Analogue to Digital converters

[steve30]

My project at college this year has been to make an analogue interface for a PC, by connecting an ADC chip to the parallel port. This is then controlled using a Microsoft QuickBasic programme in MS-DOS 6.22 (running on a speedy 1.8GHz Athlon XP system).

This is working quite well, and I was thinking about trying to get it to work on the Amiga as well if I can find out how to access its parallel port.

Something I was wondering though, is if anyone has ever tried making a simple analogue to digital converter interface type thing for a Macintosh? My project would probably be quite difficult to interface to a Macintosh as it uses parallel data and the Mac doesn't really have a suitable parallel interface. But it does have serial interfaces and there are plenty of serial ADC chips around.

 

[~tl]

Which ADCs were you planning on using? I believe that most of the serial ones use either SPI or I²C rather than what would be considered as typical computer "serial" (RS232). I would be very surprised if you could find one that would communicate directly with the differential RS422 port on a Macintosh, so you'd need to do some level conversion at least. A small µC might give you more flexibility than a standalone ADC, and a lot of them have at least one ADC built in, but you'd need to have programming hardware, etc, to get a program on to it – ask at your college, I'm sure they have something.

 

[steve30]

I'm not planning on making an ADC for the Macintosh at all. I was just wondering if anyone else had and how it may be possible.

Perhaps something could be done with a PIC? i don't really know much about any microcontrollers, but I know PICs can have ADCs built in and have some form of serial interface. I think some even use USB.

 

[wally]

These guys have thought about the problem some for RS-232 (not Mac, uses multiple handshake lines) and software controlled handshakes: http://pdfserv.maxim-ic.com/en/an/AN827.pdf

So the level conversion is done, but other Mac issues remain.

I've also done some digital oscilloscope/FFT work with the built in sound input A/Ds in PB500s, PB1400s and beige G3s along with some REALbasic programming with Sound Manager calls. I found that under some circumstances there is a startup transient when you call for a series of samples to be made, and many need to be discarded before waveform capture. In the PB5XXs the codec is a Crystal Semiconductor CS4216, and if you examine the input circuit recommendations you can see where to bypass the input AC blocking capacitors and build a DC coupled outside preconditioner that level converts to the input level of the A/D that is offset by the REFBUF voltage level, or for non-precision DC work to an approximation thereof.

My current project is to interface a National Instruments SCSI-GPIB-A box using REALbasic and some downloaded code to interface to the GPIB. I want to drive a varied collection of things beyond A/D and D/A including logic analyzers, universal counters, relay boxes and programmable lab bench supplies.

 

[Bunsen]

either SPI or I²C

Speaking of which ... I've noticed one or other of those signals are present on Mac pinouts occasionally - from memory on the Duo docking bar and the HD connector on the back of 6/7/8100s. Does anyone have any ideas for interfacing to them from the software side if they are in fact present?

 

[porter]

My project would probably be quite difficult to interface to a Macintosh as it uses parallel data and the Mac doesn't really have a suitable parallel interface. But it does have serial interfaces and there are plenty of serial ADC chips around.

What rate of change of data are you expecting? If you are doing serial, you need the baud rate to be at least 10 times the sample rate. Also two standard parallel interfaces for macs are (a) NuBUS and ( B) SCSI.

Most modern Macs have a sound input port, which I count as an ADC.