64K EEPROM Sources

rich_mays · Post by **rich_mays** » 07 March 2007, 14:02 PM

Does anyone know where I can buy retail quantities of the Atmel AT25HP512 EEPROM? Digikey has a quantity 3000 minimum.

Thanks,
Rich

Post by **dkinzer** » 07 March 2007, 14:26 PM

I saw on Atmel's site that the AT25HP512 has been replaced by the AT25F512A. The latter has the advantage of being byte-writable whereas the former only supported page writes. It also has a shorter write time. The ZX-40 can be configured for either mode and for various page EEPROM sizes.

Unfortunately, the AT25F512A is only available in the SAP-8 and SOIC-8 SMT packages. The SOIC-8 package is available at Digi-Key.

For prototyping purposes you can find SOIC-8 to DIP-8 adapters in various places on the Internet and on eBay as well.

For a while, Mouser had the ST Micro M95512 in stock but it no longer is. This device is similar to the AT25F512A and is available in SOIC-8 and TSSOP-8 packages.

dlh · Post by **dlh** » 07 March 2007, 14:32 PM

rich_mays wrote:Does anyone know where I can buy retail quantities of the Atmel AT25HP512 EEPROM? Digikey has a quantity 3000 minimum.

Mouser was listing the ST M95512 as stock and I bought 100 of them. I see now that they again list it as non-stock with 2500 minimums. I'll see if I can find out what changed.

Catalyst has a 64K SPI EEPROM listed as a future product but I haven't been able to get any details on it.

rich_mays · Post by **rich_mays** » 08 March 2007, 5:59 AM

Thanks for the information. I'm surface mount phobic, but I suppose I'll have to modernize.

Post by **dkinzer** » 08 March 2007, 7:43 AM

rich_mays wrote:I'm surface mount phobic[...]

I was as well. It's not that difficult once you've soldered a few and you develop your own technique. There is a lot of information available on the Internet describing techniques and giving tips for manual SMT soldering with wire solder as well as with paste. If you can justify it, I'd highly recommend getting a binocular zoom/boom microscope (say, 7.5x to 20x). Otherwise, a 10x jeweler's loupe works quite well for inspection.

Once you've gotten past the initial apprehension and designed and built your first SMT board you'll wish that you had done so earlier.

GTBecker · Post by **GTBecker** » 08 March 2007, 16:24 PM

I'm surface mount phobic[...]

Another small hurdle for me was hot air. One _can_ solder and remove
impossible parts, and it's not hard afterall.

Tom

twesthoff · Post by **twesthoff** » 09 March 2007, 4:45 AM

Since there is a discussion on surfacemount soldering, I thought I would pass on something that works well for me to remove surfacemount IC's. I use it to replace the ZX-44 with the ZX-44a on my boards when upgrading. You can also easily remove large 100 pin and above surface mount chips. All you need is a normal soldering iron and a tweezer or dental pick.

See: http://www.emulation.com/046/

It is about $20, but well worth it... It really is easy.

It is primarily some solder with a very low (136 F) melting point, and some liquid flux. You apply a large amount of the solder to all of the pins of the device. The thermal mass of the solder slows down the cooling so you have enough time to lift off the chip before it cools to the very low melting point of the solder. Then you put a drop of flux on a q-tip, reheat the solder, and the excess solder on the board soaks up into the q-tip.

You almost have to see it to believe it. It works well and it is easy.

mdown · Post by **mdown** » 09 March 2007, 7:29 AM

ChipQuick has leadfree kits also see

http://www.chipquik.com/store/nolead.htm

-MikeD

Post by **dkinzer** » 09 March 2007, 8:10 AM

twesthoff wrote:I thought I would pass on something that works well for me to remove surfacemount IC's.

I've used two different methods. The first one is non-destructive and uses a hot air gun (e.g. a paint stripping gun or one made for the specific purpose) to heat the entire chip. A dental pick or similar instrument under one corner of the chip is used to lift the chip when the solder melts and then tweezers are used to completely remove the chip. Of course, the heat gun needs to be carefully manipulated to avoid overheating the chip or affecting nearby chips.

The second method is destructive and is therefore only useful when the chip is known to be defective or you are willing to sacrifice it. The advantage is that it is quicker and less likely to affect nearby circuitry. In this method, an Exacto knife is used to shear off the leads where they enter the chip's package. This is fairly easy to do if you shear one or two leads at a time. When you get down to the last few leads it is recommended to hold the chip down with tweezers or a probe. When all leads have been sheared, the chip can be lifted off and then a soldering iron can be used to desolder all of the leads. Usually, you can pick up several of the leads at a time with the iron.

In both cases, solder wick can be used to remove the excess solder from the pads in preparation for reinstallation.

mikep · Post by **mikep** » 09 March 2007, 8:24 AM

I think we got a little off topic here. It might be better to start a separate thread on SMD soldering.

I have been trying to get some 64K EEPROMs for a while now and have been frustrated by price (over $6) or minimum quantities. I also missed the window for the ST M95512.

dkinzer wrote:I saw on Atmel's site that the AT25HP512 has been replaced by the AT25F512A.

This device is a 3.3V device so voltage level translators and a 3.3V regulator are needed. I have a prototype schematic and PCB layout for this that all fits into an area 0.5" square (i.e. roughly size of 8 pin DIP).

dkinzer wrote:For prototyping purposes you can find SOIC-8 to DIP-8 adapters in various places on the Internet and on eBay as well.

I have SOIC-8 to DIP-8 adapters for sale on my website. They are only 60 cents each and jiffy envelope snail mail can be arranged to reduce shipping costs on this item.

There is also one other alternative that has not been discussed. The recently announced ZX devices based on ATmega128 and ATmega1281 support up to 60K of ZBasic program memory using their internal flash.

dlh · Post by **dlh** » 30 March 2007, 12:44 PM

Digikey now stocks the Microchip 25LC1024. While it has 128KB, the ZX chips can only address 64KB. It's not available in single piece quantities but the minimums are not outrageous. The SOIC-8 version is a wide body so this may also be a factor.

Don had indicated earlier that it looked like it would work but I'm not sure how deeply he looked as it wasn't really available yet when we discussed it. It uses 256 byte page writes so that may be a gotcha.

Post by **dkinzer** » 30 March 2007, 16:40 PM

dlh wrote:Don had indicated earlier that it looked like it would work but I'm not sure how deeply he looked as it wasn't really available yet when we discussed it. It uses 256 byte page writes so that may be a gotcha.

I have not yet seen an SPI EEPROM with capacity greater than 64KB that will work as the program store (i.e. in place of the AT25256A). In all cases that I've looked at, the larger devices require three address bytes to be sent but the ZX only sends two. The VM could be modified so that that aspect is configurable but that would slow down every access that requires the address to be sent - not a good tradeoff, I think.

The larger devices can be accessed, however, using SPICmd().

mikep · Post by **mikep** » 07 April 2007, 17:09 PM

I have been able to source a supply of the ST Microelectronics 64K Byte SPI EEPROM (part number M95512-WMN6P).

This can be ordered as an option with the Oak Micros ZX-24e/ZX-24ae ZX devices or stand-alone for use with an existing microcontroller. Because this chip is in a SOIC-8 package, it is available asis in a tube or mounted on a PDIP-8 header board.

Check the Oak Micros shop. I also offer low cost United States Post Office first class mail for orders under $25.

dlh · Post by **dlh** » 12 April 2007, 6:51 AM

mikep wrote:I have been able to source a supply of the ST Microelectronics 64K Byte SPI EEPROM (part number M95512-WMN6P).

How reliable is your source? I've put my project on hold for fear I couldn't find a source for less than OEM quantities.

Will you offer any discount for quantity? Once I use up the 100 I got before Mouser changed their terms, I'll buy SOIC in 25, 50 or 100 piece lots depending on the price.

mikep · Post by **mikep** » 12 April 2007, 15:27 PM

dlh wrote:
mikep wrote:I have been able to source a supply of the ST Microelectronics 64K Byte SPI EEPROM (part number M95512-WMN6P).
How reliable is your source? I've put my project on hold for fear I couldn't find a source for less than OEM quantities.

Will you offer any discount for quantity? Once I use up the 100 I got before Mouser changed their terms, I'll buy SOIC in 25, 50 or 100 piece lots depending on the price.

The source is reliable but they only have a few in stock. Leadtime from ST is quoted as 16 weeks. And there is no telling if ST will continue to make these chips. Those are the problems with single sourcing of components.

I decided to buy a relatively small quantity and mount them on a header board as a service to other ZX users. Unfortunately no-one has yet decided to buy one

Another alternative to your problem is to use a mega128/mega1281-based ZX device that has up to 60K of program memory in the AVR flash. Oak Micros' ZX-128e/ZX-1281e are socket friendly which I believe is a requirement for roZetta. I could supply these devices without the RAM extension if desired. The ZX-128e/ZX-1281e offer one RS232 compatible serial channel and another high-speed hardware UART as well as the standard ZBasic software-based serial channels (COM3-6). Unfortunately it would require some slight redesign of your board.

ZBasic Forum