Thanks for posting this. When I run it, it throws an error that it cannot find a model for ZT450 , MPS8599. As it is packaged, it is not portable.Here a correction, I did made the ASY wrong, in last post sorry, has two npn.
I have included the MJ darlingtons, and it do sim nice, when using original like you ask, then do copy cmd file to lib for new models I did already include but you can also use the new one from #Mooly and include only the MJ transistors
all is included in mooly zip file, if you do not mind mine hijacking Mooly?
regards
Here is what I came up with. I thought it would be different from what I typically use for class AB amplifiers but actually I think class D benefits even more from having a low number of test cycles. This means there are more datapoints per cycle and so each comparator flip gets the most accuracy in the simulation. In ideal cases you can use only 5 cycles but 10 cycles is fairly foolproof.
The first setting I change if the simulation is too fast or slow is "res". I almost never change the number of cycles. If there is a large slant in the noise floor of the FFT down from LF, I increase the number of warmup cycles. If the simulation gets stuck during the warmup cycles I set "warmup" to 0 so I can see what the circuit is doing in case it is oscillating or latching up.
I changed the inductor symbols but this doesn't mean anything, it is just a different symbol for the same component.
The first setting I change if the simulation is too fast or slow is "res". I almost never change the number of cycles. If there is a large slant in the noise floor of the FFT down from LF, I increase the number of warmup cycles. If the simulation gets stuck during the warmup cycles I set "warmup" to 0 so I can see what the circuit is doing in case it is oscillating or latching up.
I changed the inductor symbols but this doesn't mean anything, it is just a different symbol for the same component.
Last edited:
FET optocoupler model?
Hi,
Does anyone know of an existing Spice model for a FET optocoupler like the H11Fxx series from ON semi?
Much obliged for any pointers.
Jan
Hi,
Does anyone know of an existing Spice model for a FET optocoupler like the H11Fxx series from ON semi?
Much obliged for any pointers.
Jan
I'm glad I know about that part now.
I tried modeling it in LTspice, but I can't seem to get the saturation curves to work right.
Here is as close as I got. I tried using some Jfet models from OnSemi hoping they were similar to what is used in the device.
Either the bad FET modeling is the issue or I don't know enough about the architecture of the device. Maybe Ian could help?
It could be using MOSFETs, although the datasheet doesn't indicate this and the results I got when I tried were not this close.
I tried modeling it in LTspice, but I can't seem to get the saturation curves to work right.
Here is as close as I got. I tried using some Jfet models from OnSemi hoping they were similar to what is used in the device.
Either the bad FET modeling is the issue or I don't know enough about the architecture of the device. Maybe Ian could help?
It could be using MOSFETs, although the datasheet doesn't indicate this and the results I got when I tried were not this close.
Attachments
That's a great attempt keantoken 
I can recall opto isolated FET's as far back as the 1980's, in fact I think Elektor (I think it was Elektor and not WW) doing a variable inter channel blend to reduce hiss on FM tuners. It reduced channel separation as signal strength reduced.
I remember thinking at the time what fantastic devices these were...
I can recall opto isolated FET's as far back as the 1980's, in fact I think Elektor (I think it was Elektor and not WW) doing a variable inter channel blend to reduce hiss on FM tuners. It reduced channel separation as signal strength reduced.
I remember thinking at the time what fantastic devices these were...
This is very nice Kean, I can use this in my project, for proof-of-principle even if not 100% accurate (and it never is of course).
To lift part of the veil, this would be part of an auto-tuning notch filter as an add-on to the autoranger. ;-)
Jan
Just realized a major issue with my idea to use an opto-FET for floating frequency control of the notch.
Normally, when using a FET for frequency or amplitude control, you supply the gate with about half the AC Vds to keep distortion low. But in an opto FET like the H11Fxx the gate is not accessible. Bummer.
Jan
Normally, when using a FET for frequency or amplitude control, you supply the gate with about half the AC Vds to keep distortion low. But in an opto FET like the H11Fxx the gate is not accessible. Bummer.
Jan
Kees, a method that is often effective to isolate the issue is to progressively delete parts of your circuit until the issue goes away. That may indicate whether it is caused by some specific circuit or sim command.
Are you sure it is not just taking a lot of time for the sim?
Jan
... the solution of course being to use a discrete FET as the floating control element and drive its gate through a photovoltaic isolator voltage generator.
Jan
I am now shure the new LTspice 64 bit has a lot of flaws, it does crash while sim, and give strange outcomes like 1 Kv on the voltage source output while it is 1 volt etc, or measure output get 12 mV on 8 ohms but 20 amps current.
here you see the triangels get messed up, the only way to get it right is shut down and restart the pc.
with the old version I have never trouble.
regards
here you see the triangels get messed up, the only way to get it right is shut down and restart the pc.
with the old version I have never trouble.
regards
Attachments
I've started experiencing a lot of crashes on LTXVII x64 following a clean install of the latest version of W10. Mine crashes when running a transient analysis, and typically at the instant the window should split into the two screens. Occasionally it appears OK but crashes when you probe anything.
I haven't found any reason for this although I suspected a graphics issue, however a lot of testing with different drivers and settings, and even using just the basic MS video driver have all shown the problem at some point.
I haven't found any reason for this although I suspected a graphics issue, however a lot of testing with different drivers and settings, and even using just the basic MS video driver have all shown the problem at some point.
Attachments
I contacted Mike Engelhardt just after posting this and received a reply within the hour. Apparently he has received complaints about this and asked was I running the latest update that came out just 3 days ago. I wasn't, but I am now. So fingers crossed. Its been OK in three or four sims run in quick succession.
Well I didn't really gave that a lot of thought - just that all these apps have the half-signal to the gate. Is it only a matter of symmetry or also a matter of non-linear transfer curve, even if symmetrical. Two different issues, no?
Jan
