I gave up on "off the shelf" power banks a long time ago. Years ago you designed a TP4056 / FS312 LiIon charger circuit. Since then I incorporate charging circuits into most of my projects. The power banks I design and use have an ON/OFF switch. That's it! NO auto-OFF.
You can add an offset to the current sense with just 2 resistors: one resistor between the 10mR and VSN, and the other from VSN to GND. The first resistor allows you to offset the voltage at VSN while only drawing a very small current to GND. Example resistor values of 2.2R for first resistor, and 18k for second resistor will only waste
Sometimes there is an Easter egg to activate low power mode. Multiple presses or long press on. Other power banks have two outputs, one high current 2.1A and the other 0.5A. Choose the lower of the two. There are also LED area lights much like the resistor and that could work too. Add a switch for when it’s not needed. Or use two jacks at the same time.
I already do this on my power bank like 2 years ago and still in use until now. I am inspired by using the 555 timer one. But I do something more less power consumption, because pulsing constant load at the USB output still uses a lot of current in my case, because my power bank module need like above 200mA to stay awake. So instead of pulsing constant load at the output, I use the same 555 timer circuit to drive an optocoupler, thus the pin 3 and 4 of the optocoupler will connected parallel to the power button. Since the power bank also will stay awake every time we press the power button to reset the timeout, why not use 555 timer circuit to "press" the power button every 10 second or something to reset the power bank timeout so it will stay awake forever. This method use more less power because optocoupler LED doesn't need so much current to light up. I only connect 75 ohm resistor in series with the optocoupler LED pin.
@@denzo_0298 My pb is DIY one. I only bought the module with the LCD display, and made a custom housing by myself, so I already thinking for extra space for the extra circuit. But if you used a commercial power bank that already premade, you could always reduce like 1 cell to give space for extra circuit, in return, replace all cells with higher capacity per cell. For example the amount of cell inside is 5cell with 2000mAh each, 10,000mAh in total, replace it with 3000mAh cell, so you can reduce 1 cell. Now you will need only 4 cells with 3000mAh each, 12,000mAh in total, but now you have an extra space from one reduced cell. You can put your extra circuit there.
As others mentioned also: insert a resistor in series with the VSN pin, the current sensing input is probably quite sensitive so the series resistor will not have much influence on its own but it makes it much easier to inject a low voltage offset on the VSN pin.
Cutting the VSN point and adding a voltage divider might help(a series resistor from shunt VSN to IC VSN, then another resistor from IC VSN to GND). That will reduce the IC VSN point voltage slightly smaller than VSP, it will think current is above threshold.
the problem with this solution is that the powerbank will never be able to go to standby mode anymore even when no load is attached to it. From that perspective n555 solution seems more flexible and less intrusive
Eplop Electronics (UK) sell a Power Bank Keep-Alive kit based on a 555 timer. Mr Carlson's Lab also did a 555 timer circuit to keep a power bank from sleeping but he chose the RC values to connect the shunt resistor for 3s and disconnect for 25s. This worked well for all but one of my powerbanks. That one fell asleep after only 12s so I added a fast option dip switch to my version of the 555 circuit.
@@sp-vt4je When coming up with the figures i settled on a quite aggressive 50% duty cycle as some powerbanks required this and i am trying to make it work with as many power banks as possible as it's a fixed device. As bornach mentioned you can get the kit version and use whichever resistor values you wish to change the duty cycle and power drawn.
I realized something better by using SOC and shorting transistor by which I can program the on/off period freely. Because some powerbanks only need about 20s pulses to stay alive and the minimum current need may also be different. The huge problem is that every powerbank behaves slightly different. Solution with USB-C PD chip is also possible but the consumption may be quite high. I will definitely measure the still draw to the chip next time.
Thanks for this hacks, I struggled with this for a while, I even had to add a boost converter together with a tp4056 just to avoid the power bank modules all together.
You could just hack in a completely manual override: Make a separate 10,000:1 voltage divider between VSP and ground, and then add a switch that switches the VSN pin between its original resistor point or the fixed voltage divider. As long as the switch is flipped, it will always think the same amount of current is flowing, regardless of what actually is (this does have the drawback of losing the ability to sense overcurrent, though).
As a regular viewer of your channel, this video is very much useful for electronics hobbyist like me , I have tested and works fine ❤❤❤ Now I can run very low current loads without "No load shutdown" problem....Once again thanks a lot 🙏
I think that using a diode would be better. Just cut the trace between VSN and the shunt pad, and add a diode between VSP and VSN. Maybe will be needed to have resistor between VSN and GND to act as a load to the diode. But, even with and 0.3V diode, the IC would think it's flowing 30A and turn off as current protections, which leads me to think about a super diode, which uses an op amp to make the voltage drop of the diode be reduce by the factor of it's open loop gain
@@greatscottlab Diode could probably work, especially with voltage divider and "adding" original shunt, so that the diode v-drop divided down would provide the minimum voltage for current measuring to keep device up, while at full load the v-drop of the diode wouldn't really change that much - so the divided V measured would like be still smaller compared to the original shunt at 1-2A so the current limit shouldn't drop much - like 2x to 3x times the minimum required current for continuing to work - Schottky would go from like
This video reminded me of a car ride when I was little. It was hot and I decided to make a portable fan to cool down (car didn't have ac). Like a day prior to the trip I got an old usb cable and a 12V fan and connected them together. The fan was spinning pretty slowly but it was enough to get a cool sensation on your skin. The drive took a few hours and 3/4 of the way I was clicking the button on my power bank to wake it up like every 30seconds becouse the fan wasn't drawing enough current to keep it awake. Good times...
Literally from the datasheet for IP5328P: "If key is pressed, whether or not load is on VOUT1 port, VOUT1 port output will be force to open; but the VOUT2 and USB C port will open only when load is detected on the according port, or the output on these port will be closed. So VOUT1 must be retained if only one VOUT port is needed, if VOUT2 or VBUS is not needed, MOSFET can be eliminated, but the 10uF capacitor should be retained." And, if you need all ports to be working - thus lesser voltage on VSN pin. So then cut the trace and attach a voltage divider(100 to 1 to get 4.95v, 1000 to 1 to get 4.995v) from shunt to the VSN, simple as that.
Not sure if you are joking, but I think he uses this technique for emphasis as he completes and explanation. He appears to have had drafting experience as well, so pretty sure he has a decent stock of engineering pens in a multiple of thicknesses. Frankly, I really like this low-tech way of identifying and emphasizing an illustration or number.
Great idea! It's very frustrating that with so many little gadgets getting USB-C nowadays for their tiny batteries or low power draw, your best quality power banks are the most likely to refuse to recognise them.
I really hate USB-C. It should be a standard, not a family of standards. There are already dedicated connectors for all the things it does, but there is no confusion what port and what cable support what standards, what bricks can deliver what PD mode, etc..
I solved this in the past by having timed pulse on one of the spare pins on Arduino controlling a fet with 150 ohm/1w resistor. To power up the power bank I wired another resistor with its power button. Works great.
I love the idea with the 555 circuit and if you want to make it even more efficient you can use as load a capacitor that discharges via a diode at the 5v output or feeds the 555
Fascinating indeed! I guess the solution would be to go with USB C and embed that circuit into the project itself (so you don't need 2 USB cables). Thanks, dude! 😊 Stay safe there with your family! 🖖😊
I believe the PD protocol requires some kind of handshake between the device and the power source, and this means you actually need to integrate some suitable ic in your PCB
The older, single cell powerbanks don't automatically shut off, no matter the load. I upgraded the cells in a bunch of them, to have more available capacity.
It can be insanely frustrating trying to buy a power bank when most don't publish the full specs or features they have. It can often feel like a gamble. One feature I like is pass through which allows for the battery to charge while also discharging it great to use like a sort of UPS. Also nice for connecting small solar panels too things you need to power throughout the night until the sun comes back up. Maybe this would be hard but it would be cool to be able to buy one endorsed by you
I built my own using a TP4056 and a SDB628 (MT3608) boost converter module for less than 50 cents for both!!! Added two USB connectors, power switch, terminal blocks, and enclosed in a 3D printed case. Voila. Always on @ +Whatever volts. That boost converter is magic. works down to 2V which consuming microAmps of quiescent current. Brilliant times.
I think these approaches are still very valuable, personally I'd rather use my old power banks that have nothing else to do to power my projects rather than buying more new expensive ones and wasting the old, otherwise perfectly good ones.
If you're going to that much trouble, just get the lithium battery and the battery bank circuit and use it for charging only, then wire the lithium battery to a voltage circuit or just use as it for your project.
For low power projects it's easier to take 2s liio batteries with xt30 which is usually made for RC stuff and add simple usb converter and over discharge protection
IMHO just buying a cheap type C trigger board that lets me use any of my existing power banks is much easier than effectively building a custom power bank from scratch
A better option is another lower power piggybacked power bank of your own design. Just need to isolate and regulate output from charging input so as no over voltage or spike on the low power device at the very end of the chain! Just need to turn on the first in the chain occasionally to charge the second. Easy and less work. Simplest option of all: plug into a mains usb charger which won't cut out for low power devices! :-)
I use a lot of those power bank modules in my projects. I go around this limitation in my projects by using one of the MCU pins in the main project board to drive a transistor to flash an LED or just shunt current through a resistor intermittently to keep the circuit "alive".
@@fusseldieb Mine is actually part of the circuit and not an attachment to the USB port. Instead of a 555, the flashing is handled by the MCU of the main circuit
From the datasheet"Discharge: If key not pressed, only when sink device attached will the output port open, non-attached output port will not open. If the opened output port current is less than 180mA @ 10mOhm, it will wait for a period of time before close the port automatically. The output current is detected by the voltage drop between VSN to output, 180mA current is equivalent to 1.8mV on 10mOhm, when the resistance (including MOSFET internal resistance) between VSN and ouput is larger than 10mOhm, the current threshold will be decreased proportionally." So all you need to do is attach a 555 timer to the KEY pin and pulse it periodically. It should stay on without drawing extra current or messing with the current sense
Funny, because i recently made exactly the opposite experience. 😄 I bought absolutely cheap beard trimmer for traveling. But this one had a) old NiMh batteries and b) an extra adapter. So i used a step down converter to the used 3V and an old USB cable to use it with a standard USB adapter. But than the charging always stopped after less than a Minute! I thought a made a mistake but than i realized that this only happens when the battery is full and the flow is dropping. Which is perfect because the original Adapter dopes not stop charging, so you never knew when the battery was already full. But because of the auto-stop of the converter now it does. So now i have a USB cable to charge and additional an auto-stop. Win-win. 😄
Instead of trying to increase voltage of the IC VSP pin , it might be better choice to cut the VSN pin of the ic and put a resistor between shunt VSN TO IC VSN and a another resistor between IC VSN TO GROUND so that the ic vsn pin voltage get reduced than IC VSP voltage, may be it will think output current is above the threshold value
You can just use an usb-c powerbank with a type-c breakoiut board, set both CC 5.1K to GND with a resistor to get 500mAh output on that port and keep all other ports awake with higher output.
If the powerbank comes with wakeup button (or the control IC has a pin for that), you could use a 555 timer to trigger it in certain intervals (depending on IC's sleep timer... check datasheet). Should be even more efficient, than the 555 Timer with load solution. Or, you could check the Datasheet, whether the sleep timer itself can be set/changed (e.g. via resistor). Though that is unlikely possible in most. Oh, and to save some more power, you use a CMOS variant of the 555.
Simulating dead man's button... nice... but you need to open the powerbank and solder. If you for some device use multiple powerbanks (e.g. replace with another one newly charged) you must do this modification for each of them.
Simulating dead man's button... nice... but you need to open the powerbank and solder. If you for some device use multiple powerbanks (e.g. replace with another one newly charged) you must do this modification for each of them.
@@vencdee I mean, there are powerbanks, that can be easily opened. I personally prefer powerbanks that are filled with 18650 cells, that you can easily replace. On those you have easy access to the PCB. And if you do many projects, you might want to add one additional to the top of the list, a custom powerbank. There are plenty of good pre-built powerbank circuits available, along with 18650 cells. But of course, there are different solutions for different kinds of people with different skillsets and levels.
for these earlier power banks, you can put a switch on that resistor circuit, so you can get what most commenters are saying: a "trickle charge" or "low power" mode where you basically switch between the ultra-low power mode and normal operation, depending on your needs
Great designs! I think the 555 circuit is plenty good enough for people that need the convenience of a powerbank. The "pre-made" power banks are meant for expedient use anyway so as long as it's not discharging the battery for nothing it's fine. People that actually care about doing a "most efficient battery powered IoT whatever" will just embed the battery and add a basic BMC in their device (ie the little boards that just protect the li-ion cells from over/under voltage and overcurrent). Nobody should be getting a power bank and then try to hack it, just get a BMC board and cells and make a "dumb" battery
I highly recommend big powerbank empty cases for 18650. It's the best case for DIY electronics. - no need for hack like in video, just connect pin as input and u can read button pushes for custom special actions and every 10-30s set the same pin as output high for few ms to mimic single button push and that will prevent powerbank from turning off. - dont use all 18650, take one of two less and put mcu there.. - cut plus or minus line from usb port and put mosfet in the middle and u have 5v line that is being controlled by PWM - if u need more control then single button can provide u can remove LED and put pushbutton there (no need to drilling that is never perfectly centered so if have OCD u still be happy how project looks in the end) - case for 8 cells (and u can put up to 7) costs only 5$ (cheaper then plastic case, usb ports and li-on charger separately)
Another annoying problem is that most powerbanks turn off the output when you connect a charger. Otherwise you could use them as a small UPS. For a raspberry pi for Example...
I've figure out that this also depends on the voltage of the charger. If charger has a good cable and the voltage on the input port stays close to 5V my power bank doesn't turn of the load and indeed works like a small UPS. But if you plug it in a weak charger and the voltage on the input drops to say 4.5 it disconnects the load until fully charge itself.
I wonder if instead of pulsing a load resistor you could supply pulses to a capacitor that then powers your project, so it would pull the same average current but in narrow spikes that ought to trigger the load detection. On the plus side it should still shut down if your circuit is properly off, but on the minus side you'll have ripple which might or might not be a problem?
With the units I have they seem to play the game of averages... which is likely why he went after 100mA pulses when the verified steady current threshold was only 50mA. It's hard to tell *exactly* what is going on inside them but it would seem that the longer a load has been near the threshold, the more readily it will trip with just a slight further decrease in current. I think the pulse approach has a benefit efficiency wise because it may take a bigger hit on the averaging cap, keeping it further away from the cutoff point. That said, this is essentially what buck and boost converters do. It's an "everything you got for a moment" followed by "let that soak in", on repeat. So you could probably half get around this simply by using another converter that follows the one in the battery. It isn't going to really "gain" you much tho, you're adding on another "base" load that doesn't actually "do" anything, putting you back in simple minimum load resistor category.
A couple years ago I made a 3S pack using a cheap BMS board and a buck converter, put a toggle switch on it and wrapped it in fish paper (after I had an oops and had to replace the BMS board). Crude but effective. I did have to dedicate a power supply to it for charging with a current limiting circuit hacked onto it to prevent the supply from shutting down on overcurrent when charging.
Hi, as many projects require 3.3V and not 5V, you can use a 3.3 LDO regulator with overdischarge protection directly on the li-ion or li-po battery inside the powerbank. 3V is enough for low power projects (SMt32 + data acquisition + oled for exemple) , a low cost LDO is sufficient in this case
Im sure I’m not the first one to mention this but most usb powerbank have two outputs, so you could put the timer in one and your application in another. It would probably be fairly easy to miniaturize into a tiny dongle.
I've built such a circuit on a breadboard a while back. I chose the 555 timer route because it's simple and dosen't draw much power. My powerbank (20000mAh) turns off after 30 seconds with no or too little load. That current threshold is around 70mA with my powerbank. So my circuit draws around 150mA for around a second every 20 seconds. My (USB operated) NiMh battery charger can sometimes draw too little current for the powerbank to stay on.
It's nice to see Scott using Everyone's Favorite Timer Chip in a project again. Coulda done it with a cheap, ubiquitous microcontroller, but where's the sport in that? (Seriously, as much as I love the 555, there's something to be said for a chip that only requires a bypass capacitor and a load resistor to perform the same function, and the micro could easily be heat-shrinked into a tiny USB pass-through dongle.) By the way, Paul Carlson built a similar 555 timer "stay awake" circuit a few years ago. It still bugs me that power bank enginerds don't realize that many consumers need this functionality.
It feels wrong using an entire microcontroller to just pulse a circuit though :/ As for power bank engineers, they do realise, it didn't actually take very long after low power type C devices like true wireless headphones came on the market for power banks to start including low power modes. The problem is that the cheap power banks just recycle the old designs so you usually only get these modes on mid to higher end ones.
@@bosstowndynamics5488 I meant the ubiquitous $0.03 one-time programmable 6- or 8-pin chips. Just a few lines of code to grab a brief pulse of current every few seconds, 100 nanofart bypass capacitor, appropriately sized load resistor, USB connector traces directly on the PCB...Bob's your auntie.
What size are those shunt resistors (1206)?, you should be able to find or stack resistors rather than just unsoldering it, however the testing to figure out how to fix the shutdown problem is pretty awesome! Keep em coming!!!!
Those are great! I love the pulse circuit! Ingenious! had that shutdown problem once, and since then, I have bought power banks that support low current devices. With the Anker 737 I have right now you can hold the button down for a few seconds to lock on. And draw as little as you want.
Although he wasn't the first to make use of a 555 for this purpose. I recall Mr Carlson's Lab made a version in a video titled "Power Bank Stay Awake!" over 3 years ago.
If you cut the trace to the LOWER of the two current sense inputs and add a 1:10000 voltage divider, the measured current will be about .5mV higher than what it really is. No opamp required. The only drawback is that the "amplification circuit" inside the chip may expect an about 10mOhm output resistance of the current sense resistor so it might be designed to draw a significant current. But I'd start out with a pretty high-impedance voltage divider. At such a ratio the big resistor actually becomes "almost a current source" So you almost get an actual .5mV offset as you wanted. Try 1M/100 Ohm, if that doesn't work, try 100k/10 and 10k/1 ohm respectively.
Some powerbanks have what's called a "Trickle charge" mode.. This mode forces the powerbanks output to always stay on. I've seen it being a thing on some of Anker's powerbanks, where you longpress the powerbutton and the indicator LED turns from blue to green...
IP5328 have I2C communication, u can easily configure this behavior throuhg it. First power on by button , to power the mcu, mcu sending "always on" to IP5328. This shunt 10mOhm it's for charging/boost current sense. I bet the whole IC get crazy after you changed it :)
The datasheet lists a parameter about light load detection which uses the voltage between the positive USB voltage rail (connected to pin VOUT1/2, VBUS) and VSN. Seems like they use the output FET as a shunt. If less than 1.8mV is measured, light load is detected. Maybe a low forward voltage diode (e.g. Schottky diode) between the positive USB rail and VOUT could do the trick without affecting charge/discharge current protection. Although, the datasheet is a bit conflicting at that point 🤔
Maybe someone else already suggested this solution, but probably the easiest one is to modify a power bank by removing the stock PCB, and installing two premade devices: 1) Li-Po charger with USB-C input 2) Dc-dc converter to have constant 5v (or another voltage as required) 3) (optional) for those power banks having more outputs, add a secondary dc-dc converter configured for an higher output voltage, like 12V In order to avoid discharging, all dc-dc converters must be connected to a switch, in order to cut the power when not in use. 😊
I just replaced the momentary button with a self locking one it kept the power bank on forever the bonus here that stand by mode could be switched on or off 😁😁
I use the pulse method in a small circuit mounted on top of the 555 itself that I put inside the power banks and it pulses the output every 30 seconds for 1 second, and it works fine...!!! Great video.
If your project contains microcontroller and you have spare output pin and some space on board you can program pulse draw, or use second small dedicated microcontroller like Attiny5 and some transistor with load resistor to do it.
Many power banks do come with a low power draw mode. What I would like to see is to **how to use them as UPSes** since most of them turn off for a second when AC power turns off or on.
I think rather the cheap ones turn off when AC-power changes to have a cheap but still safe renegotiation of protocols and capabilities. But you can probably be super cheap and simply add a large enough capacitor to survive the downtime. While they do work greatfor UPS applications, such on-line UPS systems wear out the battery quickly. better would be to use it as standby once power drops (like most USPs do)
Modifying a power bank to act as a UPS would be much more complex than these mods, so it probably makes more sense to just buy a power bank with passthrough charging/UPS mode to begin with (fairly common these days).
I bought a power bank that has 4x 18650 protected button top cells wired in parallel and they are user replaceable with a slide off battery cover. I slide the battery cover off and insert the jumper leads to + and - bypassing the the electronic circuit board if I want to use with low power devices like micro controllers. It provides plenty of runtime for small loads, and if I want to be more exact with voltage or use maximum amount of power from the 18650, then I can plug a small DC-DC boost converter.
My powerbank can actually be set to a "disable auto shutdown"-mode which also switches off any fast charging. It just provides stable 5V. It's a ROMOSS Sense 8P+. The mode can be activated by a double tap on the on/off button while it's on.
Alternatively and what I think is nicer: a small 3.3 V battery that is connected to the power bank via a relay/Mosfet to wake it up. (when most power banks are last switched on) The small battery is permanently connected to the Arduino. Advantage: The small battery can be recharged directly from the battery pack depending on the duration and wakes it up at the same time. :)
I experienced it alot, An led can also be used as a load for output spikes preferably an SMD one. Or just force button press of the power bank to keep it turned on all the time using of course a signal instead of manually physically doing it.
I like the second solution: Periodically drawing enough power to keep the power bank activated. You could incorporate this in your project, by periodically enabling an output of your Arduino, for example, to power a LED.
Yay for the 555. Useful in 1978* Still useful in 2024! * Marketed in 1972. But I was using it as a kid a tad later 😅 Think the cmos version came out in my youth too...
Also a problem with some small synthesizers. I found that some old small cheap powerbanks don't have a switch-off (ir it is very low). One soultion: using a USB hub and plugging in a USB lamp additionally. Then it stays on.
Great video! I like the pulse circuit. Could dial it in to trigger right before the shut down of the power bank to minimize how much power is drawn instead of the 8 second interval used for demonstration. Maybe as a subsystem added to the small projects themselves to allow function with any power bank without external modules. I know the current draw is quite small, but were there any voltage dips presented on the output side from the pulse?
Could you not bypass the shunt completely - just swap it with a 0ohm, so 5V goes to both VSP and the USB port without interruption, then put a 1ohm resistor from that 5V line to VSN, so regardless of the current actually drawn, VSN sees the drop from 5V to 4.98V (assuming VSN sinks 30mA) and this is over the 18mA threshold to keep it on?
You could also add a switch to bridge your higher value sense resistor with the smaller resistor for the times when you need to supply higher currents.
I found another hacky solution for my powerbank: I just press it's button once every few seconds (with a NE555 timer), which lets it stay on. That does not even draw any wasted current (except a few mA for the NE555) and works perfectly (for me)
One thing I would like to see is a low voltage input cutoff on powerbanks. I have several that will draw all the way down and not shut off when they get low. I don't like the batteries going that low. Luckily I'm using recycled laptop batteries that were free. They would last longer if they didn't draw down so low.
low power mode can be activated on majority of branded powerbanks. the real problem is that some powerbanks may turn off after charging your device instead of acting like a wall charger (until power bank drains).
Hi Scott, here is my solution: I had an esp32 lying around, wrote a short micropython-script, it connects to wifi and starts WebRepl (both not necessary), then goes into deepsleep for 28 seconds. I also "removed" the LED (with an electrician screwdriver). I'm sure it will draw more energy, but it was simple for me to implement. Thx for your videos from bremen!
I work mainly with phones. For example, just by the current that the iPhone draws, you can tell what is damaged in it. I have a 20,000mAh and 20W Powerbank from BaseUS and it drives me nuts when it turns off after 15 seconds of no or low power consumption (around 0.05A). I was thinking about making a PCB for it to artificially consume power, but I put it off until later. Btw my old Xiaomi 10 000mAh Powerbank with 5V2A output supports low power mode by simply clicking power button 2-3 times. Then it will never turn off, but I can't always use it, especially when I need Fast Charging. Thanks GreatScott!
there are two common problems in see with power banks. 1. threshold voltage over time. your video addresses this. most i have tested cut off after around 30 seconds with small current used, or the consumption is too low to even keep it running. yours sounds to solve this. 2. passthrough charging use case: sbc like a raspberry pi using a power bank like a ups. it should only allow mains power until there is an outage then kick in. most power banks do not have passthrough chargung. it would be helpful to have a power switcher that would activate any powerbank when an outage occurs, probably with a capacitor in order to keep continuous power right when the power bank is about to activate. that way you can use any powerbank in whatever useful form factor to go with your project. by default, powerbanks simply drain out. you don’t have the option to use mains then power bank as secondary. some solutions are clunky with a custom circuit and no power bank while others just use a power bank tethered to mains but it is still using the battery first.
This is a great succinct example of why everyone hates software and technology: A designer or company makes one assumption, and there will always be a huge number of people that don't want that as a feature. And thus nothing ever does exactly what people want
The shunt resistor mod gives me an idea... what about a voltage divider on the 555's reset pin, carrying (or transistor buffering) current? Pull RESET low when current exceeds a certain threshold, so we're not unnecessarily pulsing the load when the attached device exceeds the minimum draw. Might be useful for something like a remotely deployed ESP32, where it''s mostly idle, but occasionally has a bunch of Wi-Fi connected work to do, for example.
You can probably get away with not having a transistor, because the NE555 can easily sink 100ma. These old school ICs are tough cookies 😂. I normally drive relay coils directly with the NE555. Maybe we could do a very minimalist version of the circuit. I love the idea, and I also have been frustrated with 5v power banks turning off 🤔😩
Disconnect the VSN from the circuit. Then make a voltage divider circuit that will feed in a delta 50mV to VSN. Let say you constantly draw 10uA on your divider circuit. 50ohm from VSP to VSN, and 500k from VSN to GND.
A simpler way to do the pulse load would be to use an ATMega328. With the 5v supply it would not need any external regulator, just a bypass capacitor at VCC. It draws around 8 µA with the watchdog enabled in sleep. Just wake it up every x seconds and parallel four or five output ports with a 47 ohm 1/4 w resistor attached to Vcc. The outputs can handle 20 mA each (easily) and you can use the internal clock so you just need one chip, one resistor, one bypass cap and you’re done.
If you already have a microcontroller within your project, you can have keep-alive at the cost of one pin, a random tiny little MOSFET, and a couple resistors, one to tie down the gate against floating and one for the actual load. Or a random BJT then one resistor is to limit base current, it doesn't necessarily need anti-floating.
The engineer who thought that shutting off the power bank if it does not detect 500ma is a id10t. Its in the same field as a cloud service for appliances.
Well these things were mostly designed for phone charging, and even old phones usually drew at least 500ma. Newer banks take into account that you might be using them for other things.
The IP 5328 chip is programmable, granted it only mentions charge curves, but probably also has provisions for setting the auto shut off value. Using TP4056 and a 18650 cell we bypass the cut-off problem altogether. And while the TP4056 draws 20µA that will take awhile to discharge with no load attached.
I gave up on "off the shelf" power banks a long time ago. Years ago you designed a TP4056 / FS312 LiIon charger circuit. Since then I incorporate charging circuits into most of my projects.
The power banks I design and use have an ON/OFF switch. That's it! NO auto-OFF.
You can add an offset to the current sense with just 2 resistors: one resistor between the 10mR and VSN, and the other from VSN to GND. The first resistor allows you to offset the voltage at VSN while only drawing a very small current to GND. Example resistor values of 2.2R for first resistor, and 18k for second resistor will only waste
Sometimes there is an Easter egg to activate low power mode. Multiple presses or long press on. Other power banks have two outputs, one high current 2.1A and the other 0.5A. Choose the lower of the two. There are also LED area lights much like the resistor and that could work too. Add a switch for when it’s not needed. Or use two jacks at the same time.
I already do this on my power bank like 2 years ago and still in use until now. I am inspired by using the 555 timer one. But I do something more less power consumption, because pulsing constant load at the USB output still uses a lot of current in my case, because my power bank module need like above 200mA to stay awake. So instead of pulsing constant load at the output, I use the same 555 timer circuit to drive an optocoupler, thus the pin 3 and 4 of the optocoupler will connected parallel to the power button. Since the power bank also will stay awake every time we press the power button to reset the timeout, why not use 555 timer circuit to "press" the power button every 10 second or something to reset the power bank timeout so it will stay awake forever. This method use more less power because optocoupler LED doesn't need so much current to light up. I only connect 75 ohm resistor in series with the optocoupler LED pin.
How do you fit an extra circuit in your power bank??
@@denzo_0298 My pb is DIY one. I only bought the module with the LCD display, and made a custom housing by myself, so I already thinking for extra space for the extra circuit.
But if you used a commercial power bank that already premade, you could always reduce like 1 cell to give space for extra circuit, in return, replace all cells with higher capacity per cell. For example the amount of cell inside is 5cell with 2000mAh each, 10,000mAh in total, replace it with 3000mAh cell, so you can reduce 1 cell. Now you will need only 4 cells with 3000mAh each, 12,000mAh in total, but now you have an extra space from one reduced cell. You can put your extra circuit there.
Pure genius... dead man's button for powerbank... excellent idea!
As others mentioned also: insert a resistor in series with the VSN pin, the current sensing input is probably quite sensitive so the series resistor will not have much influence on its own but it makes it much easier to inject a low voltage offset on the VSN pin.
I did a pulse circuit with a Attiny13, it had far less components, and cost around the same if you are not mass-producing it.
Cutting the VSN point and adding a voltage divider might help(a series resistor from shunt VSN to IC VSN, then another resistor from IC VSN to GND). That will reduce the IC VSN point voltage slightly smaller than VSP, it will think current is above threshold.
Yup, was about to suggest that
the problem with this solution is that the powerbank will never be able to go to standby mode anymore even when no load is attached to it. From that perspective n555 solution seems more flexible and less intrusive
@@borincod yes you’re right it would probably drain batteries much faster when the Power bank is stored with no load.
Eplop Electronics (UK) sell a Power Bank Keep-Alive kit based on a 555 timer. Mr Carlson's Lab also did a 555 timer circuit to keep a power bank from sleeping but he chose the RC values to connect the shunt resistor for 3s and disconnect for 25s. This worked well for all but one of my powerbanks. That one fell asleep after only 12s so I added a fast option dip switch to my version of the 555 circuit.
Many thanks for the mention.
It has a 50% duty cycle tho, which is too much
@@sp-vt4je When coming up with the figures i settled on a quite aggressive 50% duty cycle as some powerbanks required this and i am trying to make it work with as many power banks as possible as it's a fixed device. As bornach mentioned you can get the kit version and use whichever resistor values you wish to change the duty cycle and power drawn.
I realized something better by using SOC and shorting transistor by which I can program the on/off period freely. Because some powerbanks only need about 20s pulses to stay alive and the minimum current need may also be different. The huge problem is that every powerbank behaves slightly different.
Solution with USB-C PD chip is also possible but the consumption may be quite high. I will definitely measure the still draw to the chip next time.
Thanks for this hacks, I struggled with this for a while, I even had to add a boost converter together with a tp4056 just to avoid the power bank modules all together.
Glad I could help :-)
You could just hack in a completely manual override: Make a separate 10,000:1 voltage divider between VSP and ground, and then add a switch that switches the VSN pin between its original resistor point or the fixed voltage divider. As long as the switch is flipped, it will always think the same amount of current is flowing, regardless of what actually is (this does have the drawback of losing the ability to sense overcurrent, though).
As a regular viewer of your channel, this video is very much useful for electronics hobbyist like me , I have tested and works fine ❤❤❤ Now I can run very low current loads without "No load shutdown" problem....Once again thanks a lot 🙏
Not that long ago I had to browse the internet in the search of something like this for a project. Really nice to see it well documented in a video.
I think that using a diode would be better. Just cut the trace between VSN and the shunt pad, and add a diode between VSP and VSN. Maybe will be needed to have resistor between VSN and GND to act as a load to the diode.
But, even with and 0.3V diode, the IC would think it's flowing 30A and turn off as current protections, which leads me to think about a super diode, which uses an op amp to make the voltage drop of the diode be reduce by the factor of it's open loop gain
Hmmmm also sounds interesting.
@@greatscottlab Diode could probably work, especially with voltage divider and "adding" original shunt, so that the diode v-drop divided down would provide the minimum voltage for current measuring to keep device up, while at full load the v-drop of the diode wouldn't really change that much - so the divided V measured would like be still smaller compared to the original shunt at 1-2A so the current limit shouldn't drop much - like 2x to 3x times the minimum required current for continuing to work - Schottky would go from like
This video reminded me of a car ride when I was little. It was hot and I decided to make a portable fan to cool down (car didn't have ac). Like a day prior to the trip I got an old usb cable and a 12V fan and connected them together. The fan was spinning pretty slowly but it was enough to get a cool sensation on your skin. The drive took a few hours and 3/4 of the way I was clicking the button on my power bank to wake it up like every 30seconds becouse the fan wasn't drawing enough current to keep it awake. Good times...
Haha thanks for the story :-)
@@greatscottlab Thanks!
You had USB when you were little!? Now I feel old
@@stefflus08 Yes, in a while I will start to feel old.
@@stefflus08 and powerbanks too... yupp, feeling old over here too...
Literally from the datasheet for IP5328P:
"If key is pressed, whether or not load is on VOUT1 port, VOUT1 port output will be force to open; but the
VOUT2 and USB C port will open only when load is detected on the according port, or the output on these port
will be closed. So VOUT1 must be retained if only one VOUT port is needed, if VOUT2 or VBUS is not needed,
MOSFET can be eliminated, but the 10uF capacitor should be retained."
And, if you need all ports to be working - thus lesser voltage on VSN pin. So then cut the trace and attach a voltage divider(100 to 1 to get 4.95v, 1000 to 1 to get 4.995v) from shunt to the VSN, simple as that.
but that ruins his example... lol
Someone needs to get this man a bolder marker so that he doesn't have to go over everything he writes multiple times.
I like it that way though ;-)
😂😂😂 rightly said
Leave his OCD alone. 🤣
Not sure if you are joking, but I think he uses this technique for emphasis as he completes and explanation. He appears to have had drafting experience as well, so pretty sure he has a decent stock of engineering pens in a multiple of thicknesses.
Frankly, I really like this low-tech way of identifying and emphasizing an illustration or number.
@@allthings3d I was joking, yes. I appreciate the drafting information though. I was not aware of that. Thanks!
Great idea! It's very frustrating that with so many little gadgets getting USB-C nowadays for their tiny batteries or low power draw, your best quality power banks are the most likely to refuse to recognise them.
@LewisOSLife2984Patreon.
@LewisOSLife2984the video is released earlier for people who pay
@LewisOSLife2984 Patreon viewers can watch early 🙂
@LewisOSLife2984Time Travel
I really hate USB-C. It should be a standard, not a family of standards. There are already dedicated connectors for all the things it does, but there is no confusion what port and what cable support what standards, what bricks can deliver what PD mode, etc..
I solved this in the past by having timed pulse on one of the spare pins on Arduino controlling a fet with 150 ohm/1w resistor. To power up the power bank I wired another resistor with its power button. Works great.
Nice Video. But a few warnings are missing for unexpierenced people. Especially for soldering Lithium Batteries
I love the idea with the 555 circuit and if you want to make it even more efficient you can use as load a capacitor that discharges via a diode at the 5v output or feeds the 555
Fascinating indeed! I guess the solution would be to go with USB C and embed that circuit into the project itself (so you don't need 2 USB cables).
Thanks, dude! 😊
Stay safe there with your family! 🖖😊
Good point!
Does USB C even need anything more complicated than a couple of resistors on the CC pins?
@@Vaionko That's a good question, I don't really know.
I believe the PD protocol requires some kind of handshake between the device and the power source, and this means you actually need to integrate some suitable ic in your PCB
@@giovannigrosso6920You can buy little PD boards for around $2. Very useful
The older, single cell powerbanks don't automatically shut off, no matter the load. I upgraded the cells in a bunch of them, to have more available capacity.
It can be insanely frustrating trying to buy a power bank when most don't publish the full specs or features they have. It can often feel like a gamble. One feature I like is pass through which allows for the battery to charge while also discharging it great to use like a sort of UPS. Also nice for connecting small solar panels too things you need to power throughout the night until the sun comes back up. Maybe this would be hard but it would be cool to be able to buy one endorsed by you
I built my own using a TP4056 and a SDB628 (MT3608) boost converter module for less than 50 cents for both!!!
Added two USB connectors, power switch, terminal blocks, and enclosed in a 3D printed case. Voila. Always on @ +Whatever volts.
That boost converter is magic. works down to 2V which consuming microAmps of quiescent current. Brilliant times.
You just need to spend more on your powerbanks. The ones I have can be put in a trickle mode which disables the low current shutoff.
Yeah I have an Anker one. I just double click the button on it, and it goes into trickle mode. Works great
@@mattanderson111 Mine are Ankers too, l wouldn't buy one without the trickle feature, it's very useful.
Honestly never heard of this feature. But awesome that something like this is out there.
I think these approaches are still very valuable, personally I'd rather use my old power banks that have nothing else to do to power my projects rather than buying more new expensive ones and wasting the old, otherwise perfectly good ones.
If you're going to that much trouble, just get the lithium battery and the battery bank circuit and use it for charging only, then wire the lithium battery to a voltage circuit or just use as it for your project.
For low power projects it's easier to take 2s liio batteries with xt30 which is usually made for RC stuff and add simple usb converter and over discharge protection
IMHO just buying a cheap type C trigger board that lets me use any of my existing power banks is much easier than effectively building a custom power bank from scratch
I was looking for this topic previously, now I know how to keep them alive
Glad I could help :-)
A better option is another lower power piggybacked power bank of your own design. Just need to isolate and regulate output from charging input so as no over voltage or spike on the low power device at the very end of the chain! Just need to turn on the first in the chain occasionally to charge the second. Easy and less work. Simplest option of all: plug into a mains usb charger which won't cut out for low power devices! :-)
I use a lot of those power bank modules in my projects. I go around this limitation in my projects by using one of the MCU pins in the main project board to drive a transistor to flash an LED or just shunt current through a resistor intermittently to keep the circuit "alive".
Basically what he showed in the video lol
@@fusseldieb Mine is actually part of the circuit and not an attachment to the USB port. Instead of a 555, the flashing is handled by the MCU of the main circuit
Ive been using that 555 timer to wakeup my bluetooth board for years. Never failed, works great and cheap.
From the datasheet"Discharge:
If key not pressed, only when sink device attached will the output port open, non-attached output port will
not open. If the opened output port current is less than 180mA @ 10mOhm, it will wait for a period of time before
close the port automatically. The output current is detected by the voltage drop between VSN to output, 180mA
current is equivalent to 1.8mV on 10mOhm, when the resistance (including MOSFET internal resistance) between
VSN and ouput is larger than 10mOhm, the current threshold will be decreased proportionally."
So all you need to do is attach a 555 timer to the KEY pin and pulse it periodically. It should stay on without drawing extra current or messing with the current sense
Funny, because i recently made exactly the opposite experience. 😄 I bought absolutely cheap beard trimmer for traveling. But this one had a) old NiMh batteries and b) an extra adapter. So i used a step down converter to the used 3V and an old USB cable to use it with a standard USB adapter. But than the charging always stopped after less than a Minute! I thought a made a mistake but than i realized that this only happens when the battery is full and the flow is dropping. Which is perfect because the original Adapter dopes not stop charging, so you never knew when the battery was already full. But because of the auto-stop of the converter now it does. So now i have a USB cable to charge and additional an auto-stop. Win-win. 😄
Have faced this a lot in the past. But never tried working on this as I didn't need it too much and also got busy in work.
Instead of trying to increase voltage of the IC VSP pin , it might be better choice to cut the VSN pin of the ic and put a resistor between shunt VSN TO IC VSN and a another resistor between IC VSN TO GROUND so that the ic vsn pin voltage get reduced than IC VSP voltage, may be it will think output current is above the threshold value
You can just use an usb-c powerbank with a type-c breakoiut board, set both CC 5.1K to GND with a resistor to get 500mAh output on that port and keep all other ports awake with higher output.
If the powerbank comes with wakeup button (or the control IC has a pin for that), you could use a 555 timer to trigger it in certain intervals (depending on IC's sleep timer... check datasheet). Should be even more efficient, than the 555 Timer with load solution.
Or, you could check the Datasheet, whether the sleep timer itself can be set/changed (e.g. via resistor). Though that is unlikely possible in most.
Oh, and to save some more power, you use a CMOS variant of the 555.
Simulating dead man's button... nice... but you need to open the powerbank and solder. If you for some device use multiple powerbanks (e.g. replace with another one newly charged) you must do this modification for each of them.
Simulating dead man's button... nice... but you need to open the powerbank and solder. If you for some device use multiple powerbanks (e.g. replace with another one newly charged) you must do this modification for each of them.
@@vencdee I mean, there are powerbanks, that can be easily opened. I personally prefer powerbanks that are filled with 18650 cells, that you can easily replace. On those you have easy access to the PCB.
And if you do many projects, you might want to add one additional to the top of the list, a custom powerbank. There are plenty of good pre-built powerbank circuits available, along with 18650 cells.
But of course, there are different solutions for different kinds of people with different skillsets and levels.
for these earlier power banks, you can put a switch on that resistor circuit, so you can get what most commenters are saying: a "trickle charge" or "low power" mode where you basically switch between the ultra-low power mode and normal operation, depending on your needs
Great designs! I think the 555 circuit is plenty good enough for people that need the convenience of a powerbank. The "pre-made" power banks are meant for expedient use anyway so as long as it's not discharging the battery for nothing it's fine. People that actually care about doing a "most efficient battery powered IoT whatever" will just embed the battery and add a basic BMC in their device (ie the little boards that just protect the li-ion cells from over/under voltage and overcurrent).
Nobody should be getting a power bank and then try to hack it, just get a BMC board and cells and make a "dumb" battery
I highly recommend big powerbank empty cases for 18650. It's the best case for DIY electronics.
- no need for hack like in video, just connect pin as input and u can read button pushes for custom special actions and every 10-30s set the same pin as output high for few ms to mimic single button push and that will prevent powerbank from turning off.
- dont use all 18650, take one of two less and put mcu there..
- cut plus or minus line from usb port and put mosfet in the middle and u have 5v line that is being controlled by PWM
- if u need more control then single button can provide u can remove LED and put pushbutton there (no need to drilling that is never perfectly centered so if have OCD u still be happy how project looks in the end)
- case for 8 cells (and u can put up to 7) costs only 5$ (cheaper then plastic case, usb ports and li-on charger separately)
“Micro watt hour per minute” is my new favourite unit
Another annoying problem is that most powerbanks turn off the output when you connect a charger. Otherwise you could use them as a small UPS. For a raspberry pi for Example...
I've figure out that this also depends on the voltage of the charger. If charger has a good cable and the voltage on the input port stays close to 5V my power bank doesn't turn of the load and indeed works like a small UPS. But if you plug it in a weak charger and the voltage on the input drops to say 4.5 it disconnects the load until fully charge itself.
I wonder if instead of pulsing a load resistor you could supply pulses to a capacitor that then powers your project, so it would pull the same average current but in narrow spikes that ought to trigger the load detection. On the plus side it should still shut down if your circuit is properly off, but on the minus side you'll have ripple which might or might not be a problem?
With the units I have they seem to play the game of averages... which is likely why he went after 100mA pulses when the verified steady current threshold was only 50mA.
It's hard to tell *exactly* what is going on inside them but it would seem that the longer a load has been near the threshold, the more readily it will trip with just a slight further decrease in current.
I think the pulse approach has a benefit efficiency wise because it may take a bigger hit on the averaging cap, keeping it further away from the cutoff point.
That said, this is essentially what buck and boost converters do. It's an "everything you got for a moment" followed by "let that soak in", on repeat.
So you could probably half get around this simply by using another converter that follows the one in the battery. It isn't going to really "gain" you much tho, you're adding on another "base" load that doesn't actually "do" anything, putting you back in simple minimum load resistor category.
A couple years ago I made a 3S pack using a cheap BMS board and a buck converter, put a toggle switch on it and wrapped it in fish paper (after I had an oops and had to replace the BMS board). Crude but effective. I did have to dedicate a power supply to it for charging with a current limiting circuit hacked onto it to prevent the supply from shutting down on overcurrent when charging.
Hi,
as many projects require 3.3V and not 5V, you can use a 3.3 LDO regulator with overdischarge protection directly on the li-ion or li-po battery inside the powerbank. 3V is enough for low power projects (SMt32 + data acquisition + oled for exemple) , a low cost LDO is sufficient in this case
The 555 timer ic makes a come back! 🙌
Best DIY electronics channel on TH-cam.
Im sure I’m not the first one to mention this but most usb powerbank have two outputs, so you could put the timer in one and your application in another. It would probably be fairly easy to miniaturize into a tiny dongle.
I've built such a circuit on a breadboard a while back. I chose the 555 timer route because it's simple and dosen't draw much power. My powerbank (20000mAh) turns off after 30 seconds with no or too little load. That current threshold is around 70mA with my powerbank. So my circuit draws around 150mA for around a second every 20 seconds. My (USB operated) NiMh battery charger can sometimes draw too little current for the powerbank to stay on.
It's nice to see Scott using Everyone's Favorite Timer Chip in a project again. Coulda done it with a cheap, ubiquitous microcontroller, but where's the sport in that? (Seriously, as much as I love the 555, there's something to be said for a chip that only requires a bypass capacitor and a load resistor to perform the same function, and the micro could easily be heat-shrinked into a tiny USB pass-through dongle.) By the way, Paul Carlson built a similar 555 timer "stay awake" circuit a few years ago.
It still bugs me that power bank enginerds don't realize that many consumers need this functionality.
Should be considered basic functionality to charge something like a headlamp :/
It feels wrong using an entire microcontroller to just pulse a circuit though :/
As for power bank engineers, they do realise, it didn't actually take very long after low power type C devices like true wireless headphones came on the market for power banks to start including low power modes. The problem is that the cheap power banks just recycle the old designs so you usually only get these modes on mid to higher end ones.
@@bosstowndynamics5488 I meant the ubiquitous $0.03 one-time programmable 6- or 8-pin chips. Just a few lines of code to grab a brief pulse of current every few seconds, 100 nanofart bypass capacitor, appropriately sized load resistor, USB connector traces directly on the PCB...Bob's your auntie.
What size are those shunt resistors (1206)?, you should be able to find or stack resistors rather than just unsoldering it, however the testing to figure out how to fix the shutdown problem is pretty awesome!
Keep em coming!!!!
Those are great! I love the pulse circuit! Ingenious!
had that shutdown problem once, and since then, I have bought power banks that support low current devices. With the Anker 737 I have right now you can hold the button down for a few seconds to lock on. And draw as little as you want.
Although he wasn't the first to make use of a 555 for this purpose. I recall Mr Carlson's Lab made a version in a video titled "Power Bank Stay Awake!" over 3 years ago.
If you cut the trace to the LOWER of the two current sense inputs and add a 1:10000 voltage divider, the measured current will be about .5mV higher than what it really is. No opamp required.
The only drawback is that the "amplification circuit" inside the chip may expect an about 10mOhm output resistance of the current sense resistor so it might be designed to draw a significant current. But I'd start out with a pretty high-impedance voltage divider.
At such a ratio the big resistor actually becomes "almost a current source" So you almost get an actual .5mV offset as you wanted. Try 1M/100 Ohm, if that doesn't work, try 100k/10 and 10k/1 ohm respectively.
Some powerbanks have what's called a "Trickle charge" mode.. This mode forces the powerbanks output to always stay on. I've seen it being a thing on some of Anker's powerbanks, where you longpress the powerbutton and the indicator LED turns from blue to green...
IP5328 have I2C communication, u can easily configure this behavior throuhg it. First power on by button , to power the mcu, mcu sending "always on" to IP5328. This shunt 10mOhm it's for charging/boost current sense. I bet the whole IC get crazy after you changed it :)
The datasheet lists a parameter about light load detection which uses the voltage between the positive USB voltage rail (connected to pin VOUT1/2, VBUS) and VSN. Seems like they use the output FET as a shunt. If less than 1.8mV is measured, light load is detected. Maybe a low forward voltage diode (e.g. Schottky diode) between the positive USB rail and VOUT could do the trick without affecting charge/discharge current protection. Although, the datasheet is a bit conflicting at that point 🤔
Maybe someone else already suggested this solution, but probably the easiest one is to modify a power bank by removing the stock PCB, and installing two premade devices:
1) Li-Po charger with USB-C input
2) Dc-dc converter to have constant 5v (or another voltage as required)
3) (optional) for those power banks having more outputs, add a secondary dc-dc converter configured for an higher output voltage, like 12V
In order to avoid discharging, all dc-dc converters must be connected to a switch, in order to cut the power when not in use. 😊
I just replaced the momentary button with a self locking one
it kept the power bank on forever
the bonus here that stand by mode could be switched on or off 😁😁
I use the pulse method in a small circuit mounted on top of the 555 itself that I put inside the power banks and it pulses the output every 30 seconds for 1 second, and it works fine...!!! Great video.
The venerable 555 still being useful!
If your project contains microcontroller and you have spare output pin and some space on board you can program pulse draw, or use second small dedicated microcontroller like Attiny5 and some transistor with load resistor to do it.
Yup, I've done this. Spare PWM pulsing the power with a MOSFET.
Many power banks do come with a low power draw mode. What I would like to see is to **how to use them as UPSes** since most of them turn off for a second when AC power turns off or on.
I think rather the cheap ones turn off when AC-power changes to have a cheap but still safe renegotiation of protocols and capabilities.
But you can probably be super cheap and simply add a large enough capacitor to survive the downtime.
While they do work greatfor UPS applications, such on-line UPS systems wear out the battery quickly. better would be to use it as standby once power drops (like most USPs do)
@@tarakivu8861 Would it be safe to add a 1.5F 5.5V Super Capacitor across the power bank's output? I would like to use it as a UPS for raspberry pi 4.
Modifying a power bank to act as a UPS would be much more complex than these mods, so it probably makes more sense to just buy a power bank with passthrough charging/UPS mode to begin with (fairly common these days).
I bought a power bank that has 4x 18650 protected button top cells wired in parallel and they are user replaceable with a slide off battery cover. I slide the battery cover off and insert the jumper leads to + and - bypassing the the electronic circuit board if I want to use with low power devices like micro controllers. It provides plenty of runtime for small loads, and if I want to be more exact with voltage or use maximum amount of power from the 18650, then I can plug a small DC-DC boost converter.
My powerbank can actually be set to a "disable auto shutdown"-mode which also switches off any fast charging. It just provides stable 5V.
It's a ROMOSS Sense 8P+.
The mode can be activated by a double tap on the on/off button while it's on.
Ultimately, the 555 timer solution is the most practical and universally applicable.
Alternatively and what I think is nicer: a small 3.3 V battery that is connected to the power bank via a relay/Mosfet to wake it up. (when most power banks are last switched on) The small battery is permanently connected to the Arduino. Advantage: The small battery can be recharged directly from the battery pack depending on the duration and wakes it up at the same time. :)
Use a shunt resistor that is 2-3 times higher reisstance than the one on the borad.
I experienced it alot, An led can also be used as a load for output spikes preferably an SMD one.
Or just force button press of the power bank to keep it turned on all the time using of course a signal instead of manually physically doing it.
I like the second solution: Periodically drawing enough power to keep the power bank activated. You could incorporate this in your project, by periodically enabling an output of your Arduino, for example, to power a LED.
Thank you for the 555 schematic! I keep having this problem with older usb power banks I want to use to run LED lights with.
Glad to help
Yay for the 555.
Useful in 1978*
Still useful in 2024!
* Marketed in 1972. But I was using it as a kid a tad later 😅
Think the cmos version came out in my youth too...
Thank you a lot for making this video, I was SO over with this feature.
Glad I could help!
Also a problem with some small synthesizers. I found that some old small cheap powerbanks don't have a switch-off (ir it is very low). One soultion: using a USB hub and plugging in a USB lamp additionally. Then it stays on.
Of all the options you provided, I liked the one that gives pulses the most
Great video!
I like the pulse circuit. Could dial it in to trigger right before the shut down of the power bank to minimize how much power is drawn instead of the 8 second interval used for demonstration. Maybe as a subsystem added to the small projects themselves to allow function with any power bank without external modules.
I know the current draw is quite small, but were there any voltage dips presented on the output side from the pulse?
I recently had that problem, thanks for sharing your solutions!
thank you so much !!! i use an old powerbank to light up my gunpla, but i had to "turrn it on" when i wanted to show to someone.
Could you not bypass the shunt completely - just swap it with a 0ohm, so 5V goes to both VSP and the USB port without interruption, then put a 1ohm resistor from that 5V line to VSN, so regardless of the current actually drawn, VSN sees the drop from 5V to 4.98V (assuming VSN sinks 30mA) and this is over the 18mA threshold to keep it on?
Yeah what other me said... I feel like I'm missing something obvious though!
Add custom circuit directly to battery inside, add switch to flip between original circuit (charge device) and bespoke circuit (low power output)
The 555 timer was the first IC I learned how to make circuits with back in highschool in my electronics class.
... Back in 2000, (Class of '02)
You could also add a switch to bridge your higher value sense resistor with the smaller resistor for the times when you need to supply higher currents.
Assuming your project has microcontroller you can use it to time pulses to load thus skipping 555 timer circuit entirely
I also go back over my sharpie marks repeatedly until it bleeds though the other side of the paper. 50-60% saturation is just not the same as 85-100%
I found another hacky solution for my powerbank: I just press it's button once every few seconds (with a NE555 timer), which lets it stay on.
That does not even draw any wasted current (except a few mA for the NE555) and works perfectly (for me)
One thing I would like to see is a low voltage input cutoff on powerbanks. I have several that will draw all the way down and not shut off when they get low.
I don't like the batteries going that low. Luckily I'm using recycled laptop batteries that were free. They would last longer if they didn't draw down so low.
low power mode can be activated on majority of branded powerbanks.
the real problem is that some powerbanks may turn off after charging your device instead of acting like a wall charger (until power bank drains).
Hi Scott, here is my solution: I had an esp32 lying around, wrote a short micropython-script, it connects to wifi and starts WebRepl (both not necessary), then goes into deepsleep for 28 seconds.
I also "removed" the LED (with an electrician screwdriver).
I'm sure it will draw more energy, but it was simple for me to implement.
Thx for your videos from bremen!
use trimmer and set it on 1 ohm every time you need max output current, set it higher if you need small current draw without power bank turning off.
I work mainly with phones. For example, just by the current that the iPhone draws, you can tell what is damaged in it. I have a 20,000mAh and 20W Powerbank from BaseUS and it drives me nuts when it turns off after 15 seconds of no or low power consumption (around 0.05A). I was thinking about making a PCB for it to artificially consume power, but I put it off until later. Btw my old Xiaomi 10 000mAh Powerbank with 5V2A output supports low power mode by simply clicking power button 2-3 times. Then it will never turn off, but I can't always use it, especially when I need Fast Charging. Thanks GreatScott!
I'll bet this guy never saw a bad day. What a pleasent thought.
there are two common problems in see with power banks.
1. threshold voltage over time. your video addresses this. most i have tested cut off after around 30 seconds with small current used, or the consumption is too low to even keep it running. yours sounds to solve this.
2. passthrough charging
use case: sbc like a raspberry pi using a power bank like a ups. it should only allow mains power until there is an outage then kick in. most power banks do not have passthrough chargung. it would be helpful to have a power switcher that would activate any powerbank when an outage occurs, probably with a capacitor in order to keep continuous power right when the power bank is about to activate. that way you can use any powerbank in whatever useful form factor to go with your project.
by default, powerbanks simply drain out. you don’t have the option to use mains then power bank as secondary. some solutions are clunky with a custom circuit and no power bank while others just use a power bank tethered to mains but it is still using the battery first.
This is a great succinct example of why everyone hates software and technology: A designer or company makes one assumption, and there will always be a huge number of people that don't want that as a feature. And thus nothing ever does exactly what people want
The shunt resistor mod gives me an idea... what about a voltage divider on the 555's reset pin, carrying (or transistor buffering) current? Pull RESET low when current exceeds a certain threshold, so we're not unnecessarily pulsing the load when the attached device exceeds the minimum draw. Might be useful for something like a remotely deployed ESP32, where it''s mostly idle, but occasionally has a bunch of Wi-Fi connected work to do, for example.
You can probably get away with not having a transistor, because the NE555 can easily sink 100ma. These old school ICs are tough cookies 😂. I normally drive relay coils directly with the NE555. Maybe we could do a very minimalist version of the circuit. I love the idea, and I also have been frustrated with 5v power banks turning off 🤔😩
You could probably just put a 10k or so(would depend on input impedance of course) resistor between the shunt and the VSN pin
Voltaic brand power banks have an 'always on' selectable mode. Useful for when you use it as a 'UPS' for USB devices...
Disconnect the VSN from the circuit. Then make a voltage divider circuit that will feed in a delta 50mV to VSN.
Let say you constantly draw 10uA on your divider circuit. 50ohm from VSP to VSN, and 500k from VSN to GND.
A simpler way to do the pulse load would be to use an ATMega328. With the 5v supply it would not need any external regulator, just a bypass capacitor at VCC. It draws around 8 µA with the watchdog enabled in sleep. Just wake it up every x seconds and parallel four or five output ports with a 47 ohm 1/4 w resistor attached to Vcc. The outputs can handle 20 mA each (easily) and you can use the internal clock so you just need one chip, one resistor, one bypass cap and you’re done.
If you already have a microcontroller within your project, you can have keep-alive at the cost of one pin, a random tiny little MOSFET, and a couple resistors, one to tie down the gate against floating and one for the actual load. Or a random BJT then one resistor is to limit base current, it doesn't necessarily need anti-floating.
The engineer who thought that shutting off the power bank if it does not detect 500ma is a id10t. Its in the same field as a cloud service for appliances.
500mA is quite big.
@@greatscottlab Agree it is a dumb circuit lol
Well these things were mostly designed for phone charging, and even old phones usually drew at least 500ma. Newer banks take into account that you might be using them for other things.
The IP 5328 chip is programmable, granted it only mentions charge curves, but probably also has provisions for setting the auto shut off value. Using TP4056 and a 18650 cell we bypass the cut-off problem altogether. And while the TP4056 draws 20µA that will take awhile to discharge with no load attached.