how easy is it to accidentally erase an eprom?

[fastline]

Long story short. I was working on an industrial control cabinet yesterday outdoors. Basically a 4x4x8ft tall control cab and then a small PC type enclosure inside that. The PC box door came open so I blew out the dust in it. The sun was not really out much and I know no direct light hit the chips but I didn't even realize until today the friggin UV erase windows are NOT covered? Who does that? In a high end industrial device? A little sticker maybe?

Anyway, they have some rather important data on them so hoping we are OK?

 

[Pommie]

If it's still working then you're fine. Put new covers over the windows (foil - not paper) and consider copying them if they are socketed and irreplaceable.

Mike.

 

[fastline]

It won't be powered again for several weeks. No way to know right now. There were no covers over the windows from the factory. About 10 chips on the PCB and not a trace there was ever a sticker on any.

 

[Pommie]

From my experience, using the sun to try to erase UVEPROMs doesn't work. However old ones that are approaching loosing data may just be pushed over the edge. I've seen many EEPROMs without stickers but they've normally fell off and blown away.

Mike.

 

[fastline]

Well, I didn't even grab any numbers off of them. I noticed the open windows tonight and immediately thought EPROM, because I know of no other chip that has a clear window on it's back. I didn't really want to shine a flashlight in there at that point. even loosing "partial" data will be enough to brick the device. I am just glad I was not in there shining lights, flashing pics, etc just yet. I read about erasing chips with strong sunlight and taking hours but not much mention about how much to start affecting them at all.

I am just totally baffled how these don't have stickers. This is a high end OEM and if they put a sticker on, it would still be there. Almost like they did it on purpose.

 

[KeepItSimpleStupid]

They really should not be "stickers", but sticky foil. The EProms are going to be sensitive to a certain wavelength of UV light I don;t think it actually has to be bright. I've done sunlight exposing of PCB's.

I would actually worry about degradation.

This https://books.google.com/books?id=7...q=eprom degradation sunlight exposure&f=false link says "extended exposure" can degrade EPROMS.

Your "probably" OK based on this: https://goughlui.com/2016/09/11/experiment-time-to-eprom-erasure-by-uv-lamp-sun/

 

[fastline]

In normal operation, these chips are double covered. Both inside the large enclosure, and inside another PC type enclosure. I know they got no direct sun and the door on the PC enclosure was open for about 10min. I blew out the dust, then shut the door and put tape on every hole so I could blow out the rest of the controls.

 

[Nigel Goodwin]

Long story short. I was working on an industrial control cabinet yesterday outdoors. Basically a 4x4x8ft tall control cab and then a small PC type enclosure inside that. The PC box door came open so I blew out the dust in it. The sun was not really out much and I know no direct light hit the chips but I didn't even realize until today the friggin UV erase windows are NOT covered? Who does that? In a high end industrial device? A little sticker maybe?

Anyway, they have some rather important data on them so hoping we are OK?

You're panicking over nothing, while it is possible to erase EPROM's using sunlight, it needs to be direct, high in UV, and takes a number of days of continuous exposure.

 

[gophert]

In normal operation, these chips are double covered. Both inside the large enclosure, and inside another PC type enclosure. I know they got no direct sun and the door on the PC enclosure was open for about 10min. I blew out the dust, then shut the door and put tape on every hole so I could blow out the rest of the controls.

EEPROM chips take hours in direct sunlight. If it was not direct light, UV doesn't reflect well, so, at most, you'll see 30% of the initial UV (one reflection off of a mirror-bright polished metal surface) and much less for painted, glass (even mirrors) or other surfaces.

If you are in the northern hemisphere (December/January) you have quite low UV in the first place. Overcast skies will further reduce.

In other words, why worry. It's not even your fault.

Also, tell the plant manager of the risk you've identified and he should abate the risk by obtaining duplicates of the ROM and covering the UV lenses.


Relative UV intensity by month and latitude... (from world health organization)

The UV Index Worldwide


Country (City) . J F M A M J J A S O N D
Argentine (Buenos Aires) 35°S 9 9 7 4 3 2 2 4 5 7 9 10
Australia (Darwin) 13°S 12 13 12 10 8 8 8 10 11 13 12 12
Australia (Melbourne) 37°S 8 8 6 4 2 2 2 3 5 6 8 9
Australia (Sydney) 34°S 9 9 7 5 3 2 3 4 6 7 9 10
Brazil (Rio de Janeiro) 23°S 12 11 9 7 5 5 5 7 9 10 12 12
Canada (Vancouver) 49°N 1 1 3 4 6 7 7 6 4 2 1 1
Cuba (Havana) 23°N 6 8 9 10 10 11 12 11 10 8 6 5
Falkland-Islands. 58°S 5 4 2 1 0 0 0 1 2 3 5 5
France (Paris) 49°N 1 1 3 4 6 7 7 6 4 2 1 0
Germany (Berlin) 52°N 1 1 2 4 5 7 7 5 3 1 1 0
Greece (Iraklion) 35°N 3 4 5 8 9 9 10 9 7 4 3 2
Japan (TokYo) 36°N 2 4 5 8 9 9 10 9 7 4 2 2
Kenya (Nairobi) 1°S 12 13 13 12 11 10 11 11 12 12 12 11
Madagascar (Tananarive) 19°S 12 12 11 9 7 6 6 8 11 11 12 12
Mozambique (Maputo) 26°S 11 11 9 7 5 4 4 6 8 10 11 11
Mongolia (Ulan Bator) 48°N 1 2 3 5 6 7 8 6 4 2 1 1
New Zealand (Wellington) 42°S 7 7 5 3 1 1 1 2 4 6 7 8
Panama (Panama) 9°N 9 11 12 12 11 11 12 12 12 11 9 9
Russia (St Petersburg) 60°N 0 0 1 3 4 5 5 4 2 1 0 0
Singapore (Singapore) 1°N 11 12 13 13 11 11 11 11 12 12 11 10
South Africa (Cape Town) 34°S 9 9 7 5 3 2 3 4 6 7 9 10
Spain (Palma de Mallorca) 39°N 2 3 4 6 8 9 9 8 6 4 2 1
Sri Lanka (Colombo) 13°N 8 10 12 12 11 11 12 12 12 10 8 8
Thailand (Bangkok) 14°N 8 10 12 12 11 12 12 12 11 10 8 8
USA (Los Angeles) 34°N 3 4 6 8 9 10 10 9 7 5 3 2
USA (New York) 41°N 2 3 4 6 7 8 9 8 6 3 2 1
Vietnam (Hanoi) 21°N 6 8 10 11 11 11 12 12 10 8 6 6

 

[MaxHeadRoom78]

As someone who has serviced CNC equipement, I have seen it happen all too frequently where cabinet doors are left open close to sunlit windows resulting in eproms becoming erased.
Max.

 

[sagor1]

Max, it is possible, but even mercury vapor lamps give off UV rays, so even inside an industrial setting with no windows but with those lamps, it is possible to erase EPROMs over a period of time if exposed to those lights.
Windows can pass UV-A through, but block UV-B (and UV-C which seldom passes thru atmosphere). Seems just about any UV will erase EPROMs over time. Sunlight normally does take "days" to erase a chip, so for this question, I would say the EPROM should be safe.

 

[MaxHeadRoom78]

This was cases where electrical cabinets were left open because of cooling issues etc, in these cases, the cooling issue should have been attended to in order to keep the cabinets closed.
I also used to stick a lable over the eprom window.
Max.

 

[fastline]

The control cabinet might have been left open at some point but the actual PC control cab (roughly a PC sized cab) was always closed. I had it open partially for about 15min when outside. I then opened it again indoors to give it a final blow out when I noticed the chip windows were not covered. I have foil tape I can put on them. Just hoping they did not see enough light to have any issues. Seems they need a pretty direct and intense shot of UV to do any harm? They are 18yrs old.

 

[Nigel Goodwin]

Just hoping they did not see enough light to have any issues. Seems they need a pretty direct and intense shot of UV to do any harm? They are 18yrs old.

As you've already been told, it takes days of direct exposure to erase in sunlight - you've no issues whatsoever. Although I'm happy you've now covered EPROM's, as I don't like to see them uncovered (even if the risk is pretty well zero).

However, you might like to consider:

"A programmed EPROM retains its data for a minimum of ten to twenty years, with many still retaining data after 35 or more years"

As they have important (and presumably irreplaceable?) data on them, it might be an idea to make copies.

There's an awful lot of pretty well antique machinery still out there working, where my son-in-law works (at a University) they collect old XP computers, in order to have replacements for the VERY expensive equipment controlled by them.

 

[unclejed613]

In normal operation, these chips are double covered. Both inside the large enclosure, and inside another PC type enclosure. I know they got no direct sun and the door on the PC enclosure was open for about 10min. I blew out the dust, then shut the door and put tape on every hole so I could blow out the rest of the controls.

if you look at an eraser box, it's got a UV-B lamp in it that is almost touching the tops of the chips. UV from sunlight is primarily UV-A after travelling through the atmosphere. i have seen stickers that have come off EPROMs without leaving any residue at all. it would not have been very wise to put EPROMs in something without something to identify the firmware version. the version number is often embedded in the ROM, but that does no good if the chip fails. i worked for a few years repairing banking equipment, and i had about 1 to 2 gigs worth of EPROM images. if it had firmware, i grabbed a copy of it.

 

[dknguyen]

i worked for a few years repairing banking equipment, and i had about 1 to 2 gigs worth of EPROM images. if it had firmware, i grabbed a copy of it.

Firmware collector

 

[KeepItSimpleStupid]

That;s a hobby I haven't heard of before.

 

[Nigel Goodwin]

That;s a hobby I haven't heard of before.

It's like a really sad version of train spotting

But to be more serious, I understand his point - it makes sense to make backup's of a memory that has a finite life, who might only need a backup once in your life, but having one could be hugely valuable.

On a similar issue, I used to have EEPROM backups for various televisions - EEPROM corruption was a common issue, and having a backup meant you could either reprogram the original one, or program a new blank one.

 

[gophert]

It's like a really sad version of train spotting

But to be more serious, I understand his point - it makes sense to make backup's of a memory that has a finite life, who might only need a backup once in your life, but having one could be hugely valuable.

On a similar issue, I used to have EEPROM backups for various televisions - EEPROM corruption was a common issue, and having a backup meant you could either reprogram the original one, or program a new blank one.

Is there any industry knowledge (actual data) or single-data-point experience (folklore) on the reliability/durability of a 35-year-old eeprom that is reprogrammed with same/different data?

 

[dknguyen]

Is there any industry knowledge (actual data) or single-data-point experience (folklore) on the reliability/durability of a 35-year-old eeprom that is reprogrammed with same/different data?

I don't think it's the age of the EEPROM that matters as much as how often it's been written or how long it's held the same data. I'd assume that an EEPROM that's 100 years old would work just as good as one that was brand new as long as the data it was holding was fresher and it's write endurance was not exceeded? No one ever talks about the age of the EEPROM itself...only it's data retention life and write endurance.