00:20 <sorear> ATF15xxBE does not support 5V?

00:23 <sorear> no mention of DDR but the max clock is high enough this may not matter (unless that means it might need a dedicated PLL)

00:27 <sorear> 1504 has the best pins/$

00:29 <sorear> although the AS is a fair amount slower than BE

00:30 <whitequark> you can use negedge clocks

00:33 <sorear> would there be any frontend circuitry beyond esd protection?

10:22 <_whitenotifier> [glasgow] freddii opened pull request #258: fixed wrong spelled words - https://git.io/JLbKP

17:20 <d1b2> <sharebrained> @NF6X Sounds like some of our Glasgow use cases are the same (including SCSI). I didn't closely all of the last 500 or so messages, but my only remaining concern for Glasgow in the case of SCSI or Shugart/floppy emulation or controller is the total drive current available per pin group (I/O bank). Each bank has its own 150mA regulator, which if asked to drive more than three signals at the SCSI maximum sink current, will theoretically crash your

17:20 <d1b2> supply rail or put you in current limiting. If, in your experiments or mine, that is an actual problem, I've been putting together an add-on "kilt" that would provide cable connectors for 34 and 50-pin floppy variants, SCSI, and any other interesting retro-bus I can think of, along with a separate power supply input, appropriate pull-ups, and drivers to achieve 48mA on each open-collector output.

19:13 <d1b2> <NF6X> Cool. From previous discussion here, it sounds like the Glasgow drives should be able to handle sinking the required current. Extra power on the addon board may be needed if it also needs to source a lot of current, e.g. for providing terminator power.

19:17 <d1b2> <Attie> individually, yes I/O pins should be able do manage the +/- 50mA... but collectively (i.e: all 8x of a port = 400mA) you may run into issues, as @sharebrained says...

19:21 <d1b2> <Attie> *to