Post by thread:[PIC]:: PIC18F4580 EEPROM Problems

I have a product based on the PIC18F4580. It works well with RS-232, RS-485 and SDI-12. But I have one application at 4000 meters of altitude where the EEPROM settings are getting corrupted at unusual times. Once the EEPROM is wiped, I loose contact with the unit. There is a 5V LDO regulator. This application has a radio with 1/10W output. The case is just aluminum. The antenna is a 3db omni that is about 5cm from the PC board. The ambient temperature did not drop below -29C this past winter. The PIC18F4580 is rated for -40. I can't believe the radio would cause this problem. Out of 40 stations, 15 were made non-functional because the address, baud rate, communication mode are all stored in EEPROM. Power is turned on every hour for 40 seconds by a data logger with RTCC. I have hundreds of these units in remote areas using RS-232 with no problems. I use the same UART for all communication modes. I use the same command parsing and EEPROM Read/Write for all modes. I just enable! the appropriate IC for the mode. The 40 stations all use the SDI-12 communication. I ran it in the office for weeks without issue. I would like to hear about any theories on why this is happening. Thank you to the PICList Best Ragards, Zipwiz

At 03:20 PM 5/21/2012, you wrote: >I have a product based on the PIC18F4580. It works well with RS-232, >RS-485 and SDI-12. But I have one application at 4000 meters of >altitude where the EEPROM settings are getting corrupted at unusual >times. Once the EEPROM is wiped, I loose contact with the unit. >There is a 5V LDO regulator. This application has a radio with 1/10W >output. The case is just aluminum. The antenna is a 3db omni that is >about 5cm from the PC board. The ambient temperature did not drop >below -29C this past winter. The PIC18F4580 is rated for -40. I >can't believe the radio would cause this problem. Out of 40 >stations, 15 were made non-functional because the address, baud >rate, communication mode are all stored in EEPROM. Power is turned >on every hour for 40 seconds by a data logger with RTCC. I have >hundreds of these units in remote areas using RS-232 with no >problems. I use the same UART for all communication modes. I use the >same command parsing and EEPROM Read/Write for all modes. I just enable! Have you run it in an environmental chamber at low temperatures? > the appropriate IC for the mode. The 40 stations all use the > SDI-12 communication. I ran it in the office for weeks without issue. > >I would like to hear about any theories on why this is happening. My thought is that you are using capacitors that do not maintain their full value at low temperatures, and as a result your noise margin is reduced at low temperatures. >Thank you to the PICList > >Best Ragards, > >Zipwize

Battery internal resistance goes way up at low temperature. The exact temperature depends on the type of battery. I assume you are using some type of lithium primary battery if you are at -29C. I had to use lithium thionyl to get reliable performance below -40C. Bob On Mon, May 21, 2012, at 07:20 PM, .....zipwizeKILLspam@spam@comcast.net wrote: {Quote hidden}> I have a product based on the PIC18F4580. It works well with RS-232, > RS-485 and SDI-12. But I have one application at 4000 meters of altitude > where the EEPROM settings are getting corrupted at unusual times. Once > the EEPROM is wiped, I loose contact with the unit. There is a 5V LDO > regulator. This application has a radio with 1/10W output. The case is > just aluminum. The antenna is a 3db omni that is about 5cm from the PC > board. The ambient temperature did not drop below -29C this past winter. > The PIC18F4580 is rated for -40. I can't believe the radio would cause > this problem. Out of 40 stations, 15 were made non-functional because the > address, baud rate, communication mode are all stored in EEPROM. Power is > turned on every hour for 40 seconds by a data logger with RTCC. I have > hundreds of these units in remote areas using RS-232 with no problems. I > use the same UART for all communication modes. I use the same command > parsing and EEPROM Read/Write for all modes. I just enable! > the appropriate IC for the mode. The 40 stations all use the SDI-12 > communication. I ran it in the office for weeks without issue. > > I would like to hear about any theories on why this is happening. -- http://www.fastmail.fm - A fast, anti-spam email service.

{Quote hidden}> I have a product based on the PIC18F4580. It works well with RS-232, > RS-485 and SDI-12. But I have one application at 4000 meters of > altitude where the EEPROM settings are getting corrupted at unusual > times. Once the EEPROM is wiped, I loose contact with the unit. There > is a 5V LDO regulator. This application has a radio with 1/10W output. > The case is just aluminum. The antenna is a 3db omni that is about 5cm > from the PC board. The ambient temperature did not drop below -29C this > past winter. The PIC18F4580 is rated for -40. I can't believe the radio > would cause this problem. Out of 40 stations, 15 were made non- > functional because the address, baud rate, communication mode are all > stored in EEPROM. Power is turned on every hour for 40 seconds by a > data logger with RTCC. I have hundreds of these units in remote areas > using RS-232 with no problems. I use the same UART for all > communication modes. I use the same command parsing and EEPROM > Read/Write for all modes. I just enable! > the appropriate IC for the mode. The 40 stations all use the SDI-12 > communication. I ran it in the office for weeks without issue. > > I would like to hear about any theories on why this is happening. > > Thank you to the PICList > > Best Ragards, > > Zipwize This is typical of space originated radiation, as in particles from the sun, causing erasure of the part. My quick fix would be to glue a 1mm thick piece of tantalum on top of the chip, and on the other side of the PCB, under the chip. I have seen exactly this done when using COTS parts in instruments that have been flown, as well as on EEPROMS used for onboard firmware. You should be able to get suitable pieces of ready cut tantalum from any supplier of scientific metal sheet. -- Scanned by iCritical.

At 23.44 2012.05.21, you wrote: {Quote hidden}>> I have a product based on the PIC18F4580. It works well with RS-232, >> RS-485 and SDI-12. But I have one application at 4000 meters of >> altitude where the EEPROM settings are getting corrupted at unusual >> times. Once the EEPROM is wiped, I loose contact with the unit. There >> is a 5V LDO regulator. This application has a radio with 1/10W output. >> The case is just aluminum. The antenna is a 3db omni that is about 5cm >> from the PC board. The ambient temperature did not drop below -29C this >> past winter. The PIC18F4580 is rated for -40. I can't believe the radio >> would cause this problem. Out of 40 stations, 15 were made non- >> functional because the address, baud rate, communication mode are all >> stored in EEPROM. Power is turned on every hour for 40 seconds by a >> data logger with RTCC. I have hundreds of these units in remote areas >> using RS-232 with no problems. I use the same UART for all >> communication modes. I use the same command parsing and EEPROM >> Read/Write for all modes. I just enable! >> the appropriate IC for the mode. The 40 stations all use the SDI-12 >> communication. I ran it in the office for weeks without issue. >> >> I would like to hear about any theories on why this is happening. >> >> Thank you to the PICList >> >> Best Ragards, >> >> Zipwize > >This is typical of space originated radiation, as in particles from >the sun, causing erasure of the part. My quick fix would be to glue a >1mm thick piece of tantalum on top of the chip, and on the other side >of the PCB, under the chip. I have seen exactly this done when using >COTS parts in instruments that have been flown, as well as on EEPROMS >used for onboard firmware. > >You should be able to get suitable pieces of ready cut tantalum from >any supplier of scientific metal sheet. Why not lead?

On Tue, May 22, 2012 at 7:19 AM, <alan.b.pearceKILLspamstfc.ac.uk> wrote: > > >You should be able to get suitable pieces of ready cut tantalum from > > >any supplier of scientific metal sheet. > > > > Why not lead? > > While lead may be heavier, I believe tantalum is denser, in terms of > stopping particles that occur in upper atmosphere/low earth orbit > conditions. > > "Where space is at a premium and radiation protection is important, lead is often prescribed. It is recognized that lead is not the densest element (e.g., tantalum, tungsten, and thorium are higher on the density scale), but lead is readily available, easily fabricated and the lowest cost of these materials." from: www.marsmetal.com/sheet-lead/radiation-shielding -- Carey Fisher Chief Technical Officer New Communications Solutions, LLC 678-999-3956 .....careyfisherKILLspam.....ncsradio.co

I did some more research into my problem and I think that the Brown-out voltage protection is solution. I did not have brown-out protection on in the configuration switches. When power is removed, it takes several seconds for the voltage to drop. The PIC spends enough time in the brown out voltage range to cause the EEPROM wipe out. I was not aware that the EEPROM was susceptable to being erased or modified by low Vdd. I assumed the PIC would just shut off clean. I have not needed to use brown-out protection before. This low power application caught me by surprise. I am using Alkaline batteries. We change them twice a year. Lithium was too expensive and too difficult to ship by air. Solar panels were not allowed due to the ground cover spray that would coat the panels. As far as radiation shielding, I once worked as a test engineer at a major a satellite company that used Molybdenum slabs on every DIP package. It was very expensive but required for space exposure. On occasion, I do remote monitoring for a Molybdenum mining company. Thank you to the PICList for your valuable comments

On Wed, May 23, 2012, at 12:03 AM, zipwizespam_OUTcomcast.net wrote: > protection before. This low power application caught me by surprise. I am > using Alkaline batteries. We change them twice a year. Lithium was too > expensive and too difficult to ship by air. I have found that alkaline batteries basically stop working by -20C (you'll get about 10% capacity and current), you can verify that easily enough. Best regards, Bob -- http://www.fastmail.fm - A no graphics, no pop-ups email service

> I have not needed to use brown-out protection before. You HAVE needed to. You just have not realised the fact. Brownout protection is needed as of right in every application EXCEPT where you can utterly guarantee the need will never arise and you are competent and qualified to make that guarantee. In the latter case it is not essential , just almost certainly needed :-). I debated whether to add the smiley and decided to - BUT Murphy does not do smileys and will have no probl;em crashing you system with no BOD when he really really wants to. Brownout protection is not optional. If the processor has it, use it. If the processor has not got it, don't use the processor. The ONLY exception is where you do external brownout control. Note that asserting a reset line or NMI line is NOT enough to recover from brownout with certainty. The only thing that is guaranteed is a full power off for an adquate [tm] period. I have seen IC's locked in browned out mode with 0.2 V on their VCC line. > This low power application caught me by surprise. I am using Alkaline > batteries. We change them twice a year. Lithium was too expensive and too > difficult to ship by air. There should be no problem whatsoever in shipping primary Lithium cells by air if this is within the US and probably within most of Europe. In eg China you'll need specific certification at least. People like Fedex have standard "shipping batteries by air" documents that show what you can and can't do and how you need to prepare or certify batteries for air transport. Note that Alkaline batteries may also be subject to air transport restrictions. As Alkaline capacity drops very substantially at low temperatures. You could probably use a much lower capacity Lithium in their place. > Solar panels were not allowed due to the ground > cover spray that would coat the panels. Knowing the application would help make informed comment on that. > As far as radiation shielding, I once worked as a test engineer at a major > a satellite company that used Molybdenum slabs on every DIP package. It was > very expensive but required for space exposure. On occasion, I do remote > monitoring for a Molybdenum mining company. And, amongst other things, the man who commented on the radiation aspects makes satellites !!! :-) Russell McMaho

> And, amongst other things, the man who commented on the radiation aspects makes > satellites !!! :-) And I had this woman as a colleague for a while ... http://www2.physics.ox.ac.uk/contacts/people/aplin Under the Research tab, note the last bullet point ... Now look at this page ... http://en.wikipedia.org/wiki/Snowdon to get the height, and you will see that it is lower than where you are placing your equipment. Others may be correct in that it is electrical problems getting you with brownout, but you need to be aware of the cosmic ray problem as well. It may be doing similar things as the brownout does. -- Scanned by iCritical.