GBSCSI - A cheap(er?) solution to replacement storage

[GRudolf94]

Ooh you made me realize me something I said was wrong. LVT series can sink 64 mA and source 32 mA, opposite of what you said. So the resistance when driving high is more like 10-20 ohms, not 5-8 ohms like I said. This makes the problem a bit better but still not great.

It could be the opposite like you said. I haven't looked at that datasheet in a good while.

Running the '245 at 2.85V sounded good at first but apparently the drive strength is not as well specified and may only be 24 mA:View attachment 78211
Too bad they don't have the "0.55V" or "0.6V" spec for Vcc = 2.7V. If they did it might be 48 mA and we'd know this solution would work well.

Hm. Indeed. It's a value that can be experimentally obtained, though. If I solidly characterized that value, and it met spec, then I'd have no qualms using that.

As long as the 125 and your MCU are both tristated, no current would flow.

Realistically since we're dealing with CMOS parts it's probably a non-issue even when the circuit is active, but... I'd still like to think it thru carefully. Ideally stick the device on something connected to a SCSI analyzer, which I also don't have.

Like Zane said above, having a worst-case test setup (and time to prototype solutions) would be ideal. Sadly, I don't get much thinking/experimenting time as of lately.