EPROMs & obsolescence

I wouldn't go as far as calling it foolhardy, but it's a significant risk you don't want to take unless you really have to. And if you do, you'll have to do some risk management. As in: stock up on the part the moment you finalize that design decision.

That said, it may not be a particularly wise choice to design in *any* EPROM these days, unless you have very strict requirements mandating them. Flash or EEROM is just so much more easy to use during development --- and for final products, you can always use OTPs instead.

--
Hans-Bernhard Broeker (broeker@physik.rwth-aachen.de)
Even if all the snow were burnt, ashes would remain.
Reply to
Hans-Bernhard Broeker
Loading thread data ...

Considering the pace of chip manufacturing today, ANY chip you buy today IS obsolete! Really,something 'smaller,better,more features' is always touted the week after my projects go to the fab shop. If you're thinking of supporting your products in the future buy a good supply of EPROMs now and stockpile them. Depends on what 'future' means, 1 year,5 years,20 years? I've got equipment with 2K eproms running after 20 years. No need to update and clients are happy. Now if you standardize your products to use the same 32K EPROM, then you're doing good. Common parts for several products reduces inventory, eases designs,etc.

I intend to keep at least one PC with an ISA slot for my EPROM reader/writer as new motherboards seem to be 'missing' this slot !

hth jay

Reply to
j.b. miller

denis forde schrieb:

I would suggest using a 32 pin socket. Since those devices are pin compatible, you can then insert anything from 32kx8 to 1Mx8. (provided you program the right section of the larger types).

--
Dipl.-Ing. Tilmann Reh
Autometer GmbH Siegen - Elektronik nach Maß.
 Click to see the full signature
Reply to
Tilmann Reh

Morning all,

Is it foolhardy to design a new product today with external 32kx8 EPROM? I see that all EPROMs smaller than 32Kx8 are obsolete. How quickly will this and the larger ones follow?

Denis Forde

Reply to
denis forde

Direct part number replacement is a problem, and certainly windowed devices are getting harder to find.

There are however, FLASH or MTP versions that are EPROM compatible ( eg Winbond 27E512 - pgms like FLASH, but replaces EPROMS, no WE pin ). This can replace any 28 pin EPROM, with minor attention to unused pins, and memory mapping.

There are even MTP (Flash) devices like the W27C520, that target NEW designs using external EPROM flows, and swallow the '573 into a TSSOP20 package.

-jg

Reply to
Jim Granville

Why on earth would you want to use an EPROM instead of a Flash?

--
Best Regards
Ulf at atmel dot com
 Click to see the full signature
Reply to
Ulf Samuelsson

So that you can be SURE the system cannot 'accidently' self program ? That's why some FLASH devices are designed to need high voltage to PGM, and are designed to NOT pgm in-circuit. ( eg W27E512, W27C520 ) The underlying process is essentially FLASH, but the process flow handling, and integrity is that of EPROM.

-jg

Reply to
Jim Granville

By tying WP* low, most Flash parts can be rendered just as non-volatile as an EPROM, and perhaps more than a windowed part covered with a $0.05 label.

--Gene

Reply to
Gene S. Berkowitz

My airplane has a LORAN from ~1989 (was in the plane when I got it, don't have the records in front of me to say when). It (and several newer models) can no longer be updated because the process involves replacing an EEPROM that is no longer made.

--
Ben Jackson

http://www.ben.com/
Reply to
Ben Jackson

Why does an (E)EPROM need to be replaced to be updated?

Reply to
Everett M. Greene

Yes, so this is not a good reason. Some Flash memories are designed to need high Vpp, because the manufacturer couldn't do a device that programs at Vcc...

--
Best Regards,
Ulf Samuelsson   ulf@a-t-m-e-l.com
 Click to see the full signature
Reply to
Ulf Samuelsson

Winbond, Macronix, SST etc can clearly make both 5V and Higher Vpp devices, so 'couldn't do a device' as a reason is not valid ?

They make EPROM pinout/FLOW FLASH because they see a clear market need for this, and it makes good sense :

  • Can ship in std 28 Pin EPROM footprint, so can target ALL existing OTP EPROM sockets with their MTP devices.
  • More modern process gives faster pgm times.
  • No charge pump needed on die = Smaller device area
  • No Charge pump means spikes CANNOT repgm the FLASH
  • Lower cycle specs, == less testing cost.

eg SST's 27SF512 is a Nov 2003 data sheet, and specs a 1.4 sec pgm time for a 27SF512 device. ( appx 5x faster than the OTP EPROM AT27C512R it might replace )

Now, not ALL flash vendors have seen this market, so that may explain Ulf's comments ?

Some customers in the automotive sector specify a PGM pin (Typically high voltage) so that it is _impossible_ to PGM the flash 'by accident'. I think Intel uses high voltage optionally, to give faster programming speeds.

FLASH memories have WRN triggered state engines : the internal pulses are much wider than the write strobes, so there is no solid hardware lockout - if the device glitches, you can get unwanted PGM cycles - even with WRN tied.

-jg

Reply to
Jim Granville

I think the risc in using 32Kx8 EPROMSs is very low. Direct equivalents in flash is already available. There are also quite many sections of the market where a software change done with the replacement of a physical device is almost mandatory. Most of the products we use that are designed and manufactured in asia use EPROMs for at least part of the firmware. Often font/language information etc.

Regards Anton Erasmus

Reply to
Anton Erasmus

OK. I agree that for companies without an EPROM process, this allows them to adress an existing market and improved programming time is nice.

The original question from Denis was if it was OK to use an EPROM, but still no reason why the EPROM was wanted in the first place. Maybe he has the problems you mentioned, and maybe not. Maybe a totally different solution, like a serial EEPROM is better for him in the end.

No Still selling EPROM in volume unfortunately... I guess if customers vote with their feet, then Atmel will react.

accident'.

speeds.

--
Best Regards,
Ulf Samuelsson   ulf@a-t-m-e-l.com
 Click to see the full signature
Reply to
Ulf Samuelsson

I forgot to also mention : better programming yield than EPROM.

True, and from this thread, he has found out

- EPROM footprints are fine for new designs, can use MTP

- EPROM flows are fine for new designs, can use MTP

- Atmel still make EPROM in volumes

There may be applications, even today, where an EPROM footprint/flow is the right design decision. The choice of true EPROM, OTP, or MTP (FLASH core, EPROM pinout), is up to the customer/puchasing officer.

Over to those customers ::

Mouser prices from a web scan : SST27SF512-70-3C-NH SST - MTP PLCC-32 64KX8,70ns $1.060/1+ AT27C512R-70JC Atmel EPROMs PLCC-32 64kx8 70ns $2.030/1+

-jg

Reply to
Jim Granville

... snip ...

As an historical note, Intel developed semi-conductor memories, including both static and dynamic RAM, and EPROMs. They then developed microprocessors as a means of creating a market for those memories. The tail wags this dog.

--
Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net)
   Available for consulting/temporary embedded and systems.
 Click to see the full signature
Reply to
CBFalconer

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.