The problem occured while parsing the line:
.MODEL M
I couldn’t even find where this line is
I think that something has changed in Qspice in the last updates so now the file does not work.
I am at a loss with this. Any idea?
Thanks!
1.- UC1846 - push-pull-current-mode-open-loop.qsch (23.7 KB)
[UC1846.qsym|at
UC1846-save_as_sub.txt (6.7 KB)
tachment](upload://pPMf2duJJjyhyzRzMtwdlzay0ka.qsym) (3.6 KB)
P.S.
It seems there is a typo in the word “occured.” The correct spelling is “occurred.”. Maybe Mike would want to correct this.
Sorry the symbol was missing
UC1846.qsym (3.6 KB)
Current solution is to update to latest Qspice version, and have a supply source setup in the similar way as TI PSpice demo circuit.
1.- UC1846 - push-pull-current-mode-open-loop.qsch (23.8 KB)
UC1846.txt (6.7 KB)
TI PSpice example
3 Likes
I got it. Thank you very much!
I made a modernization of the chip model. In particular, I replaced the trigger model.
But I found that the Qspice with this circuit is completely false. The transistors put in the circuit are too powerful and have a large capacitance. But in spite of that the output of the chip is almost full amplitude. I took part of the circuit - output buffers from the model and transistors with transformer etc. In this part of the circuit the control pulses turned out quite different from the
whole circuit and approximately as they should be. Of course you can say - put the right transistors, but the calculation should be correct with any transistors. So this is a gaffe ( BUG ) of Qspice!
UC1846.txt (3.0 KB)
Buf.qsch (17.1 KB)
1.- UC1846 - push-pull-current-mode-open-loop (2).qsch (24.0 KB)
Export netlist of Buf.qsch, add {} for Lm and model of BSC123N08NS3, run in LTspice and it give a result similar to Qspice.
(remove .txt from this attachment)
Buf.cir.txt (1.4 KB)
Simplified the sub-circuit to only drive a resistor in its current setup and ran it in LTspice, it even seemed to have difficulty driving a very light load (the result is the same as in Qspice). This is not meant to prove or disprove if it is a bug, but just to verify that both Qspice and LTspice are giving the same set of results in the current setup.
1 Like
In the subcircuits, X1, X2 1846OUT, pins 3 and 7 need to be swapped
1 Like
