My Nano is going in a tight space and the host device is currently fed by 4 X AA batteries, alkaline or NiMH, whatever I have handy at the time. Trouble is either way I am > 5Vdc which exceeds the 5V regulated input but < 7Vdc, so the onboard regulator would be underfed. USB is not an option AFAICT.
I've considered a 2S LiPo battery, have several that "fit" but then charged they = 8.4Vdc, which exceeds the power input for the host unit.
I see a few options, help me walk through them, please.
2S LiPo w/ linear regulator (7805?) gets 5Vdc. Creates issue because LiPo has to be cutoff at 3.0Vdc/cell, so 6.0Vdc in this case. 7805 heat dissipation is 0.136W so no issues there ((8.4 - 5) * .04mA)
2S LiPo w/UBEC switching regulator, has option for 5Vdc Creates same cutoff issue as linear regulator. Maybe could build voltage divider and monitor LiPo on an analog input... here we go...
Voltage divider to feed regulated input with AA batteries(seems to fail in my test for some reason)
Can you explain how this helps? 4.5Vdc? When I power the AVR with 3 cell alkaline it never boots, just dimly flickers.
polymorph:
An MC34063 or similar switching regulator wired as a buck regulator can work right down to one diode drop above 5V.
Thanks for the info Polymorph. Would this http://bit.ly/lm2596step down buck package based on the LM2596 work? It seems so, just don't know about the heat. Since the input is close to output and the mA is low maybe ok?
As the amount of voltage you need to lose is so low, I'd just use a low dropout linear regulator to reduce the 6V to 5V (it's a shame that they didn't use one on the Nano, instead of a 78M05). This one http://uk.farnell.com/microchip/mcp1700-5002e-to/ic-v-reg-ldo-250ma-to-92-3/dp/1331481 has a dropout voltage of just 75mV typical @ 40mA load current. This is much less than the "one diode drop" you will get from a switching regulator, as well as being simpler solution. It's 83% eficient @ 6V input, and I doubt you will do much better with a switching regulator.
Put the 3AA into the 5V pin, not the Vin pin. Bypasses the regulator.
I have run 16 MHz '2560 designs this way for initial bootloading before powering from a 24V source (and appropriate switching regulator down to 5V). (Not a Mega, but a 2560 based design).
OK guys, I called and verified the absolute bottom end for powering the host device is 4.3V, so the 3xAA route seems sketchy and rules out NiMH. Maybe I should mention, the device is an RC airplane transmitter (Tx), so I'd hate to have a power issue because of running so closely to the power floor.
What about this? Instead of a 2S LiPo pack, use a 3S LiFe pack, nominally a 9.9V pack. Discharged it would still be above the minimum Vin of the on-board regulator, so if I ran that into Vin could I power the Tx from the 5V AVR pin? I still need to meter the mA draw the Tx in stock setup, but if it's low enough what do you think? And if it seems possible, would the 40mA per pin apply in this configuration?
You run it from 9.9V, outputting 5V, half of whatever current you draw is wasted as heat, dissipated by the chip. And its not very well heatsinked.
How much current do you need?
Maybe a separate switching regulator would be better - step-up for a single battery, step-down for multiple batteries. Examples are here:
Is everyone forgetting that battery voltage drops as they are used up? 6V will turn into 5V eventually, but with a 75mV low dropout regulator that will work fine. However, three AA alkalines at 4.75V max when brand new won't last long at that voltage, and in both cases you are wasting a large chunk of battery life.
I metered the mA draw from the stock setup. It spikes around 135mA for a split second, then settles into a 72 - 81mA constant draw. I question the accuracy of the initial spike since even if I stop the DMM from auto ranging it seems to drop before I trust the reading I get. So lets say double the reading, under 300mA then drops. Still within the 78M05 capability.
But...
I got to thinking, is the regulated +5Vdc pin that bypasses the regulator limited to just +5V?According to this thread +6V is the limit, and the OP there safely put +5.2V in the 5V pin. In that case, fresh off the charger, 4 NiMH = 5.6V. After a few days that drops to 5.2V, and drained to 1.1V/cell would be 4.4V. Imagine that... 0.1V above the floor of the Tx! If it's too sensitive for 5.6V what about a diode to drop some forward voltage?
I was able to get the Nano to run at 4.3V so IDK what was happening earlier, may have had it accidentally on the Vin pin.
Guys/gals, I really appreciate the guidance here. I'm new to Arduino so everything helps.