4
\$\begingroup\$

We have a system that uses OTP (one-time programmable) EPROMs, specifically the AT27C256. We would like to replace this EPROM with an EEPROM or Flash chip such that we can make our units field upgradeable, more serviceable, etc. Obviously, we want this change to require little to no hardware modifications. These EPROMS are currently the code storage for an 68HCP11 micro.

It looks like the AT29C256 would be a viable option, as we could use it with only one pin change (change Vpp to WE_N). However, it appears that these chips are not made any more and sourcing is very unreliable, especially as we need ~70K of them. As such, we're looking at any other reasonable options.

We've considered adding a daughter board that sits on our EPROM's footprint (PLCC32) and contains another flash chip (AT28 series?) along with whatever conversions are necessary to make it work with our HC11 micro. We've also considered adding an FPGA to the daughter board between the micro and flash such that a series of "magic" reads can perform writes to the chip, but this seems to be getting a bit excessive for a problem many others must have solved. It would, however, allow us to not use jumper wires on our existing boards.

What have others done in this situation?

\$\endgroup\$

1 Answer 1

3
\$\begingroup\$

I'm seeing <5MHz clock rates listed for that micro, so I would be very tempted to get a modern micro with a high enough pin count and 32K bytes of spare internal flash, and emulate it in software. With a 3-wire serial port hanging off the end of course!

Major challenge I see would be 5v - 3.3v conversion. You could probably do it with series resistors, assuming that the output of the 3.3v micro is enough for a reliable read by the old micro (which it usually is, if you aren't in a high-noise environment)

If the clock is low and you don't need the full capacity, you might be able to use an ATMEGA328 running at 5V.

I guess another challenge would be insuring that the add-in comes out of reset first, though you could fix that with a jumper wire to hold the main processor in reset until the new device comes out of power-on-reset.

Of course another option if the processor is socketed is to emulate the processor rather than the EPROM.

\$\endgroup\$
2
  • \$\begingroup\$ I'd use level-translators/FETs instead of series resistors. The space hit is cheap and reliability is paramount. \$\endgroup\$
    – Joel B
    Jul 11, 2012 at 20:23
  • \$\begingroup\$ Plenty of PICs run at 5v native, and I believe you can arrange them to multiply up an external clock for internal purposes. \$\endgroup\$
    – pjc50
    Jul 12, 2012 at 16:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.