2532 / 2732 Adaptor Board

[KenLowe]

Does this look ok if I want to use a 2732 in place of a 2532 in the Atom? In particular, I've wired:

2532 pin 12 (Gnd) to 2732 pin 18 (nCE)
2532 pin 18 (A11) to 2732 pin 21 (A11)
2532 pin 20 (PD/nPGM) to 2732 pin 20 (nOE)
2532 pin 21 (VPP) left floating (it's connected to +5V on the Atom motherboard)

Schematic

PCB layout

3D render

[danielj]

I'd be tempted to be a bit more elaborate so you can use sane, and available, 28pin eproms

https://www.radix.co.uk/download/2532-a ... asheet.pdf

d.

[KenLowe]

Hmmm. They've done the opposite from me with the connection to nCE & nOE on their newer 27C512 EPROM.

On my 27C32 EPROM I've tied nCE to GND, and I'm driving nOE from pin 20 of the 2532 socket. Whereas they've tied nOE to GND and are driving nCE from pin 20 of the 2532 socket.

I wonder which is best??? I went with toggling nOE because the timing diagram suggested that I needed to set nOE low AFTER nCE was set low. It's maybe not that important.

Regarding using a larger IC, yes, that is definitely an option. Particularly if I need to share the same Atom IC24 socket for two different ROM images:

viewtopic.php?p=410642#p410642

[1024MAK]

On modern CMOS devices, it's less critical as they are rather faster than older parts.
For older parts, an EPROM (or ROM) will respond quicker if /CS (/CE) is tied low and you use /OE (output enable) to control the output buffers.

The disadvantage of having /CS (/CE) always low is that if the chip has a lower power standby mode, it will never go into this mode. But that's not normally a problem for mains powered systems.

For slow parts, using /CS (/CE) may cause problems with timing, depending on how slow the device is.

Mark