Battery operated Propeller

max72

Hi,
I would like to ask opinions or suggestions on battery operated propellers.

I'm not talking about ultra low power solutions, but of projects using rechargeable batteries and a propeller at full power.

I'll make a couple of examples:

1) Propeller with sensors, display, radio modules, requiring both 5V and 3.3V.
- The easiest solution is 6 NiMh AA batteries. You get a nominal 7.2V, and a couple of linear regulatos are enough. Is the solution I'm using now.
- Same as above but with switching regulators, like the LM2594 series. More expensive, requires external inductors, offers a longer battery life
- 4 NiMh AA or a LiIon battery. It settles to a voltage in the middle. I could not find a good solution, at least woth the effort.

2) Propeller with sensors, display, radio modules, requiring only 3.3V.
- this time 4 NiMh AA are good enough for a linear regulator, a switching one would be overkill
- Could be used a LiIon? I guess the handelds are using this solution, offering both recharge from USB, regulated output and so on. All the solutions I found are extremely complex, or missing an aspect or the other.

Thanks in advance
Massimo

Mike Green

It's not clear what you're asking.

A switching regulator is the best solution since it extracts the most energy from the battery, allowing the device to operate the longest. When you use linear regulators, the excess voltage is just wasted as heat. Switching regulators are more complex and more expensive, but the incremental cost has dropped and they're now pretty simple to implement using modern ICs.

The linear regulator solution is the cheapest and simplest, but the least efficient.

Rechargable lithium batteries require careful charge and discharge control. They require special regulators for both functions.

Miner_with_a_PIC

Mike makes good points, use of Li-ion batteries is very desirable(high energy density, low self discharge, good cycle life etc) but creating systems for charging and discharge control are a complete project in and of themselves.

The second option (using a common 3.3v rail) is very desirable as most lower voltage devices pull less current, you don't have to use multiple regulators and there is no need to condition I/O voltages between devices.

max72

Ok, I'll try to better explain...

I'm doing some experiments with displays and GPS. On the other hand I would like to make a small portable clock with temperature sensor, and maybe some other stuff.
I made a small test unit using a demoboard, a couple of serial LCDs and a GPS. At the moment I need both 3.3 and 5V. I settled to the 6 NiMh solution, and using the demoboard the linear regulators.

I'm trying to find solutions with a better display, maybe graphical, but in direct sunlight it is difficult to find anything as good as the serial 4X20 green displays.
I have a 3.3V GPS which is working rather well, but I don't know if I can find a 3.3V display as good. Otherwise I'll settle to a 5V solution.
At the moment this is a general questions, trying to understand if there are solutions available or best practice options.

If I'll need a 3.3 and 5V then the 6 NiMh with switching regulators looks as the best solution, offering longer battery life, but I wondered if other solutions were possible.

On the other hand if I'll settle to a 3.3V only solution the Li-ion is intriguing, even if dangerous. A 4NiMh + linear regulator or switching are still a good option, but the Li Ion-batteries look promising..
For charge control I found solutions like the Max1555 or the LTC3440. Both could be coupled with an ultra low dropout like the LTC3035.
Would it work? Are there better options?

The other problem is related to the breadboard. While both Max1555 and LTC3440 have an Msop package, which is tricky but still solderable, the LTC3035 is too much.

So I'm wondering if such a solution would work, and if other components with a frendlier package are available.

Massimo

icepuck

Here's a link to some switching regulators that are drop in replacements for the the 78XX.
the link is for the 3.3volt but they have 5volt and adjustable regulators as well.
hope this helps...
-dan
www.dimensionengineering.com/DE-SW033.htm

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Arguing with myself--sometimes me, myself, and I don't always agree.
(Former) Caterpillar product support technician

Post Edited (icepuck) : 1/13/2010 5:25:46 PM GMT

Ramboman

I use ONE a123 cell that gives 3.3V, witout any regulation.
Charged with a RC charger, a CC-CV labo power unit or the following DIP:
http://ww1.microchip.com/downloads/en/DeviceDoc/51850a.pdf

icepuck

Here's some info on LiFePO4 chemistry vs. other battery chemistries.

www.batteryspace.com/lifepo4cellspacks.aspx

Here's some links that have premade backs packs.

duraliteflightsystems.com/
www.batteryspace.com/li-ion18650battery74v2600mah1924whbatterypackmodulewithpcbprotectionic.aspx

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Arguing with myself--sometimes me, myself, and I don't always agree.
(Former) Caterpillar product support technician

Bobb Fwed

I have been using Li-Ion for a project I am working on, and I use both 5V and 3.3V. The batteries I buy (a 2x3, 6Ah at a nominal 7.4V) run about $40 and come with a complete discharge protection solution. I then use the MCP73862 to charge it (they have a single-cell version too).

I also use a 5V switching regulator (LM2674M-5.0) that then powers my LDO 3.3V regulator (a LM2674M-3.3 exists, but none of my suppliers ever carry any real stock, so I settled for the slightly lower efficiency LDO -- my 5V circuit is pulling more than my 3.3V anyway, and 30 days was the given requirement). I can get over 35 days of battery life out of my setup.

Another option would be to use the Ni-Cad or NiMh (at any voltage) and use a buck-boost regulator to extract even the last drops of power out of it (not really an option on the li-ion).

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
April, 2008: when I discovered the answers to all my micro-computational-botherations!

Some of my objects:
MCP3X0X ADC Driver - Programmable Schmitt inputs, frequency reading, and more!
Simple Propeller-based Database - Making life easier and more readable for all your EEPROM storage needs.
String Manipulation Library - Don't allow strings to be the bane of the Propeller, bend them to your will!
Fast Inter-Propeller Comm - Fast communication between two propellers (1.37MB/s @100MHz)!

Tubular

max72 said...
Max1555 and LTC3440 have an Msop package, which is tricky but still solderable, the LTC3035 is too much.

So I'm wondering if such a solution would work, and if other components with a frendlier package are available.

Massimo

The max1555 is in a simple SOT-23 package, easy enough to solder to breakouts from Sparkfun and many others.

The maxim datasheet confuses things by calling it a TSOT/5...

Toby Seckshund

Watch the quiesent current draw of the regulator as the dropout voltage is neared. They quote the dropout to be "xxx mV, @ xxxmA" but many spike up the quiesent to 100mA or more. So a 3.7V battery into a 3.3V out reg sounds wonderful, but more could be wasted in the reg than the load. Hence the attraction of 2 x 3.7V into a switching reg.

Depending on what the Prop project is the supply rail could be fed dirrectly from the Li Ion giving 4.2V max down to 3.7 min. To be kinder a diode could drop this by 0.7V, to 3.5V to 3.0V. This would give more protection for free at the expence of slightly lower regulation. becareful on charging the Li Ions they must be voltage capped and current limited to a proportion of its capacity. Battery packs have voltage balancing for the cell voltages and thermal sencing so that they can be caharged as heavily as possible ( and so that they do not last very long )

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Style and grace : Nil point

mctrivia

i use a 1.8 to 5.5V buck/boost regulator on my propmod-us_ps_sd to give a steady 3.3V out for the entire life of 2AA bateries.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
24 bit LCD Breakout Board now in. $24.99 has backlight driver and touch sensitive decoder.

max72

Thanks all for the feedback.

I'll investigate further the many options.

Massimo

Ramboman

Par info:

http://www.hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=6444&Product_Name=GENUINE_A123_Systems_26650A_LiFePo4_Cell_w/_tabs

Dr_Acula

I'm using switching regs LM2574 and LM2575. It depends on the current draw, but once you get to heatsinks with linear regulators, I think switchers might actually end up cheaper.

Also it depends on weight. LiPo is the lightest, but as stated, you must have proper over and under voltage protection etc.

NiMH is probably the simplest.

For me, my boards are in fixed locations (away from mains) with a solar panel supply and weight is not a problem. $ for mAH, I did some sums a while back and lead acid gel cells worked out the cheapest. And with switchers, efficiency is very similar whether you use 7V or 12V or 20V as the input. So I've got 13V 7AH gel cells, 10W solar panels with off the shelf charge controllers and switching regulators.

But every solution is different. This wouldn't be the answer for a mobile robot, for instance.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.smarthome.viviti.com/propeller

Ramboman

Try a123, without any protection...

Javalin

You can run the propeller happily from 3v and associated eeprom etc. I've got a miniGPS data logger thats completely 3v, except for a USB (5v)- but thats powered from the computer. It goes nicely on a 3.7v single cell lipo. The single cell lipo's are easy to charge from a 5v (aka USB) power source - several small IC's that do it all for you

James

max72

I had good results with NiMh and the protoboard (recent NiMh batteries are impressive), but a more efficient solution looks viable.

I haven't even considered the possibility to go for an unregulated solution. The VLDO linear regulators are in effect not a solution. I was looking for Buck-Boost solutions, but if all my gear is tolerant it would be a much simpler setup.

Just a question checking the data on the internet: the LiFe batteries show a flatter discharge curve, and are much less nervous, but they look only a little bit lighter than the NiMh equivalent, and heavier than a LiPo. Am I right?

Massimo

Miner_with_a_PIC

Massimo,

If you want to forgo the regulator all together it would seem that flatness of a discharge curve and voltage of the plateau region would be central to your battery choice decision. The LiFe batteries seem to be ideal in this regard as the plateau is right at or slightly below 3.3V for the majority of the battery's discharge cycle. Its is almost as if these batteries were customized for use on the propeller without regulation.

The power density is of less concern as you would need many more NiMH batteries and a regulator to achieve the same result as a non-regulated LiFePO4 set-up.

Additionally don't forget that NiMH have horrible self-discharge >>

From wiki : The self-discharge is 5–10% on the first day and stabilizes around 0.5–1% per day at room temperature.

I didn't do any serious in depth research but a cursory check seems to indicate that LiFePO4 cells have anywhere from 3-5% loss per month. Its common for all Li-ion battery types to have relatively low self-discharge rates.

Rayman

You can see a simple circuit with DIP boost chips that turns two rechargeable AA cells into +5 and +3.3 V here:

http://www.rayslogic.com/Propeller/Products/PAB/PAB.htm

Ray

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm

WBA Consulting

Haven't been able to try it out yet, but I just picked up a really nice 12v 2000mAh Li-Ion pack made by Energizer for $20 at WalMart. I dis-assembled it and the internal parts are put together nicely for hacking. Overall, I was very impressed with its quality of design. I think it will work out great for a handheld project I am working on. Looks like a great match for my PowerTwig too.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Andrew Williams
WBA Consulting
WBA-TH1M Sensirion SHT11 Module
Special Olympics Polar Bear Plunge, Mar 20, 2010
Propeller-Based Reverse Geo-Cache Birthday Present Project

Michael O'Brien

Just an aside,
·· but my Wife needed a portable temperature and humidity reader - the solution was to use any battery powered stamp board like the super carrier board and insert a spinstamp propeller 24 pin module in it instead.
The 9v battery runs the spinstamp, the sensirion sensor, 2 74hc595 chips and 8 multiplexed 7seg led modules fine from the 3.3v regulator on the spinstamp - all out of the box.

·· thank you
·· /michael

evanh

The newer "Ready-to-Use" NiNH's have very good specs. The self-discharge figures, for example, compete with alkalines so can be left charged for years. Lower internal resistance and higher capacity than previous, thousands of recharges, handles high and low temperatures, and don't self-degrade.

The one advantage LiIons still have is energy density but only when they are new, as LiIons are constantly degrading with time, no matter whether you use them or not. Also, this self-degradation is accelerated greatly with high temperature.