Main Site Documentation

Panda II RTC backup battery


#1

Has anyone used 3v rechargeable lithium coin cells to back the RTC? These life test units run into years and the clock is key. They sit on UPS backups but the hyro occasionally goes out for days here.

A coin cell would be easiest but I need a bunch of these units so if something more robust is needed it would be good to know before I build them out.


#2

Just look at the Fez Connect schematics : http://www.ghielectronics.com/downloads/FEZ/Shield/FEZ_Connect_sch.pdf : this board extends eventualy the panda2 with ethernet and other features, like RTC battery.

GHI is using a 0.22F supercapacitor for the RTC backup and it can last fo days. Maybe for weeks… It is way simpler than a lithium battery… :slight_smile:


#3

Thanks for the heads up on the connect schematic. I am fairly new to FEZ and any good info is appreciated. I was using Arduinos until I ran into a Panda spec sheet and the choice was obvious although I do give a lot of credit to the Arduino community for bringing these gadgets into the main.

If the .22F cap will keep one alive for days then a rechargeable lithium should be bulletproof. I have 17ma/hr 3.1v coins with solder tabs, and I am seeing 3.2v on the v-bat line which should keep them alive indefinitely.


#4

I have used these http://www.sparkfun.com/products/10319 for RTC power and found they are beefy enough to power the whole board for a short time! Definately overkill, both on size and power, but if your not space constrained, they do last a long time. The holder and breakout for these that SF sells make for a nice prototyping package. I also have a Connect shield, but have not tested how long the supercap will keep the clock going.


#5

Thanks. They are high capacity coins. One interesting note in the description is that you are not supposed to trickle charge them. On the units I have the running overcharge would be about .1 volt. Since they cost a whopping 1.68 I guess I can afford to experiment a bit and find out why not. A limiter might be necessary. Does anyone know what the minimum keep alive voltage is for the RTC?


#6

Just tested RTC and backup RAM with 1000uF 10V (thank you Intel for dead motherboard) connected to my Panda II.
It can keep clock and ram for like between 4-5 minutes.

I found that there are supercaps 1.0F-1.5F at 5.5V, like Panasonic EECF5R5U105 1.0F x 5.5V.
If I’m right, this big guy will last 1000x longer (Not sure if dynamics will be the same). So thats should be about 4000 to 5000 minutes ~ 3 days.

If it’s not enough, just use CR2032 @ 240mAh, which is de-facto standard for RTC and CMOS backup on big brother =) Those coins and sockets are widely available.


#7

The calculus is not that simple :wink:
You are probably using most of your 1000uF stored energy not to power the RTC chip, but “going away” by the internal resistor of the verrrrry bad quality of this kind of cheap filtering DC capacitor. And probably very used.

So a supercap wich is [italic]designed[/italic] for storing energy with a very high inner impedance will probably last much longer than you expect. :stuck_out_tongue:


#8

I have ordered rechargeable Lithium with solder tabs that should let me mount it right on the panda II below the board if I have read the dimensions correctly. They are inexpensive. The life test runs that I do run into years. All the filenames and timestamps are based on the rtc and it has to keep time.

In this new era of more violent weather hydro is much less reliable. As an easy example we lost it twice for about 6 hours this weekend. The panda will reboot when the power to the ups is restored, but if the RTC is wrong the data is useless to me.


#9

(A bit OT, but…)More violent than what/when? One logical fallacy we all suffer from is judging the entire world from the limited historical data of our own lives. Add in the fact when anything happens anywhere in the world we know about it almost instantly now and it is easy to reach incorrect conclusions.

I was watching a documentary the other night where they were tracing back the migration of humans out of Africa and the routes that different groups went. The climate has been a lot worse in the history of human kind, to the point of nearly wiping out the entire species.


#10

Better than saying “More violent weather” I’d rather say “lack of investment/maintenance in the hydro infrastructure during the past 30 years, resulting in the poor actual network quality” :smiley:


#11

Didn’t mean to start a weather controversy. I was just suffering to the local situation and my own lifetime observations. I have spent from two to six weeks every summer for the last twenty years on my sailboat in the North Channel (great lakes) and the last five years has been much more extreme than the previous 15.


#12

That is what I mean my friend, our lifetimes are to short a measure for something like climate. When I was a kid we had several really bad winters in the 70’s in Ohio, and then it was not too bad. They have had a few more bad winters in the past handful of years though. It is always changing.

I know my comments were OT, but it is one of those things that pushes my buttons…


#13

Controversy? Nah, the more opinions, the better! All the usual tech talk can get pretty dry… we went from batteries to the near extinction of humankind in one thread ;D I personally think we are way overdue for a (un)natural smackdown of some kind. The real bummer is dragging all the other critters down with us.

Conspiracy theories are encouraged! At least, scientifically plausible ones…