FRAM FM24C256 from RAMTRON

Tell us about the cool things you've recently got
victor_pv
Posts: 1650
Joined: Mon Apr 27, 2015 12:12 pm

FRAM FM24C256 from RAMTRON

Post by victor_pv » Mon Sep 11, 2017 4:00 pm

I ordered some of these a couple of weeks ago:
https://www.aliexpress.com/item/10PCS-L ... 0.0.yWvKDH

There are a few manufacturers with part FM24C256, so I trusted the picture that it would be a RAMTRON FRAM rather than an EEPROM, which Fairchild produced with the same name:
http://www.jameco.com/jameco/products/p ... 580fsc.pdf

I had looked at a bunch of FRAM datasheets, both i2c and SPI, and decided to give the i2c one a shot, and failed to notice the one I was ordering was supposed to work at 5V only.
They arrived during this weekend, and are indeed marked RIC (Ramtron). I soldered 1 to a dip adapter, and plugged it thru a level shifter first, and all was good. Next I plugged it without the level shifter (but powered at 5V) and all was good (write and read test). Then powered it at 3V3, and still seems to work fine.
I'm sure is not ideal, and perhaps even a little further voltage drop start's causing issues with reads or writes, but so far so good. I plan to drop the voltage a bit lower, perhaps 2.5V-3V and see what happens, but worst case I can power it at 5V (PU resistors at 5V too) and still use it fine.

If anyone knows of a library specific for i2c FRAM memories let me know. I am using one for eeproms, and increased the buffer lenght in Wire to test that it's not really a relabelled eeprom (since eeproms are organized in blocks and writing more than their internal buffer at once will wrap and overwrite from the start of the block again). I checked the test program I was using in a real eeprom and oculd see the overwrite effects. With the FRAM I tested with really large buffers and large writes (up to 1024), and never saw the overwrite effects, but if anyone has a library or a test for FRAMs, would be happy to test it.

User avatar
martinayotte
Posts: 1222
Joined: Mon Apr 27, 2015 1:45 pm

Re: FRAM FM24C256 from RAMTRON

Post by martinayotte » Mon Sep 11, 2017 4:26 pm

I don't know if this one would work with FRAM, but I usually use that one for EEPROM :

https://github.com/JChristensen/extEEPROM

BTW, last summer, MrBurnette gave me few FRAM samples he got, but they were SPI FM25V10, for those ones, I've updated SPIFlash from LowPowerLab, I've sent them a PR, but they didn't wish to merge it. If someone wish to get it, I still have my fork here :

https://github.com/martinayotte/SPIFlash

BTW2, those FM25V10 are 3.3V ...

victor_pv
Posts: 1650
Joined: Mon Apr 27, 2015 12:12 pm

Re: FRAM FM24C256 from RAMTRON

Post by victor_pv » Mon Sep 11, 2017 5:59 pm

Thanks Martin. The first one is the one I am using right now (with small modifications), but is not optimal for FRAM since it takes page size in consideration, which doesn't exist in FRAM. I have used it rising the page size to it wastes less cycles sending addresses.
Those FM25V10 look interesting, have you done anything with them yet?

User avatar
martinayotte
Posts: 1222
Joined: Mon Apr 27, 2015 1:45 pm

Re: FRAM FM24C256 from RAMTRON

Post by martinayotte » Mon Sep 11, 2017 8:03 pm

victor_pv wrote:
Mon Sep 11, 2017 5:59 pm
Those FM25V10 look interesting, have you done anything with them yet?
Not really other real things except making the tweak to the library.

User avatar
Pito
Posts: 1529
Joined: Sat Mar 26, 2016 3:26 pm
Location: Rapa Nui

Re: FRAM FM24C256 from RAMTRON

Post by Pito » Mon Sep 11, 2017 8:49 pm

I messed with FRAMs (FM25V20) and MRAMs (MR25H40) in past (SPI) and it works with the same protocol as SPI SRAMs, ie 23LC128++.
So you may use the library as-is (maybe 2 lines to edit).
It has got an autoincrement, so after initial 4/5bytes header (1or2 bytes command and 3b address) you stream the bytes in/out full speed..
You write full speed (ie 40MHz SPI) and you read full speed with both FRAMs and MRAMs. Like an SRAM. Unlimited endurance and none-volatile.
Nothing more to add :). Only issue is they are still small sizes :)
PS: Ramtron was aquired by Cypress I think. I got MRAM's from https://www.everspin.com/serial-peripheral-interface .
Pukao Hats Cleaning Services Ltd.

User avatar
mikikg
Posts: 14
Joined: Tue Mar 07, 2017 4:08 pm
Location: Serbia
Contact:

Re: FRAM FM24C256 from RAMTRON

Post by mikikg » Sun Sep 17, 2017 10:40 am

Also take an look on Fujitsu line of FRAMs:
http://www.fujitsu.com/us/products/devi ... index.html

I have used SPI version MB85RS64 and I'm very satisfied how that works.
Due some requirements in my application I have to test it's endurance in real usage scenario, I have passed 10^6 reads/writes at same memory location, after that I give up : )

The only small thing that is missing in this 64Kbit model is Manufacture/Device ID which is useful to read as confirmation that we have valid model and that connection is OK before we do anything else later.

User avatar
Pito
Posts: 1529
Joined: Sat Mar 26, 2016 3:26 pm
Location: Rapa Nui

Re: FRAM FM24C256 from RAMTRON

Post by Pito » Sun Sep 17, 2017 11:24 am

I have passed 10^6 reads/writes at same memory location, after that I give up
I had a chat with a guy close to ramtron several years back, and he told me their actual FRAM endurance is unlimited, the information in the datasheet is extremely conservative under extreme env conditions (as the product was just introduced)..
The MRAMs (magnetoresistive - like today's harddisk technology) is a different story. They are "writing n. unlimited" by default, there is a limitation in the magnetic fields strength the memory can operate in without a data upset, though..
Pukao Hats Cleaning Services Ltd.

victor_pv
Posts: 1650
Joined: Mon Apr 27, 2015 12:12 pm

Re: FRAM FM24C256 from RAMTRON

Post by victor_pv » Mon Sep 18, 2017 3:59 am

I checked one of those Fujitsu datasheets and seems like the claim 10^12,
How long did it take to do 10^6 writes?
If we multiply it by another 10^6. I bet the product would be obsolete before you can complete the test :D
I remember there is a thread in a forum testing endurance of SSD hard drives. I wonder if someone at some point started a FRAM endurance test and is either still running, or completed and the results somewhere.

User avatar
Pito
Posts: 1529
Joined: Sat Mar 26, 2016 3:26 pm
Location: Rapa Nui

Re: FRAM FM24C256 from RAMTRON

Post by Pito » Mon Sep 18, 2017 8:07 am

Ramtron claimed 10^14 writes. One write (1byte) takes, say, 300ns at 40MHz SPI.
So you have to wait 249000 years to wear out a 256kB large Ramtron's FM25V20 FRAM.
@victor: with your I2C you certainly have to wait a dozen of ice ages :)
Pukao Hats Cleaning Services Ltd.

victor_pv
Posts: 1650
Joined: Mon Apr 27, 2015 12:12 pm

Re: FRAM FM24C256 from RAMTRON

Post by victor_pv » Mon Sep 18, 2017 2:10 pm

Well I was thinking more in the line of testing a few positions repeatedly, I don't have that amount of time to spare ;)

There are two papers from NASA on the subject, one a stress test with conditions that could happen in space:
https://ntrs.nasa.gov/archive/nasa/casi ... 036989.pdf

And other for endurance under normal conditions:
https://trs.jpl.nasa.gov/bitstream/hand ... sequence=1

In the endurance test they reach their goal and kept testing, but I didn't find the details on how far the test kept going until finding a failure.
In the stress test they get the chips to fail in most extreme conditions, but in the normal cold/heat cycles none of the failiing.
It seems that as you said the specs are conservative and within operating conditions the memory should exceed it well enough.

There are 2 other papers from RAMTROM and TI about endurance, they pretty much caused accelerated aging with overvoltage to test their specs, which supported their claims of endurance.

Post Reply