I do not have any special techniques to offer as I do not use a simulator very often. And when I do use a simulator I am using it to focus on an instruction flow pattern to fine tune an algorithm. I do not use a simulator to "play make believe" with peripheral devices. There are just too many peripheral types and variations out there to expect any simulator to do a robust job at simulating the part. When you add to this the multitude of connection configurations that are available it quickly becomes a non-starter for using a simulator as a general case tool.

Keep in mind too that a simulator software vendor may very well have implemented a peripheral simulation to a very specific or limited use case that matches the hardware usage model for some hardware they had. In that case the model for the peripheral would likely fail to give a reliable behavior when used in another way.

Another thing to think about is that the peripherals that a simulator may support well are likely to be older style parts that may very well be out of the mainstream of design choices by the time you get to implementing your own gadget. If a simulator was up to date with all the modern parts and peripherals being supported you would expect to have to pay a lot of $$$$ for the tool and that would likely include $$$ yearly maintenance fees. This is unlikely to be the case for low cost tools.

Many simulators come with an extensible interface that allows the addition of new peripheral devices to the simulator. This can be "cool" to implement and play around with but the time spent is way better used to make the real world implementation work. This of course presumes that you place a value on your time and are working to make a successful product. On the other hand if you are a developer working a high $$$$ tool set company earning a nice salary to make peripheral simulations then by all means have at it and make it robust as you can. But while you have fun doing that do hope that the managers that fund the simulation development projects don't come to sites like this and find out how little value that experienced developers see in such tools. They may change their ideas on the types of things that they want to fund in the future!!

Michael Karas