I've simulated this:
In LTspice but it behave weird I can't understand my fault, I think maybe the model have problem?I've downloaded the model from ADA8282 ADI page.
Also I've followed it's evaluation board approach UG-846:
still the problem exist.
New question rises: why they've used C5, R5 and R4?
I have simulated ADA8282 using LTspice and the provided Spice model. I don't see any issues, as shown below:
I followed this procedure to create the symbol within LTspice:
I also made sure I have a path to ground from each input (E1, and V1).
Here is the simulation file and the symbol file *.asy, for your review:
7725.ADA8282 Ezone question 5_13_19.zip
Please review and let me know.
Thank so much, have you tested it for long time, will it stay stable for 1 secound?
Look, in first one millisecond I have what you've achieved, but it deviate as the time passes:
Thanks for your sharing of model, as I think our models are the same except for it's model pin arrangement, I didn't upload it but only check on your model for long time simulation.
The best requard
Yes, I see the same strange behavior of the model if I let Transient Response run beyond about 45ms. I cannot tell what's causing the ADA8282 model to misbehave like that!
The Spice model file does hint that the model may be very minimal, copied below:
"*This model was created for the purposes of simulating functionality only.*Parameters modeled include:* Output Voltage Swing* Bandwidth* Functionality at 36dB gain* Supply current"
I think in normal applications dealing with typical signal frequencies / periods, that's a probably a "don't care".
Is there a particular reason that you need to let your simulation run that long?
I think you are asking about the source E1 in my schematic?
That's just a "voltage dependent voltage source" to help me create the differential input by specifying its gain to be "-1V/V" so that it'd be inverted compared to the other source V1.
I'll forward the question about the misbehavior of ADA8282 pspice time-domain transient response issue going past 45ms to people in charge and if I get a resolution I'll post it here.
Hi Hooman, thank you so much.