Cem Eryasar wrote:
> I have to store some runtime values somewhere so that I do not lose them
> over power downs and restarts. However on my tiny LPC-2138 board there
> is no external EEPROM for my use. Therefore I am thinking of using the
> unsued flash area of the 2138 as a tiny file storage segment. From the
> chip manual point of view it seems possible with IAP commands. But there
> is not much example out there.

There is an example for IAP in the file-archive of the LPC2000
yahoo-group. You have to sign up to this group with a yahoo-account to
get access to the file-archive. I have not used this code so far.

For my "T-Clock" project I have implemented an "EEPROM-emulation" inside
the LC2106 flash. See
http://www.siwawi.arubi.uni-kl.de/avr_projects/arm_projects/ (files
iap*.*). It's not a very well-structured code and from the days when I
started using LPC controllers but at least it worked in my project and I
know that others used the code too.

> Has anybody tried that? Is it useful?

Obviouly I have tried it for the "T-Clock" and it worked. Be aware the
you have to write a sector (several bytes, depends on LPC) and you have
to delete a flash-bank before you can write new data in a sector of a
bank. At least on an LPC2106 it's not possible to delete only a sector.
It's not a "byte"-access as in an EEPROM.

The read-modify-write approach in the T-Clock-code:

- read a sector from the flash-bank into a buffer
- modify buffer with updated values
- if bank is full erase bank, reset sector-number
- write to next sector in bank (or the first after erase)

Inside the sector I keep track which of the sectors holds the "newest"
values for the read-routine. This is basicly a emulation of an EEPROM of
sector-size bytes. This approach avoids that the bank has to be erased
every time new data gets stored.

> Any ideas for alternatives?

For example an external EEPROM with I2C- or SPI-Interface.


Hope this helps,
Martin Thomas

Hi Martin,

> There is an example for IAP in the file-archive of the LPC2000
> yahoo-group. You have to sign up to this group with a yahoo-account to
> get access to the file-archive. I have not used this code so far.
>
> For my "T-Clock" project I have implemented an "EEPROM-emulation" inside
> the LC2106 flash. See
> http://www.siwawi.arubi.uni-kl.de/avr_projects/arm_projects/ (files
> iap*.*). It's not a very well-structured code and from the days when I
> started using LPC controllers but at least it worked in my project and I
> know that others used the code too.

I looked at your code. You disabled the PLL prior to calling IAP
function. Then after the IAP function you enabled it. Why did you need
that? I did not see anything like that on the LPC2138 user manual.

I wrote a simple test and it runs without a problem so far without
disabling the PLL. I tested the write/erase cycles with and without the
PLL disabling, but I did not notice any difference. The cpu runs at
14Mhz crystal with 2x PLL multiplier.

Thanks
Cem

Cem Eryasar wrote:
>
> Hi Martin,
>
>> There is an example for IAP in the file-archive of the LPC2000
>> yahoo-group. You have to sign up to this group with a yahoo-account to
>> get access to the file-archive. I have not used this code so far.
>>
>> For my "T-Clock" project I have implemented an "EEPROM-emulation" inside
>> the LC2106 flash. See
>> http://www.siwawi.arubi.uni-kl.de/avr_projects/arm_projects/ (files
>> iap*.*). It's not a very well-structured code and from the days when I
>> started using LPC controllers but at least it worked in my project and I
>> know that others used the code too.
>
> I looked at your code. You disabled the PLL prior to calling IAP
> function. Then after the IAP function you enabled it. Why did you need
> that? I did not see anything like that on the LPC2138 user manual.
>
> I wrote a simple test and it runs without a problem so far without
> disabling the PLL. I tested the write/erase cycles with and without the
> PLL disabling, but I did not notice any difference. The cpu runs at
> 14Mhz crystal with 2x PLL multiplier.

Hmm, it has been to long since I have written the code. Maybe there is
something about this in the LPC2106-manual (LPC2106 has "IAP" and
LPC2138 has "IAP2"). IRC the IAP-calls receive the target-frequency as
parameter. I don't know now if this is the frequency "before" or "after"
the PLL. I expect the frequency-information is used for the timing of
the flash write/erase-sequence. Double-check the datasheet about the
IAP-frequency and from where it is sourced.  It may work for you now but
you may also "overclock" something.

Martin Thomas

Thanks a lot for all the support.
Really appreciated your time and patience.
Regards
Cem