Favorite way to power Moteinos?

[Felix]

Would that work to "level translate"? I think so, that's what it's for, never used that particular one but should work.
Before you cut the trace, try it as is, they provide 5V power and 3.3V logic. See this: http://lowpowerlab.com/shop/FTDI-Adapter

[LazyGlen]

... But I would like it to be more compact so I'm going to use two CR2032 Lithium (not rechargeable) button cells, 20mm diameter. Each CR2032 is 250mAh @ 3V so I'll have 500mAh at 6V. Since a 9 V battery is also around 500mAh I should get 9+ months out of two small batteries.
...

It has been a long time since I studied this stuff, but I think you are mistaken.
Qty (2) 2032 batteries with 250mAH @ 3v does not get you 500mAH @ 6v. If you put them in series (one atop the other) you get 250mAH at 6v. If you put them in parallel, (side by side, if you will) you get 500mAH at 3v.

The other thing to be careful of is - what conditions are used to generate the rating? I pulled up a Sony  CR2032 data sheet via a web search: http://www.cr2032.co/cms/prodimages/sony_cr2032_datasheet_13.pdf and at the top of the page it states that the standard discharge current is 0.2mA or 200 nano amps. According to Felix's testing posted here: http://lowpowerlab.com/blog/2012/12/24/moteino-coin-cell-battery-tests/ depending on what you are doing, you can easily exceed that.

Don't let me stop you, but be prepared for results that differ from expected.
[edit]
I knew I had seen some measurements! The coin cell tests were based on RFM12B, the rating for RFM69W is 45mA in transmit mode, 16mA in receive mode. So you will certainly be exceeding the standard discharge current when the radio is active.
[/edit]

[SoLongSidekick]

I did try, I think I either have a software problem on my end or there might be something wrong with my FTDI Adapter. I tried using it on both my Moteinos and nothing. Then just to sanity check tried it on a Arduino Pro Mini and still nothing.

Will try again when I get home from work.

[Felix]

You got an FTDI Adapter from my shop and it's not working? That would be a very low possibility.
How exactly are you using it, what steps are you taking? (Software wise...)
What is the target board you are choosing?

[KanyonKris]

It has been a long time since I studied this stuff, but I think you are mistaken.
Qty (2) 2032 batteries with 250mAH @ 3v does not get you 500mAH @ 6v. If you put them in series (one atop the other) you get 250mAH at 6v. If you put them in parallel, (side by side, if you will) you get 500mAH at 3v.

The other thing to be careful of is - what conditions are used to generate the rating? I pulled up a Sony  CR2032 data sheet via a web search: http://www.cr2032.co/cms/prodimages/sony_cr2032_datasheet_13.pdf and at the top of the page it states that the standard discharge current is 0.2mA or 200 nano amps. According to Felix's testing posted here: http://lowpowerlab.com/blog/2012/12/24/moteino-coin-cell-battery-tests/ depending on what you are doing, you can easily exceed that.

Don't let me stop you, but be prepared for results that differ from expected.
[edit]
I knew I had seen some measurements! The coin cell tests were based on RFM12B, the rating for RFM69W is 45mA in transmit mode, 16mA in receive mode. So you will certainly be exceeding the standard discharge current when the radio is active.
[/edit]

The capacity does add when multiple batteries are used together (think 2 buckets of water). But you are correct, those coin cells can't supply much current and that's a problem. In a similar microcontroller + radio application, one guy used a capacitor in parallel with the coin cell(s) to supply current to the radio when it transmitted every 10 seconds.

I don't feel like fiddling with coin cells + capacitor so I'm going with 3 AAA batteries. The enclosure won't be quite so tiny but it will be small enough for my application.

Thanks, LazyGlen, for helping me avoid some frustration.

[LazyGlen]

The capacity does add when multiple batteries are used together (think 2 buckets of water). But you are correct, those coin cells can't supply much current and that's a problem. In a similar microcontroller + radio application, one guy used a capacitor in parallel with the coin cell(s) to supply current to the radio when it transmitted every 10 seconds.

I don't feel like fiddling with coin cells + capacitor so I'm going with 3 AAA batteries. The enclosure won't be quite so tiny but it will be small enough for my application.

Thanks, LazyGlen, for helping me avoid some frustration.

Not to belabor the point, but for clarity and future reference. You double EITHER the voltage OR the capacity (AH) with 2 batteries, depending on series or parallel. A more thorough writeup is here: http://www.zbattery.com/Connecting-Batteries-in-Series-or-Parallel

I ordered some extra NiMH packs for my cordless phones a while back for an abandoned project, they look like (3) AAA batts shrink wrapped with a JST(?) connector. They are not much bigger than the Moteino with the FTDI header soldered on. They look to be perfect for powering up these little suckers. Plus with 4 cordless phones around the house, to recharge, I can just swap a pack, phone to Moteino and drop the phone into it's charger.

I'm glad I saved you some frustration, I always seem to cause myself more!

[KanyonKris]

Not to belabor the point, but for clarity and future reference. You double EITHER the voltage OR the capacity (AH) with 2 batteries, depending on series or parallel. A more thorough writeup is here: http://www.zbattery.com/Connecting-Batteries-in-Series-or-Parallel

How about some more belaboring? Yes, the amp hours will not double, but 2 batteries do hold twice as much energy as 1. Amp hours are how much current a battery can supply in an hour. However batteries have internal resistance that determines how much current can flow out of a battery which is somewhat independent of how "big" the battery is, which you wisely pointed out with the small discharge rate of coin cell lithiums. Think of two buckets with holes in the bottom with one above the other (serial) - you won't get more water flowing with 2 buckets compared to 1 bucket, but you will have twice as much water so you'll get twice as much total water expelled.
Amp hours are useful for comparing batteries of the same type (ie alkaline or lithium ion), but they don't tell the whole story in terms of how much much electrical energy is stored. Here's a decent discussion - http://www.dcbattery.com/faq.html

I ordered some extra NiMH packs for my cordless phones a while back for an abandoned project, they look like (3) AAA batts shrink wrapped with a JST(?) connector. They are not much bigger than the Moteino with the FTDI header soldered on. They look to be perfect for powering up these little suckers. Plus with 4 cordless phones around the house, to recharge, I can just swap a pack, phone to Moteino and drop the phone into it's charger.

Be careful, many NiMH have high self-discharge rates and are often dead in a month, maybe 2 (like my cordless phones and previous generation cordless drill). There are low self-discharge NiMH like Eneloops that have a long shelf life and can be used essentially like alkalines.

[LazyGlen]

My wife tells me I argue too much and no one cares 

We may even be arguing the same side, and I realize that you have already moved past the coin cell for your project. I'm feeling dense, but I'm trying to learn, and I'll keep asking questions till you get tired of answering them!

Point 6 on your link page says the same thing I did. I realize that putting 2 cells in the circuit doubles the energy available, but you can only double 1 variable, capacity or voltage. Capacity == volume and voltage == head or pressure.

The water-as-electricity analogy works, and I've used it myself, but don't forget that any voltage over 3.3v gets burned off by the on board regulator, putting an "extra hole" in our buckets that only spills on the floor (or heats up the enclosure in the case of electricity).

Be careful, many NiMH have high self-discharge rates and are often dead in a month, maybe 2 (like my cordless phones and previous general cordless drill). There are low self-discharge NiMH like Eneloops that have a long shelf life and can be used essentially like alkalines.

Thank you! This explains why the battery packs were dead on the shelf where I left them. The one that I popped into the phone does seem to have taken a charge, I don't know if or how badly they are damaged.

I guess since this directly relates to the "How do you power Moteinos" title of the thread, I'll put this out for critique:
Building a slight variation of a WattMote. Would like to monitor power conditions long term, including date, time and duration of any fluctuations or interruptions. So I intend to add this small 3.6v 1000mAH battery pack inside the Kill-a-Watt with a LM317 regulator setup for a (1000Mah/40 =) 25ma (or so) trickle charge sourced from the full wave rectified  ~14.4v signal. The Moteino should then not have problems interfering with the nearly fully used supply available from the on board 7805 that others have reported having issues with and requiring the monster cap to mitigate with the xBee solution. This allows me to not really worry about current draw (within reason) for what I can stuff into the case, as well as enabling me to store in on board RAM (or an SD-card - stuffed in) a time stamp (DS1307 stuffed in) of power loss and recovery. When I power up or come in rage with a receiving unit, I can get a full log of data.

LazyGlen

[KanyonKris]

While we were agreeing on some things, I was wrong on the primary issue you brought to my attention: 2 same batteries in series will double the voltage output, but the Amp-hours will only be that of one battery. I was the dense one, but it finally clicked. For some reason the Point 6 of the FAQ I referenced was more clear to me than the one you first pointed out, and your mention of voltage as head pressure finally got the light bulb to come on. I see my mistake now. Indeed 2 batteries have twice the energy, but in series that is already accounted for by the voltage doubling. If the Amp-hours also doubled then the power would be 4x, and that's wrong. Thank you for persevering.

Yes, the regulator burns more power the higher the voltage goes above 3.3. Three alkalines shouldn't be too bad (4.5 V), but I would do better to use Eneloops that fairly quickly drop to 1.2 V and stay pretty flat until nearly out of charge - and that would give me 3.6 V which should result in very little lost to the regulator. Hmm, could I risk bypassing the regulator? I'll check the datasheets.

As for your WattMote variation, the general idea makes sense but I wasn't able to follow the details enough to be sure it will work as you intend. Wire it up and try it?

[LazyGlen]

... Indeed 2 batteries have twice the energy, but in series that is already accounted for by the voltage doubling. If the Amp-hours also doubled then the power would be 4x, and that's wrong. Thank you for persevering.

I'm glad we are now on the same page, though I was hoping you had a secret that I just didn't 'get' method of getting more energy out of those cells.

Yes, the regulator burns more power the higher the voltage goes above 3.3. Three alkalines shouldn't be too bad (4.5 V), but I would do better to use Eneloops that fairly quickly drop to 1.2 V and stay pretty flat until nearly out of charge - and that would give me 3.6 V which should result in very little lost to the regulator. Hmm, could I risk bypassing the regulator? I'll check the datasheets.

My point on the regulator was more regarding the 6v of our previous discussion. I think you are fine with 3 Eneloops, bypassing the regulator seems like it would be more trouble than it is worth. A quick web search indicates the freshly charged voltage is 1.4 volts, which gives you 4.2v into the Moteino, which is more than I would feel comfortable with.

As for your WattMote variation, the general idea makes sense but I wasn't able to follow the details enough to be sure it will work as you intend. Wire it up and try it?

I intend to. I just put it in here as a sanity check, in case anyone jumps up yelling "DON'T DO THAT!" or "but if you do this...".

I will write it up over on the projects page as I get further into it.

[hexibot43]

     I am truly enjoying this respectful debate.  I feel that I've learned a lot as well.  It has prompted me to  think hard about using the 9v that I have become so accustom to using.  Thanks,  Mab

   

[Felix]

9V batteries surely won't be as efficient as 3 AA/AAAs, but it's convenience for me...

[KanyonKris]

9-10 months on a 9 V battery is good, nothing wrong with that.

[Felix]

you can prolly get more if you transmit once a minute instead of every 8 seconds.

[KanyonKris]

I'm using this for my remote Moteino (using the Lowpower library):

Code Select Expand

LowPower.powerDown(SLEEP_8S, ADC_OFF, BOD_OFF);

SLEEP_8S is the longest duration predefined. My guess is 8 seconds is the longest the watchdog timer can be set for.

So to go longer than 8 seconds between transmitting data, I take it I'd just have a counter loop to put the Moteino right back to sleep when it wakes up until the count was up (say to 7 which is close to 1 minute).

I don't need updates every 8 seconds. I'll do 32 (4x8) or 64 seconds.

I think I will blink the LED every time the Moteino wakes up (every 8 seconds) just as a heartbeat so I can see it's alive. And double blink when it transmits data. I may remove this later to save a tiny bit more power, but will probably leave it in, could be handy.