Source & Sink Capability of GPIO Pins

I don't get the problem. Give him a spec on your input and either let them deal with the interface, or you can design an interface circuit once you see the data sheet on the DAC output.

Where is the problem?

BTW, is your input digital? If not, why do you have a "current source" on it? No, forget I asked that one. I don't really want to know.

no

There are such specs on every processor and FPGA I've seen.

And they are all different. Even one single processor often has separate specs for various (groups of) pins.

petrus bitbyter

That's correct. And the original poster was very clear with is question.

It's an analog input with a precise 100uA current source pull-up to allow the user to simply add a resistor to ground to set a threshold value for an internal comparator. Now they want to force it with an external voltage source. I'm gathering that uP DAC outputs are weenie?? ...Jim Thompson

Sure, this whole discussion depends on what is meant by "general specifications". Yes, each device has its "general specifications". Of course they're going to vary across manufacturers but I assumed JT would know that not every pin in existence would be identical. Perhaps I thought too much of his abilities. ;-)

Now, now! I don't use uP's, so I have to rely on what my customers tell me they're going to do to my chips. When they tell me nonsensical things I try to sort them out here. ...Jim Thompson

Does every single chip you've ever designed have exactly the same I/O parameters? Why should uC I/Os be any different? Get the spec and read it.

Every end user will have a different uP.

I have since squeezed from _my_ customer that it will actually be a rail-to-rail PWM... NOT analog as he first said, so the problem is moot. ...Jim Thompson

There is no standard GPIO drive capability, if that's what you mean.

In fact, my favorite line of processors (ST's with ARM Cortex cores), may have three or four different rated GPIO capabilities, depending on what pins you use (because some are 5V tolerant, some have analog input capability, some are way fast, etc.)

In general, most processors I've worked with can drive way more than

There. Now you have some smoke. Provide your own mirrors, and you can do _anything_.

Love those ST Cortex M? processors. The pin selection utility in Cube is awesome but it would be nice if they added the source and sink capability of each part to it. The Nucleo boards are so cheap that they can be used in a variety of projects.

It's even worse than that. Some of them have programmable drive strengths, so it depends on the software you are using as well as which pin you are using.

Ok, now we know the rest of the story. I can't say how "weenie" the DAC on an unspecified chip might be. I have no idea how you can draw any conclusions. Why can't you just let the customer drive the signal knowing the requirements?

Without some constraints, no answer to your question is possible.

Do you know what its drive capability is? You better find out, or at least specify the limits.

That's correct. And the original poster was very clear with is question.

That's correct. The short answer is "No". Nevertheless some people like to tell more even if the OP is not interested.

petrus bitbyter

Now, now! I don't use uP's, so I have to rely on what my customers tell me they're going to do to my chips. When they tell me nonsensical things I try to sort them out here.

...Jim Thompson

Every end user will have a different uP.

I have since squeezed from _my_ customer that it will actually be a rail-to-rail PWM... NOT analog as he first said, so the problem is moot.

...Jim Thompson