Denial

I'm thinking of using a R6532 RIOT instead of a VIA for my I2C board (to give some extra RAM to play with). Reading the datasheet I notice that the PB lines are push-pull. Is this different from the VIA? Are these safe to use for bit-banging where multiple devices can pull the line low?

After a quick look they seem to be identical. Both 6522 and 6532 port 'B' are push/pull although 6522 in some cases only can source 1mA and 6532 seem to be able to source 3mA. Port 'A' has an open drain configuration with a pull-up on both chips.

I checked the Rockwell 6522 data sheet which has nice figures (figure 2 and 3) of the output configurations. I couldn't find any figures for the 6532 chip, but the text are nearly identical in this respect.

That said I might be hesitant to deliberately short the push/pull stages. Might be better to switch the buffer between input and output, with an output set to 0.

TLovskog wrote:After a quick look they seem to be identical. Both 6522 and 6532 port 'B' are push/pull although 6522 in some cases only can source 1mA and 6532 seem to be able to source 3mA. Port 'A' has an open drain configuration with a pull-up on both chips.

Thanks for confirming that.

I checked the Rockwell 6522 data sheet which has nice figures (figure 2 and 3) of the output configurations. I couldn't find any figures for the 6532 chip, but the text are nearly identical in this respect.

Yes, shame there wasn't similar diagrams for the 6532, I'd have been less confused

That said I might be hesitant to deliberately short the push/pull stages. Might be better to switch the buffer between input and output, with an output set to 0.

That's what my code does at present but I was concerned that I'd damage the 6532 if it and another device tried to drive the line low at the same time.

I think I'll switch to using Port A, open drain will pull up is what I need anyway.

The problem is not if many pins sink the output to "0". The problem is when one or more sink "0" and one or more source "1".

Using port A fixes that.