Lpc2000

Hello,
I am using LPC2138 for industrial controller application.
I have no option but to use EEPROM for data storage of 256 byte at every 10
sec.
I am planning to use higher capacity 32Kb EEPROM and write 256 bytes all
over EEPROM, and once it is full, then erase entire EEPROM and start writing
again. This way I have to erase entire EEPROM after 128 write cycles.
My question is, will this enhance the life (100,000 cycles) of EEPROM by 125
times?

Best Regards,

Mukund Deshmukh.
Beta Computronics Pvt Ltd
10/1, IT Park, Parsodi,
Nagpur-440022
Cell - 9422113746

Mukund Deshmukh schrieb:
> I am using LPC2138 for industrial controller application.
> I have no option but to use EEPROM for data storage of 256 byte at every 10
> sec.
> I am planning to use higher capacity 32Kb EEPROM and write 256 bytes all
> over EEPROM, and once it is full, then erase entire EEPROM and start writing
> again. This way I have to erase entire EEPROM after 128 write cycles.
> My question is, will this enhance the life (100,000 cycles) of EEPROM by 125
> times?

If I calculate correct with 100k cycle, the EEPROM will wear out in
416days.
So an EEPROM 128times larger seems to be to large :-)
What livetime do you expect for your HW ?

-- 
42Bastian

May be FRAM (ferroelectric RAM) could help you ? 
It's a kind of nonvolatile memory with almost unlimited number of write-erase cycles.
See for example:
http://www.ramtron.com/doc/Products/nonvolatile

Regards,
 Max

--- In lpc2000@yahoogroups.com, Mukund Deshmukh <betacomp_ngp@...> wrote:

>
> Hello,
> I am using LPC2138 for industrial controller application.
> I have no option but to use EEPROM for data storage of 256 byte at every 10
> sec.
> I am planning to use higher capacity 32Kb EEPROM and write 256 bytes all
> over EEPROM, and once it is full, then erase entire EEPROM and start writing
> again. This way I have to erase entire EEPROM after 128 write cycles.
> My question is, will this enhance the life (100,000 cycles) of EEPROM by 125
> times?
> 
> Best Regards,
> 
> Mukund Deshmukh.
> Beta Computronics Pvt Ltd
> 10/1, IT Park, Parsodi,
> Nagpur-440022
> Cell - 9422113746
>

>
> If I calculate correct with 100k cycle, the EEPROM will wear out in
> 416days.
> So an EEPROM 128times larger seems to be to large :-)
> What livetime do you expect for your HW ?

So my assumption is correct. If I can reduce the size of data chunk to 128
bytes, life would be doubled.
Yes I am looking for 3 years life.

Will this be sound design, considering the 100K cycles may be reduced due to
higher temperature or manufacturing defects.

Best Regards,

Mukund Deshmukh.
Beta Computronics Pvt Ltd
10/1, IT Park, Parsodi,
Nagpur-440022
Cell - 9422113746

> May be FRAM (ferroelectric RAM) could help you ?
> It's a kind of nonvolatile memory with almost unlimited number of
write-erase cycles.
> See for example:
> http://www.ramtron.com/doc/Products/nonvolatile
>

I have already thought of  this possibility, but was doubt full about
availability and cost.

BTW is it possible to use internal ram of LPC2138 in Non volatile mode with
RTC battery?

I think I can steal a few location of Alarm setting, but those wont be
enough.

Best Regards,

Mukund Deshmukh.
Beta Computronics Pvt Ltd
10/1, IT Park, Parsodi,
Nagpur-440022
Cell - 9422113746

> BTW is it possible to use internal ram of LPC2138 in Non volatile mode with
> RTC battery?

AFAIK it is not.
As for the cost and availability of FRAM, here in Russia I have found Ramtron's FM25L04 (512 bytes) for as low as $0.85 by the piece and in-stock, it may be different for your place though.

Regards,
  Max

Mukund Deshmukh schrieb:

> So my assumption is correct. If I can reduce the size of data chunk to 128
> bytes, life would be doubled.
Try if you can pack the data.

> Yes I am looking for 3 years life.
With 128bytes there are still some days missing :-)
And maybe you should give some percent plus.

In the end it is a matter of cost (your time <-> chip).
If you don't have to care for the chip price, I'd go for one
which can hold 2 times lifespan or as already suggested go for
FRAM (these are also faster).

> Will this be sound design, considering the 100K cycles may be reduced due to
> higher temperature or manufacturing defects.

Your supplier should have data on this. AFAIK the 100k is the minimum in
a standard enviroment (what standard you have to ask the manufacturer).
Often it is 20 degree Celsius and some x percent moisture.



-- 
42Bastian

Actually, I calculate the lifetime at 1481 days:

128 writes * 10 sec/write => 1280 seconds per erase cycle
1280 * 100000 cycles => 128,000,000 seconds which is about 1481.48 days
This is about 4 years.

On 5/19/06, 42Bastian Schick <bastian42@...> wrote:
>
> Mukund Deshmukh schrieb:
> > I am using LPC2138 for industrial controller application.
> > I have no option but to use EEPROM for data storage of 256 byte at every
> 10
> > sec.
> > I am planning to use higher capacity 32Kb EEPROM and write 256 bytes all
> > over EEPROM, and once it is full, then erase entire EEPROM and start
> writing
> > again. This way I have to erase entire EEPROM after 128 write cycles.
> > My question is, will this enhance the life (100,000 cycles) of EEPROM by
> 125
> > times?
>
> If I calculate correct with 100k cycle, the EEPROM will wear out in
> 416days.
> So an EEPROM 128times larger seems to be to large :-)
> What livetime do you expect for your HW ?
>
> --
> 42Bastian
>
>
>
>
> Yahoo! Groups Links
>
>
>
>
>
>
>
>


-- 
Jim Parziale
nuncio.bitis@...
Malden, MA


[Non-text portions of this message have been removed]

----- Original Message ----- 

From: "theothervmax" <vmax@...>
To: <lpc2000@yahoogroups.com>
Sent: Friday, May 19, 2006 10:02 AM
Subject: [lpc2000] Re: EEPROM


>
> May be FRAM (ferroelectric RAM) could help you ?
> It's a kind of nonvolatile memory with almost unlimited number of 
> write-erase cycles.
> See for example:
> http://www.ramtron.com/doc/Products/nonvolatile

I was going to suggest that. It seems the best solution to the problem.

Leon 

---
[This E-mail has been scanned for viruses but it is your responsibility 
to maintain up to date anti virus software on the device that you are
currently using to read this email. ]

Jim
> Actually, I calculate the lifetime at 1481 days:
> 128 writes * 10 sec/write => 1280 seconds per erase cycle
> 1280 * 100000 cycles => 128,000,000 seconds which is about 1481.48 days
> This is about 4 years.

I did this for his original size and *hmm* did it wrong.

 >> If I calculate correct with 100k cycle, the EEPROM will wear out in
 >> 416days.

100k cycles à 10s = 1e6 s = 277h ~= 11days :( not 461

So you're right, the 32k EEPROM will fit the lifetime he wants.

-- 
42Bastian

--- In lpc2000@yahoogroups.com, Mukund Deshmukh <betacomp_ngp@...> wrote:
>
> Hello,
> I am using LPC2138 for industrial controller application.
> I have no option but to use EEPROM for data storage of 256 byte at
every 10
> sec.
> I am planning to use higher capacity 32Kb EEPROM and write 256 bytes all
> over EEPROM, and once it is full, then erase entire EEPROM and start
writing
> again. This way I have to erase entire EEPROM after 128 write cycles.
> My question is, will this enhance the life (100,000 cycles) of
EEPROM by 125
> times?
> 
> Best Regards,
> 
> Mukund Deshmukh.
> Beta Computronics Pvt Ltd
> 10/1, IT Park, Parsodi,
> Nagpur-440022
> Cell - 9422113746
>

Hi,
another option could be to use an EEPROM with 1 Mio write /erase
cycles. I found a brandnew datasheet from Intersil with a 32k EEPORM
and 1 Mio W/E cycles. 
No idea about availability or cost, just a thought.
http://www.intersil.com/data/FN/FN8108.pdf
Bob