Author |
Message |
chresan

Joined: Jul 11, 2006 Posts: 150 Location: Karlsruhe, Germany
|
Posted: Tue Jan 16, 2007 6:43 am Post subject:
nm1 memory endurance question |
 |
|
Hi,
as part of the Nomad - modular editor project I am writing a test suite to simulate user interaction with the nm1. Now I fear that the massive amount of memory write operations might cause damage to the memory and finally will decrease the lifetime of my synthesizer. I know that flash memory generally does not allow many erase cycles. Has someone specifics about the nm1 memory and can help me estimate this problem?
Thanks in advance,
Christian _________________ ----------------------
http://nmedit.sf.net |
|
Back to top
|
|
 |
blue hell
Site Admin

Joined: Apr 03, 2004 Posts: 24446 Location: The Netherlands, Enschede
Audio files: 297
G2 patch files: 320
|
|
Back to top
|
|
 |
chresan

Joined: Jul 11, 2006 Posts: 150 Location: Karlsruhe, Germany
|
Posted: Tue Jan 16, 2007 10:07 am Post subject:
|
 |
|
Hi Jan,
thank you for the info. I was able to find the data sheet:
http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/21503g2.pdf
It says that 1,000,000 program/erase cycles per sector are guaranteed.
I assume that selected patches are copied to a fixed address space (Slot A-D), what means there is a accumulation of write operations at the particular locations. That sounds not good .
Well then, do you know how much it would cost to replace the flash memory, if I had to ?  _________________ ----------------------
http://nmedit.sf.net |
|
Back to top
|
|
 |
blue hell
Site Admin

Joined: Apr 03, 2004 Posts: 24446 Location: The Netherlands, Enschede
Audio files: 297
G2 patch files: 320
|
Posted: Tue Jan 16, 2007 11:02 am Post subject:
|
 |
|
I don't know if an anti wear-out algorithm is implemented by Clavia. If they did it wouldn't matter where you'd store patches, otherwise it might be good to do some rotating over the storage locations yourself.
BTW, when you just send a patch to the synth and do not store it I think it is kept in RAM (as it's gone after a power cycle) and not in flash. So I think it's only permanent storage (and OS updates) that will do harm to the flash in the long run.
Re. the cost of changing the chip I'd say it could be done in an hour or maybe two for an experienced person (so let's say max. 150 € might be a reasonable price). The chip itself is probably just a few €'s. However there is always a risk in the PCB getting damaged by de-soldering larger surface mount components. Might this happen it usually is possible to repair the PCB, but it won't look very nice then - with maybe some extra wires soldered around or excessive solder being used. I wouldn't mind too much myself about that, but others may have a different opinion.
For what it's worth ... I've done some development of flash based systems myself, for the job. I never ran into a situation that a flash device wore out during the debugging phase. _________________ Jan
also .. could someone please turn down the thermostat a bit.
 |
|
Back to top
|
|
 |
chresan

Joined: Jul 11, 2006 Posts: 150 Location: Karlsruhe, Germany
|
Posted: Tue Jan 16, 2007 11:26 am Post subject:
|
 |
|
The test suite I am going to write will test if changes to a patch are in sync with the patch stored in one of the slots. So the flash memory should not be affected at all since the patch is kept in ram. This sounds much better.
Thank you for the clarification. _________________ ----------------------
http://nmedit.sf.net |
|
Back to top
|
|
 |
micro

Joined: Mar 01, 2004 Posts: 28 Location: Croatia
|
Posted: Sat Apr 28, 2007 7:58 am Post subject:
|
 |
|
I was also asking this question myself, seeing this thread it freaked me out. Now I see that the best memory is my brain, real analog modular and pencil and paper for patch schematics
Quote: | The chip itself is probably just a few €'s. However there is always a risk in the PCB getting damaged by de-soldering larger surface mount components. Might this happen it usually is possible to repair the PCB, but it won't look very nice then - with maybe some extra wires soldered around or excessive solder being used. I wouldn't mind too much myself about that, but others may have a different opinion | Maybe is the best way to solder ZIF-socket on PCB for future proof if its possible and harmless. |
|
Back to top
|
|
 |
|