2
\$\begingroup\$

Currently, I have a USB powerbank powering my arduino Uno which powers my micro servomotor. I would like to have the motor running for long periods of time (~24 hrs). Sometimes, I am able to achieve this run-time but in most cases the power bank shuts off after about 3-4 hrs of running. I am not quite sure about the variability. It seems that my powerbank has a current threshold (don't know exactly but probably 50-60 mA) and my arduino/servo motor are probably right at that cut-off.

What is the easiest way to handle this problem? Could I program my arduino to draw more current from the power bank (have it calculate pi to many decimal points?)? Or, would just buying a proper grade resistor do the job.

Thanks

\$\endgroup\$
5
  • \$\begingroup\$ No practical solution other than finding a "dumb" power bank that doesn't auto shut off. Earlier ones tend to be like that. \$\endgroup\$
    – dannyf
    Feb 14, 2017 at 22:54
  • \$\begingroup\$ @dannyf , actually your answer was probably the best, too bad you cancel it. \$\endgroup\$
    – Ale..chenski
    Feb 16, 2017 at 4:59
  • \$\begingroup\$ Ali, I didn't cancel it. I think some moron did it for me. \$\endgroup\$
    – dannyf
    Feb 16, 2017 at 12:05
  • \$\begingroup\$ @dannyf, probably the same one who downvotes my attempts to help... :-( \$\endgroup\$
    – Ale..chenski
    Feb 16, 2017 at 22:55
  • \$\begingroup\$ You got downvoted? I thought you made good, logical arguments in a reasoned way. It doesn't make sense door anyone to downvote you. This place is weird. \$\endgroup\$
    – dannyf
    Feb 16, 2017 at 23:57

2 Answers 2

Reset to default
3
\$\begingroup\$

A direct resistor connection will work, but will be very wasteful. Trying to calculate PI won't actually increase the load much. Instead, a simple intermittent pulsed load will work. A resistor and suitable N-channel transistor or mosfet, pulsed every n seconds would trick the power supply into thinking there is a large load, but also reduce the average load, hence longer battery life.

schematic

simulate this circuit – Schematic created using CircuitLab

A 1 second pulse of 75mA every 10 seconds is only a 7.5mA draw over 10 seconds, compared to a constant 75mA load.

Adjust the period and load as needed.

\$\endgroup\$
4
  • \$\begingroup\$ If you are so belligerent, your solution is also "patently" false. The idea is good, and it is true that powerbank circuitry [might] have some delay after the current drops below threshold. However, you have no idea how often the internal firmware scans the sensor ports, and how long you need to apply the load, or what should the period be. My powerbank drops the power after 7-8 seconds. So your algorithm already is not working. More, I tried 1s on, 4 s off with 100 mA, and it worked for a while (1 minute or two), then dropped off. Without knowing the algorithm, the pulsed method fails. \$\endgroup\$
    – Ale..chenski
    Feb 15, 2017 at 2:02
  • \$\begingroup\$ @alichen you think all power banks cut off on low loads. Not even half do. You think me calling that a lie is somehow belligerent. You also thought no power bank could take and provide power at the same time, only to be proven wrong by your own device. You may want to rethink your approach. \$\endgroup\$
    – Passerby
    Feb 15, 2017 at 18:15
  • \$\begingroup\$ Passerby, wrong again. Adjusting period without EXACT SYNCHRONIZATION with inner workings of powerbank controller will fail again, since your external frequency will be never the same as internal, and at some point your schema will fail catasrophically. If you know a power bank that does not shut down by itself, simply post brand/model/link, so people here will get it and not suffer from abrupt disconnect of power, myself included. So far this site has a half-dozen of questions about powerbanks, with no clear answer. \$\endgroup\$
    – Ale..chenski
    Feb 15, 2017 at 18:36
  • \$\begingroup\$ Passerby: regarding the powerbank working as UPS (charging and providing power at the same time), I have retracted my statement yesterday, after additional testing - the charge progress was disproportionally slow while external power was well available, so the bank can't be really qualified as UPS, and re-design seems necessary. \$\endgroup\$
    – Ale..chenski
    Feb 15, 2017 at 18:43
0
\$\begingroup\$

Most "powerbanks" are designed to cut internal operations off when the current falls below certain threshold (50 mA - 100 mA). Having additional load to keep it running will waste the battery runtime.

The best way is to modify the powerbank circuit to remove this protection. Some cheap powerbanks have a single control circuit that apparently has the current shunt circuit built-in, and it is not possible to disable. Some powerbanks will have low-Ohm sensing resistors on-board, so increasing their values will lower the cut-off threshold, but it may have adverse effect on the overall current delivery. It might be possible to somehow fool the cut-off circuit applying some internal bias, or else, it depends on powerbank control circuitry.

FOR ILLUSTRATION: My Powerbank (Model CJ-6000) has a pretty sophisticated circuitry, I might start playing with it: looks like a uP and three sensing resistors

See also https://electronics.stackexchange.com/a/270342/117785

ADDENDUM: after some reflection (based on powerbank circuit shown above) it becomes clear that a simple change in current sensing resistor will not work well - while the cut-off threshold gets lower, the overcurrent will also go lower, which might be unacceptable. The workaroud would be to use a non-linear element like a Schottky diode, with something like 10 Ohms in parallel (to define new cut-off threshold). This circuit has something like 3.5A overcurrent threshold (which amounts to 350 mV threshold over 100 mOhm shunt), so the diode must have a pretty low forward voltage like this one to avoid the overcurrent flip.

\$\endgroup\$
15
  • \$\begingroup\$ Thanks Ali. If there is no way for me to tweak the power bank circuitry, would you say that my best bet is to use an appropriate resistor. The powerbank is rated at 50,000 mAh (not sure if I believe this) but a 100 ohm resistor rated at .25 W should suffice. \$\endgroup\$
    – kingrumak
    Feb 14, 2017 at 19:23
  • \$\begingroup\$ @kingrumak, wow, 50Ah, this is a sizable jump-start quality pack! At 50mA load it should last about 1000 hrs, not counting for your arduino. If it is good enough, the extra load is probably your easiest solution. \$\endgroup\$
    – Ale..chenski
    Feb 14, 2017 at 19:44
  • 1
    \$\begingroup\$ All "powerbanks" are designed to cut internal operations patently false! \$\endgroup\$
    – Passerby
    Feb 15, 2017 at 1:33
  • \$\begingroup\$ @Passerby , my statement about "all powerbanks" is based on simple logistics. The powerbank industry is dominated by China, and all devices use same cheap Chinese ICs for control. Please provide evidence by pointing to any available powerbank that doesn't turn off on low loads, and I will happily replace "all" to "most". \$\endgroup\$
    – Ale..chenski
    Feb 15, 2017 at 18:57
  • \$\begingroup\$ Multiple generations. The cheapest banks, and thus the ones with the highest quantity out there, do not. Look at any youtube review of power banks. Every single one in this form factor. wildernessmastery.com/wp-content/uploads/2016/09/Power-Bank.jpg all the single cylinder ones. All the slimmest credit card shaped ones. Sold retail in any CVS, Walgreens Walmart, even dollar stores now. All the eBay multi cell ones that are sold with inflated stats and fake cells. Saying "most" have these auto off features is ignorant, as is thinking there is only one type of IC or board layout. \$\endgroup\$
    – Passerby
    Feb 15, 2017 at 19:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.