Hi All
I do not now but I use sub circuits, and these have nodes named inside them, is this maybe the cause.
I only can sim when I delete the copied part, as you see, nice plot, three level. If the subcircuit is cause I did
learn again something more but maybe subcircuit nodes do not impact schematic because I use also more
then one in a simple channel without that copy for BTL.
regards
I do not now but I use sub circuits, and these have nodes named inside them, is this maybe the cause.
I only can sim when I delete the copied part, as you see, nice plot, three level. If the subcircuit is cause I did
learn again something more but maybe subcircuit nodes do not impact schematic because I use also more
then one in a simple channel without that copy for BTL.
regards
Attachments
I have now try other gate driver models, it looks like the IR2113 Behavioral Model by analog is the cause of the trouble
or even a bug.
I did try in old 32 but LTspice, hmm sims faster more stable and cleaner but same trouble..
pic 1626 let also seen a big delay between the halves while it is exact the same circuit for BTL. that give a uge crossover
distortion.
Pic 1629 is a sim with normal models for Gan, it is slow, but it do behave as expected, with common mode
voltage as is normal with multilevel systems. But it is now with a common mode filter out of fase as needed.
I do not now what is wrong with the IR2131 subcircuit, but there it go wrong.
regards
or even a bug.
I did try in old 32 but LTspice, hmm sims faster more stable and cleaner but same trouble..
pic 1626 let also seen a big delay between the halves while it is exact the same circuit for BTL. that give a uge crossover
distortion.
Pic 1629 is a sim with normal models for Gan, it is slow, but it do behave as expected, with common mode
voltage as is normal with multilevel systems. But it is now with a common mode filter out of fase as needed.
I do not now what is wrong with the IR2131 subcircuit, but there it go wrong.
regards
Attachments
Jan. I have redrawn with LTspice own models, then things are oke. because these models are lossless there is no delay.
Th feedback circuit do work. Go try now other ways to use the models of interests.
This 5 level does have carrier 2 Mhz. 4 fold of original carrier.
Attachments
Wel last attempt is done, sub circuits do not work when use multiply versions in one schematic., special in opposite phase use like
BTL multilevel.
In some way it does see each other, and maybe the need to edit sub circuit names for each part make them uniek.
Normal models do work fine but are slow, and the ltspice behavior models do wel as expected.
Did all kinds of tryouts, nothing works, the switching pulses are also very small, but when use one btl channel things get oke.
output is not in fase, even when I put in phaseshift 180 degree it does not work, I expect a lot of common mode voltage in the
model of the subcircuit because of wrong behavior and fase shifts.
I think even of a bug, because it has to be possible to use multiply subcircuits without this trouble because uniek names stuff.
Pic with yellow model do fine, but very slow, that is why I did use subcircuits. both circuits are same connected, no mistakes.
So conclusion, I need other models, and maybe writer of LT spice can look into this.
regards.
BTL multilevel.
In some way it does see each other, and maybe the need to edit sub circuit names for each part make them uniek.
Normal models do work fine but are slow, and the ltspice behavior models do wel as expected.
Did all kinds of tryouts, nothing works, the switching pulses are also very small, but when use one btl channel things get oke.
output is not in fase, even when I put in phaseshift 180 degree it does not work, I expect a lot of common mode voltage in the
model of the subcircuit because of wrong behavior and fase shifts.
I think even of a bug, because it has to be possible to use multiply subcircuits without this trouble because uniek names stuff.
Pic with yellow model do fine, but very slow, that is why I did use subcircuits. both circuits are same connected, no mistakes.
So conclusion, I need other models, and maybe writer of LT spice can look into this.
regards.
Attachments
Ya, standard.bjt: ".model 2SCR542F3 .... Vceo=1.00E+21 Icrating=1.00E+21" should be maybe "Vceo=30 Icrating=3".
https://fscdn.rohm.com/en/products/databook/datasheet/discrete/transistor/bipolar/2scr542f3tr-e.pdf
https://fscdn.rohm.com/en/products/databook/datasheet/discrete/transistor/bipolar/2scr542f3tr-e.pdf
Last edited:
I'm not at my computer for a while but if you click and drag a box over the relevant portion of the waveform doesn't it give you a volts per second measurement in the lower left corner?
BTW it's hard to read your schematic but the green waveform doesn't look like slewing to me, it looks more like a bandwidth limit.
BTW it's hard to read your schematic but the green waveform doesn't look like slewing to me, it looks more like a bandwidth limit.
Hi
I do not understand the reading that it gives on the bottom left corner it reads x=24.52us and y=-2.80v but reading varies depending on where the + (indicator) hovers in the window area (not the cursor placement).
In the resulting window (my image) Diff (Cursor2-Cursor1) Slope: -2.25794e+007
Now I've read to calculate slew rate run a square wave to reach a triangle like form output. I've also read a freq of 100khz will do as estimate, this was what I did. I've executed a run at 100khz, input amplitude calculated at near clipping (0.6v) and at measured output max estimated full power of 50w.
Am I doing it correctly?
Thanks!
