Hello folks,

I was wondering what would happen to a 6522, if one of the bidirectional ports has been programmed as an output by mistake. If this output is connected to another output which is in a different logic level, I assume that high currents would be floating which could probably damage both devices. Is this assumption correct for the 6522, or are all of the outputs rather high impedance? How could I protect hardware interfacing the 6522 from damage by incorrect programming? Is it possible to place a resistor in series to all in-/outputs? How do I have to calculate these resistor values with respect to staying inside a TTL level?

Any help would be very much appreciated!

Regards,
Matthias Hartl

I checked all my data books covering the 6522, 65C22, and 65NC22 VIAs from Rockwell, GTE (now California Micro Devices), Synertek, and WDC, and none of them include this in their "absolute maximum" spec.s.

I tried a Rockwell VIA recently however, and found that a shorted output line would source and sink around 100mA. I put the DMM on the 200mA range with one probe on ground (for measuring pull-up current) or Vcc (for measuring pull-down current), and put the other probe on the output just long enough to see a stable current reading, which was only a second or two. I didn't leave it shorted long enough to see how long it would take to damage it; but the more I/O bits you abuse at once, the more vulnerable the part will be.

I had a situation years ago where a 6520 (or maybe 65c21?) PIA was used in a printer interface, and certain conditions produced a latch-up a few times per year. The computer and the printer were powered up by different power supplies plugged into different AC circuits across the room, so maybe there was some kind of spikes that caused the latch-up. The PIA would get so hot it would burn you to touch it, but it never sustained any damage. We never did figure out exactly what caused the problem.

As far as adding resistors: That's rather unconventional and not normally necessary. When coupled with the normal, unavoidable, stray capacitances in any circuit, resistors produce unwanted delays and slow things down.

If the VIA pin is an output, the maximum resistor value will be more-or-less 0.7V divided by the amount of current the input it's connected to needs for a logic-low state, since in TTL, the logic-low state is where more current is passed. So if the VIA can do enough current with its output at 0.1 volts, and the input of an LSTTL part (perhaps the heaviest valid load you might connect) needs 0.8V or lower at 0.36mA, you have (0.8-0.1)V/.36mA, or about 2K. The time constant produced by 2K and 15pF is only 30ns, and even a few time constants to get you down to a logic 0 often won't really cause problems with the applications you're probably interested in using the VIA for. It wouldn't hurt the VIA a bit however to limit each pin's output current to 10mA per bit instead, which when dealing with shorts, would mean 470 ohms. There shouldn't be any reason to go above 470 ohms, and you could probably go below that with no problem, even if two outputs were accidentally connected together.

Garth

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?