Couldn't you communicate with ADC and DAC?

[Krumlink]

Is it at all possible to communicate between PIC's using a DAC and a ADC.

You can send a Digital signal to a off PIC DAC and have the other PIC use its internal ADC. You would need to supply the Comm line, 5VDC and GND lines for accuracy. It could work right?

 

[Sceadwian]

Yes of course, in fact if the PIC has an external memory interface many ADC and DACs can be hooked up directly to the address bus allowing the DAC or ADC interface to be as simple as reading/writing bytes from memory on the PIC. If your PIC doesn't have an external memory bus or you don't want to try to use it no big deal, just read the data sheet on the DAC/ADC and hook it up to an I/O port.

 

[kchriste]

You mean communicate between PICs using an analog voltage? Well, yes and no. You will get an approximate value. Say you want to send the number 0x55. You convert it to analog with one PIC and back to digital with another PIC and you might get 0x55. You might also get 0x56, 0x54 or something else "close" to 0x55.

 

[3v0]

The answer is yes but why would you want to? Pics have several other ways to talk to each other. There are several serial methods. If you want parallel look at the communication between the inchworm and unicorn (schematic).

 

[Nigel Goodwin]

Is it at all possible to communicate between PIC's using a DAC and a ADC.

You can send a Digital signal to a off PIC DAC and have the other PIC use its internal ADC. You would need to supply the Comm line, 5VDC and GND lines for accuracy. It could work right?

Like everyone else says, silly idea! - use a single line and send serial data, A2D and D2A conversions aren't 100% accurate, they are at least +/- one bit.

 

[Analog]

Absolutely!

We could also have them communicate with pneumatics, but why would we want to torture ourselves??????????!!!!!

 

[picbits]

You'd be better using a current loop rather than relying on voltages over a distance.

I.e. putting a constant 20ma current through a cable will appear as 20ma at both ends.

 

[Krumlink]

I was wondering if it was possible, I know that there would be a accuracy issue. I have no plans nor do I want to start trying A/D D/A communication.

 

[Blueteeth]

You pretty much do it anyway...after all, surely an output pin is simply a 1-bit DAC and an input pin a 1 bit ADC? 'Digital' is purely theoretical .. once you go into the real world you deal with analogue voltages which represent digital information 5v= 1, 0v = 0...oh but what about RS232?

Some have said its a silly idea, but not necessary..

It's used in RF comms, and even wired comms to increase the throughput without increasing bandwidth (which in RF is very much at a premium). If you have 4 voltage levels, then that represents 2 bits....I mean our cable provider uses 64 QAM to pump data through (64 signalling levels). Of course its a trade off between 'bandwidth' and 'dynamic range'....higher quantisation increases the risks of errors.

Of course I didn't think you were actually going to use it over, say, SPI/USART/I2C, its theoretical. but interesting nonetheless. Hell, its a pretty good cheap form of very short range wireless comms...no line of sight like in IR, sounds goes through walls too...but yeah BIG hassle.

My two, possibly three, cents.

Blueteeth

 

[Sceadwian]

I can't remember the project but someone did this using a very basic direct DDS and a logic accumulator to do DAC to ADC back to DAC for a neuron project a ways back all completely on digital I/O lines. Elegant, but not very practical.