SPICE MODEL - ASD Out Noise

Hi MrT918 ,

Please upload your LTspice files, including model and symbol (zip them up).

mike

4111.OP467.zip

2818.op467libmodel.zip

Hi MrT918 ,

The OP467 model can be used in LTspice, just like in Pspice. Both use SPICE. You just need a symbol for the device. Quickest way to do this:

1. Close your schematic.
2. Place the op467.cir file in the same directory as your schematic
3. In LTspice, open the op367.cir (you'll need to change the viewable types to Netlists or All)
4. In the open netlist find the line: .SUBCKT OP467   1  2  99 50 27
5. Right-click anywhere on that text line
6. Choose Create Symbol
7. Browse to the directory that contains your netlist and schematic (should go there automatically)
8. Click OK. The symbol is generated and opened.
9. Open your schematic. 
10. Type P to Place Component.
11. Click Refresh.
12. At the top, choose Schematic Directory
13. Choose the op467.
14. Click Place.

You can get fancy and use the universal op amp symbol and modify it, but above is the quickest way.

I'll see if I can point you in the right direction for getting help re Signal Chain Designer. 

mike

Hi MrT918 ,

MrT918 said:

Could you please confirm whether I should expect feedback, or if I should proceed with creating a support ticket?
Also, would it be possible to forward this request to the appropriate product owner?

I've moved the thread to the Signal Chain Designer forum, as your question is really coming from the modeling standpoint, as opposed to the operation of the device itself. If you don't get any response, tag me again, by typing "@"  MStokowski .

mike

I am not really willing to use LTSpice for my simulation, but my guess is that the model will show the same fault noise ASD. I am willing to find the correct model that exhibits the same noise ASD as in the datasheet and signal chain designer of AD.I am not really willing to move to LTspice for the simulation, since my assumption is that the same noise ASD discrepancy will most likely appear there as well. The issue seems to be related to the macro-model itself rather than the simulator. My objective is to find a proper model that reproduces the same noise ASD behaviour reported in the datasheet and in Analog Devices Signal Chain Designer.

As you see the noise ASD is not correct even under the LTSpice.

I think that the model is not correct.

I completed a study on SPICE modeling and reviewed the macro-model provided by Analog Devices for the OP467.

It now appears that the issue is not only related to the simplified subcircuit structure of the macro-model itself, but also to the fact that the transistor models used internally specify almost no parameters beyond the forward current gain. Parameters typically associated with accurate noise simulation are missing.

*
* MODELS USED
*
.MODEL QN NPN(BF=33.333E3)
.MODEL DX D
.MODEL DY D(BV=50)
.ENDS

As observed above, the BJT model only defines the forward beta (BF), while other parameters commonly involved in noise modeling are absent.

The remaining question is therefore the following:

Why does the model implemented within Analog Devices’ Signal Chain Designer appear to reproduce the expected noise behaviour, while the SPICE macro-model provided directly by Analog Devices is not the correct one?

Hi MrT918,  I can't comment too much on the SPICE model from 1992, but can confirm that the Signal Chain Designer Tool would use information from the datasheet to get the the noise to simulate accurately. Please let me know if you have any other questions/concerns around the Signal Chain Design Tool.  Thanks, Dave