Ultra low power switching circuit & DC to DC converter t

[fastenjaw]

Hi guys! Nice place you got here I found this place while trying to find a site which might be able to help with a small crisis I have with low power components. It seems I can't find any circuit online which seems to help sufficiently and all I have is a basic idea of what will work. Well it's time to tell the story so you'll understand my problem.

I am sooooo sorry this is long.

After weeks working on a very special kind of small robot, I began to aproach completion until I made a few discoveries. The robot has enough room for a single (Not a typo>>>) AAAA Battery. The battery produces 1.5V and powers a small toy motor which drives wheels. The trouble is the robot staggers when fully powered up on a new battery. I can't make the robot any bigger without spending alot of money and its just way cheaper to boost the voltage .2 to .8 volts so the motor is slightly overpowered. Any obstruction will cause the motor to stall. There is no way I can add a higher voltage cell and the only ones available are too weak to power the motor or simply will not fit. I need a DC to DC converter which can run off as low as 1V and boost the voltage just enough without being too inefficient. I'm aiming for 90% efficiency. The battery has limited life so any power lossess should be avoided whenever possible. I have considered a IC known as SP6648 or SP6661. Both boost voltages a little as .5 volts but can't output less than 2.5 volts which is a problem since adding voltage regulators do waste energy as well. Also I can't find those inverters without having to make a company and buying something like 1000pieces or so... Not very practical when all I need is 1 or 2. I saw a converter which is 97% efficient but requires too much voltage and boosts it way too high. I refuse to strap a battery to the outside since it makes the robot look sloppy. This is the most important issue I have with my robot. Any help will be treasured dearly. Thanks

The second issue is alot easier (I think) but is still a tricky one since the circuit needs to consume no more than a few microamps so it could run off a 3V button cell for a year or more. The robot has a small switch which needs to trigger a circuit which drives the motor. Whenever the button is pressed the motor does not respond. It is only when the button is released is when the circuit responds and tells the motor to turn on or off. the first on/OFF cycle results in the circuit turning on the motor indefinitely. the next on/OFF cycle results in the circuit shutting off the motor indefinitely. The circuit needs to handle about 100mA of switching current. Also there can be no significant voltage lossess or inefficiencies since the battery again has a limited life. I have considered using logic gate circuits or even a modified timer circuit but I have the issue of draining the button cells too fast. Its possible to have the circuit shut down when no signal is detected after a duration but still needs to power a switching mosfet so that the motor remains on or off after the button is released. Another mountain I need to climb before I can hope to finish my robot. Thanks.

 

[Oznog]

You've probably got more trouble than that. Problem is, an AAAA cell can't put out much current. Measure the voltage the cell puts out under stall conditions. It may be that the power you can draw out at once just isn't possible, a converter will draw even more current than that off the cell.

Most boost converters capable of driving a motor will take up more space than adding a battery!

The glaring obvious problem is you're using the wrong battery. There are some short, small diameter rechargeable (I forget if they're NiMH or NiCd) used toy airplanes and those teeny, tiny rc cars. You could stack two or three in series where you put in an AAAA, also with multiple cells there's some leeway in where you can put them. Current capacity of those is actually quite high, but of course its runtime is low, usually only a few minutes. This is just a fact of life- tiny batteries are going to have tiny capacity.

A 3V lithium like a non-rechargeable CR123 has a lot of capacity, good current discharge capabilities, and pretty small. There's small rechargeable flat rectangular lithium prismatics at cheap surplus prices at https://www.allelectronics.com .

 

[tansis]

AAAA Battery
Alternate Designation LR6
Dimensions 42 x 08 mm

Without seeing the bot this is a stab in the dark
how about a stack of button cells wrapped in
heat-shrink tubing or a potted in silicone sealant?
Sadly one drawback leaps to mind, these cells are
not rechargable. There are some rechargable Lithium and Nimh
variants about that produce 3v but have no idea if they can be made
to fit. 30mm diameter 3.2mm thick, 3v / 100ma (200ma MAX drain)

update.. SR41 button cells are 7.9mm in diameter

 

[fastenjaw]

I tested the motor with a N battery on an earlier prototype robot. It lasted 8 hours on a new battery. I am hoping I could at least achieve 1/4 of that. I believe a AAAA battery has no less than half the capacity of an N battery. I have a theory of fitting in more than 1 cell but I dont know which cells work best and will fit. Keepin mind the motor is very small. Almost as small as a pager motor. So power consumption is actually like 50 mA or slightly less possibly 60mA tops. I have not checked the battery duration on the new one. I will do a lifespan check after I get the desired power delivered to the motor. Also I have room for the circuit just not enough room for more cells which would also weigh the robot down. Note I do have room in the pod section which is where the electronics will go. I can at best add a medium sized lithium cell to power the switching circuit. Just remember that the pod (Attached like a ball at the front of the robot) cant weigh toomuch more than it does without causing major balance problems and load on the front wheels.

I expect more power to be used up but remember the voltage boost is fractional. No major current change is posible. Maybe 15 or 25 mA increase. As long as the robot can last roughly 2 hours or more I'm happy. So does anyone know any circuits which can do slight voltage boosts at high efficiency??? .2 to 1 volt increase is not alot.

 

[fastenjaw]

WOW! I just discovered the AAAA battery has about the same or more capacity! Based on radio shack book saying a N alkaline battery has 440mA capacity. I found a Energizer Titanium Technology AAAA cell has about 625mA capacity!!! Thats even better!!! I think I could pull 10 hours off my older robot easily! The new one still needs the inverter though because of the stalling issue...

EDIT>>>

Sorry for double posting... Didnt want to write too much in 1 reply... The rechargable button cells in a definite consideration. The problem here is I'd have to buy tabbed versions so I dont solder the case. They WILL EXPLODE if I try soldering the surface as I have read in a safety section of a electronics website. The other configuration is to have them fit in individual holders and solder all negatives and positives. This is probably more complicated then just buying a chip and adding some components. Remember reducing voltage will almost certainly mean some energy is wasted with the regulator in place. The recharge or replace will be considered and they DO make rechargable AAAA batteries. My robot will not work with 1.2V. It already works marginally with 1.5V. Not without the converter can I make use of the rechargable AAAA anyways...

 

[Oznog]

Well first thing you need to know then is how much voltage and current you need to avoid stalling the motor, and you will need to know what the battery will put out for voltage at that current times the boost ratio and divided by the efficiency. You need to know these before selecting a boost converter.

It may be difficult to power this converter, most need at least 3V on the input side to drive the logic.

I still think the converter is going to take up more space than a second battery.

 

[audioguru]

Hi,
Do you believe those battery advertisments? Energizer don't make "Titanium" cells anymore since I complained that they don't have any Titanium. They sort-of agreed but said a few Titanium atoms drifted in from across the street. A very dry martini!
Check their "new" e2 cells (same battery, even higher price) at www.energizer.com and click Technical Info. Their E96 AAAA cell has a capacity of 625mA/hr to a rediculously low cutoff voltage of 0.8V. With a more reasonable cutoff voltage of 1.1V, its capacity is only 375mA/hr. At your 50mA current drain, it drops to 1.1V in 6 hrs, which is only 300mA/hr.

Did you forget that the battery cell's voltage drops with use? It ranges from 1.1V to 1.3V over most of its life. If you need a minimum of 1.7V, then a stepup converter must boost the 1.1V to 1.7V which is a ratio of 1.55. A new battery cell's 1.5V will be boosted to 2.3V and the current will be more than (add converter efficiency) 78mA. Your robot will be turbo-charged for a few minutes, the cell's life will be seriously reduced and even when running with an average cell voltage of 1.2V, the current drain will be more than 71mA.

I thought that Maxim-ic's MAX660 charge-pump IC is perfect since it operates with a supply down to 1.2V and has a 100mA output. But inspection of its datasheet shows that its output current drops to nearly nothing with a supply below 2V.

Try your motor with Energizer's L92 Lithium AAA cell. It provides 1.5V that slowly drops to only 1.4V over most of its life. Otherwise, you'll need a 3V Lithium cylindrical cell and a simple PWM circuit to drop its effective voltage to 1.7V.

 

[fastenjaw]

Hmmmm.... where to start... Ok. lets start with the titanium technology cells... I never knew the cells drop off voltage was steeper. Let alone the no titanium in there heh. Makes me think of sticking to either rechargable AAAA or AAAA standard alkaline. Still 75 mA will deplete the single AAAA cell in roughly 4 hours. 300 mA is decent terms of capacity but assuming I need at least 225 mA with at least 1.2 volts means I get a 2 hour run time before the robot starts to stagger and stop.

I figure about 2 to 3 hours of demo life is still solid and definitely doable for my robot. As for efficiency over voltage on low power converters, I DID manage to find a potentially promising converter over the weekend made by Texas Instruments. Its called Tps61010. I'm reading it again now. It can kick out 200 mA at .9V <<<Not a typo. The output can be varied from 1.5V to 3.3V Efficiency is at around 95%. Assuming I only need to keep the voltage above 1.5V to keep from stalling the motor, I could get the mA consumption to as low as 60mA with a .3 volt gain? 1/5th voltage gain should result in 1/5th current consumption gain not counting the 5% or 10% power loss. I'm estimating here. For the most part it still works at 1.5V but 1.6V to 1.8V should be all that I need and thus should suffice.

What do you think about the TPS61010? Anything better? ATM I'm sticking with AAAA and will try to succeed with it as I have shown success with an N battery. They are nearly the same other than shape...

 

[fastenjaw]

AAA batteries will NOT fit... I have a space in the robot with a 1" by 1/2" diameter space which is plenty of room for the 1 circuit board to be installed. Using SMT IC and a handful of tiny low power components is all I need. I have plenty of room for another 1/2 inch PC board too if I cram... But I cant add more weight (in batteries) in the front without loading the front drive wheels which already have alot of the weight... The battery mentioned a few posts ago is WAYYYYY TOO BIG. The robot looks like a small cylinder with wheels and cannot be made larger otherwise it wont fit in the tube it will be working in. The AAAA battery barely fits without me having to cut a hole through the robot... The rest of the space is taken up by the motor and the wiring...
/`\```\
/ B| M \ B is Battery compartment
| ----| | W is wiring space
|W_ | _ | M is gearbox with small motor
\_||_||__/ Wheels are at bottom
|| ||


______________________
| |___|
| |||||
| ___ |||||
| ___ / \ |----
|_/_o_\____/__o__\______|
\ /
--------------------------------------------------------
The motor stalls at about 1.5 or less. I tried a 3 volt cell just to see how it will perform but it was WAY OVERPOWERED. It ran like it was on steroids heh. There ARE converters designed for single cell use at optimum efficiency. Read my other post. Basically the motor is a 3mm diameter cell phone motor. Yea it has to be that small to fit... I think the current is actually less than 50 mA but I cant get a real acurate measurememt with my $40 DMM since it actually slows the motor somewhat and may be drawing added current itself.

 

[Roff]

I did a quick search and ran across **broken link removed**. Go to DC-DC converter IC manufacturers' web sites - Maxim, Linear Technology, National, etc., if this part won't solve your problem.

 

[audioguru]

Hi Fastenjaw,
The TPS61010 is a fantastic discovery that is perfect for you robot. With a tach or even just motor current sensing you can use its voltage regulator to increase its output voltage to provide more torque to push things (your opponent?)

 

[fastenjaw]

I will check the Maxim pages in a moment. I do have a theory which might work. I could make a larger motor fit. I have a kit in front of me with 2 tiny micro motors capable of higher RPM. The motors are about 1.2mm bigger and may fit with a little grinding of the gearbox. I just dont know if higher RPM may translate to more power... Any objections with the larger motor possibility? I still need the inverter though because I still may need to overpower it slightly even with a fractionally larger motor. I will definitely need the inverter for 1.2V rechargable AAAA battery usage.

Any one here know if Maxim stocks TPS61010? I called TI and asked how I can get ahold of these from a retailer without having to buy 1000 or so.

 

[Roff]

Maxim used to be very generous about sampling. I think this is still the case. The TI part looks more flexible, but Maxim doesn't make it.

 

[audioguru]

Fastenjaw,
Sorry, I previously mis-typed the TPS61010 part number.
Where do you buy electronic parts? A quick search at www.newarkinone.com shows a 41 day backorder and a single quantity price of $1.90. Another quick search at www.digikey.com shows 571 in stock for $2.48 each.
You might pay $8.00 for the courier.

 

[fastenjaw]

I'll check those places out. I originally tried google and put the words TPS61010 Order/buy. I didnt get any hits other than data specs and company websites which didnt sell it.

Oh and get this. Remember the theory about the motor being undersized for the job? Well I decided to rip out the smaller one I used in my other successful bot proto and replaced it with a larger 28000 RPM micro motor from those ZipZaps packages. The packs from RadioShack brings like 2 or 3 of em at various RPM. Anyways I did a current check and found that both motors actually take 80 mA under the load of the bot's weight. 25mA to 50mA unloaded. BUT the larger motor ran with more power and since it didnt stall as the smaller one did, it came out more efeicient!!! I grinded just enough to squeeze it in. It almost didnt fit! I was 1/2 mm from the surface! That was scary to cut since I didnt want to cut through. If I had screwed up you would be able to see a slit in the plastic housing. Not good If I were to show it as my final sample. Fortunately it looks like I may not need any other type of battery. I will soon get a charger, rechargable AAAA's, and buy the converter so that it can be displayed multiple times and recharged multiple times.

Thanks guys for all your help. You really helped me get my brain joggin. If the motor turns out to perform perfectly under its final average load, the only hurdle left is the switching device... I will test the motor with the enclosure after I get the wiring cleaned up a bit.

 

[fastenjaw]

Yup! Digikey has it! Thanks Audioguru!!!

Now for the switching...

Anyone know a ultra low consumption digital switch/circuit I could use? I need one to work on a 1.5V or 3V lithium button cell for at least year. The device needs to switch power on and off for the motor whenever a switch is pressed then released. The logic state needs to change when the switch is released. Any noise immunity and anti bounce added to the device is highly appreciated. Thanks