Another unpublished draft. sheesh!

-----

Got simpler, then went all-out... We're talking multiplexing and Write-Complete testing.

Pardon the Digital logic, Dunno how to make a rotating drum-ROM with magnets and reed-switches in this simulator. And... well... I got lazy with the gates.

RunIt!

The key, of course, is the refresh/read circuitry, at the upper-right which consist of a single relay and a few resistors.

Herein, I think I've chosen somewhat realistic components, etc... 5V 160mW bulbs, a 5V relay with 600ohm coil, 10Hz refresh.

A few last-minute changes, can't say they've been fully-pondered, but... oh snap... oscillation again? No.

Another key, of course, is finding timings/values that cause the off bulbs' heat/resistance to asymptote, rather'n keep climbing, due to periodic measurement current. I think it's doable with enough slack to not require a lot of precision.

Really, this is a mess, the concept is pretty simple, just need some old cylindrical container, a motor, magnets, and reed-switches for all them counters, diodes, and associated relays.

The write-complete circuitry, though... That's something else. That's a lotta relays and gates. And may be completely unnecessary, depending on the goal. E.G. was thinking of a 5x7 grid where pixels are set by hand. In which case the "set" and "clear" switches could be used manually with a keen eye, and all that testing circuitry removed.

But also have a "big idea" to make my millions from this... so that'd need automation for writing (which, really, could just use a R-C delay circuit...)