memory flash question

[John rodrigues]

is it possible to avoid head loss and radiation? for long-term storage without losing files

 

[barry]

I'm done. You keep asking the same question over and over and over and over and you don't pay attention to what people are telling you.

FLASH WILL LOSE DATA OVER TIME.

We can't be any more definitive that. Use another medium.

 

[betwixt]

The truth lies somewhere between. All charge based storage, flash, EPROM, EEPROM and so on, will eventually leak charge from the storage cell and lose data. Manufacturers usually quote figures of between 20 years and 100 years before no longer guaranteeing data integrity. There is some degradation of all these kinds of memory due to the writing process but that is more of a short term failure than retention once the data is stored.

Brian.

 

[John rodrigues]

Is this loss of data from flash and eeprom inevitable even with some frequency of use keeping it energized?

 

[ads-ee]

Is this loss of data from flash and eeprom inevitable even with some frequency of use keeping it energized?

Yes it is inevitable even with "refreshing the storage" because doing that will degrade the cells in the device until it can no longer be erased and programmed.

 

[John rodrigues]

If I write to the flash only once, how does this file degradation occur?

 

[betwixt]

Don't think of 'files', a memory device stores bits and nothing more. The way the bits are organized into files is dealt with by the operating system.
With the exception of fusible link devices (which you will probably never come across), memories store either a '1' or a '0' in individual cells. The difference between '1' and '0' is whether a charge is stored on a capacitor or not. The capacitor is actually fabricated into the silicon as a 'floating' conductor surrounded by insulation, a sensor picks up the charge if it is there and when requested to do so, reports it to the pins on the device. The reason they gradually lose data is simply that no insulation is perfect and some of the charge starts leaking away from the moment it is stored. The charge is incredibly tiny but the insulation is incredibly good so it can take tens of years before enough has leaked through the insulation for it to change state.

Brian.

 

[ads-ee]

Another example of charge storage that is easier to physically see and measure are capacitors. These devices also store charge.

If you charge a capacitor, disconnect it from the charging source and leave it sitting you will eventually find that the capacitor is no longer charged at some point in the future. The length of time it takes to self discharge is dependent on the insulation that Biran mentioned.

 

[John rodrigues]

My interest is to discover the frequency of use of electrolytic capacitors to prevent them from depolarizing? I have many devices with electrolytic capacitors stored

 

[ads-ee]

My interest is to discover the frequency of use of electrolytic capacitors to prevent them from depolarizing? I have many devices with electrolytic capacitors stored

Ah, it's finally clear what is going on with this member.
Someone is conducting a Turing test!

That seems to be the only explanation why we now have this non sequitur response from them based on my example of how charge based storage solutions can't be permanent.

 

[betwixt]

I've probably got over 5,000 electrolytic capacitor in storage here, some have been in stock for more than 20 years and I wouldn't hesitate to put them back into use again. I tend to test the older ones first but they never fail. It would take me a lifetime of charging and discharging them if doing it periodically was necessary.

Brian.