Oops, nodenames, not codenames.? Darn auto-correct.

?

Are you certain that those are the exact, correct codenames nodenames?? Did you get them from the expanded netlist, and were you using a version older than 24.x?

?

Andy

john23,

?

I uploaded AD797_Opamp2.zip to the Temp folder.? It shows the opamp2 symbol, applied to the "MyAD797" (wrapper) subcircuit.

?

Andy

?

On Sun, Feb 23, 2025 at 02:24 PM, john23 wrote:

Hello Andy,I managed to do it as shown in the attached file.

added the whole netlist as a directive and deleted the last pin number of the title.

Is it Ok?

That works too.

?

I was trying to make it work without modifying the original AD797 SPICE model.? Sometimes, IC manufacturers get really picky about altering their SPICE models, and I wanted to avoid that.? It does indeed make your simulation function - although the schematic looks awful when opened in LTspice, because it scales to fit the text.

?

When you downloaded the AD797 SPICE model, you agreed that you "may modify this SPICE Model to suit Your specific applications,", so apparently they do not mind you modifying it this way for your own use.? But you are prohibited from making this change to the model and then distributing a copy of the modified model to anyone else.? The change you made alters the function of the model.? That is prohibited by the License Agreement that you agreed to when you downloaded it from Analog Devices.

?

Also, remember that you did alter it, and if you come back to your simulation some 10 years from now, remember that it is a modified SPICE model.? It helps to add a Comment note to the schematic, saying that it's been altered.? 10 years from now, if you tried to update the AD797 model it contains, it won't work unless you modify it again.? That was another reason why I preferred to put a "wrapper" subcircuit around the original model instead of modifying it.

?

Andy

?

On Sun, Feb 23, 2025 at 12:42 PM, john23 wrote:

Hello Andy,I did what you said it still gives a netlist bug which says pin mismatch.

Ltspice file and the errotr is attached.

Where did I go wrong?

You did not follow what I said to do.

?

What was the point of creating a subcircuit named "MyAD797", if you did not even bother to use that subcircuit?

?

The subcircuit you created (as a "wrapper" around ADI's AD797 model) is named "MyAD797".? Therefore, change the name next to the opamp2 symbol, to that exact name.? You want your opamp2 symbol to call the MyAD797 subcircuit that you created.

?

I wrote that already in message 158621, did I not?? Carefully read those instructions again, and follow what I wrote.

?

Also, get rid of the command ".include ad797.cir" that you added to your schematic which should not be there.

?

Andy

?

Hello Andy,I managed to do it as shown in the attached file.

added the whole netlist as a directive and deleted the last pin number of the title.

Is it Ok?

?

Thanks.

?

/g/LTspice/files/Temp/op-amp%20AD797%20%281%29.asc

Hello Andy,I did what you said it still gives a netlist bug which says pin mismatch.

Ltspice file and the errotr is attached.

Where did I go wrong?

?

/g/LTspice/files/Temp/ad797_cir.zip

Group member "flyback" (massey2) uploaded a schematic earlier today "LLC to compare.asc".? ?But she or he may have forgotten to read the group's guidelines.? Files should be uploaded into the "Temp" directory only.? (I fixed that for you.)? Uploaded files must be accompanied by a message from you, telling us why you uploaded it, or what is your question.

?

The file she or he uploaded does not run without errors.? There were mistakes in the schematic.? I corrected some (not all) of them, and uploaded "LLC to compare AI1.asc".

?

But it is still not right.? "flyback" attempted to make parameters depend on a voltage in the circuit, and that can never work.? Parameter values are fixed during any simulation run, and their values must be known before the simulation begins.? I substituted a dummy value for the first one that was supposed to track a voltage.

?

If you want to make the parameters (not .PARAMs but circuit element parameters) depend on your voltage at node EA, you must find another way to do it.

?

The extensive use of Rser with your voltage sources is somewhat troubling.? Attempting to power an op-amp that way, with no bypass capacitors, might lead to failure, although Rser was small enough so it's probably OK.? Also be more careful with your typing.? Numbers such as "0..01" might not work.? (I fixed that too.)

?

Andy

?

Who?

So.. what? bandwidth are You looking for?

Thank you for your reply and the test file, Andy.

Indeed, splitting v(out) into v(out)@1 and v(out)@2 clearly revealed the missing part at the end,

and so the crazy fft plot makes perfect sense. Specifying a time range solved the fft problem.

?

--

Ryu

"In that case I would suggest capturing waveforms on a transistor driving a dummy load, and matching them by modifying Rb for a model that is known to already have correct high-frequency behavior. One step further would be to get the low-current Rb with the noise method and then adjusting Rbm and Irb to match the waveform."

?

A practical & useful suggestion. Thank you!

In that case I would suggest capturing waveforms on a transistor driving a dummy load, and matching them by modifying Rb for a model that is known to already have correct high-frequency behavior. One step further would be to get the low-current Rb with the noise method and then adjusting Rbm and Irb to match the waveform.

Hi Jerry, I am investigating a NMOSFET based class E oscillator. The BJT will be used as a driver for the mosfet and part of the feedback loop will include the rbb and the cap_be of the BJT in series with the input capacitance of the mosfet. Hence I am after the high current version of the rbb.

Thank you keantoken. I am interested in the Rbb at high current values. Would the s-parameters help for this estimation? If if does, then please provide some guidelines on how to do so. I am already familiar with deriving s-parameters for a bjt in LTspice.

IIRC, we've been told that v24.1.3? restores LTspice's ability to generate the netlist, but the netlist it generates is no longer usable as the input to new simulations?

?

Stupid.? Er, Bummer.? No, it IS stupid.

?

Andy