Introduction: DIY Portable Variable Power Supply
I've decided that i really need a small, portable, variable power supply.
It will be helpful with a lot of things that can't be moved near the bench power supply (or if it's not easy to do so), like test a small DC motor, reactivate a battery with a BMS, power up a standalone arduino project, test an ambient led in its supposed place to see the real outcome, and so on...
This project is the "making of" of a device I saw on Thingiverse, which seemed to be a really nice and compact design. I've made only a trivial mod to the orignal project but imho the OP instruction video is nice but not complete, therefore i'd like to add further info and hints.
Safety first: this project involves soldering on lithium batteries, connecting everything to not-certified electronics and everything is likely to be put together by not professional people with hobby tools. I'll do my best to explain the best practices but if you proceed the responsability is up to you.
My mod simply adds a DC connector, in this way i can easily build adapters to whatever connector, like the T-dean, XT, dupont, 9V rectangular batteries, alligator clips (for their specific tasks)...
I don't like to mandatory have "floating" alligator clips, with their exposed metal and maybe voltage applied, because I really don't want to increase tha chances to have the magic smoke escaping from my devices.
Step 1: Components
To build this power supply you'll need few things: a couple of 18650 cells, a TP4056 module, a DC/DC boost converter (step-up), a small voltmeter, a small switch and whatever dc connector you like.
Links, links, links! (affiliated)
You will need also the 3D files to print the case.
Step 2: Tools
Step 3: Electrical Wiring
The wiring scheme is quite easy, you have to perform these steps:
- connect the two 18650 (in parallel) to battery pads on the TP4056 (B+ and B-)
- connect the TP4056 negative output OUT- to the step-up negative input VIN-
- connect the TP4056 positive output OUT+ to one pin of the switch
- connect the 2nd pin of the switch to the step-up positive input VIN+
- connect the step-up out pad to both the dc connector and the small voltmeter (respect the polarity!)
The only difficult task here is to deal with the small room for the components and the cables lenght (which is quite short). As you can see if you examine the printed case, there are holes and routes for the cables, you have to use them to close properly the case.
Step 4: Electrical Considerations
The OP assembly video is quite self explanatory but imho some additional details should be discussed. In the end it's a power supply, then the main part of an assembly tutorial "should" be related to electronic instead of plastic.
First warning: i didn't see any indication about the voltmeter trimming. You have to use a small screwdriver to calibrate the voltmeter BEFORE sealing the case with the glue, otherwise you won't have a precise voltage regulation. The trimmer is shown in the image, red arrow. Pay attention not to create shorts with the screwdriver.
USB ins't the only charging option, you can use also the TP4056 +/- input pads, which accept a voltage beetween 4 and 8V. The typical supplied voltage is still 5V (like the usb) but it could be interesting to know what could be used to recharge this device. For instance: a pre-usb-era old mobile phone charger, with its original connector replaced by another one, compatibile with the one you could have soldered on the TP4056 input pad (like i did in this other project).
You have to balance the cells before any parallel connection! Otherwise one cell will discharge charging the other one, as fast as they can, and this could be way over the safe charge/discharge speed. It could be dangerous!
Keep always in mind that the 18650 li-ion cells are intrinsecally dangerous! Soldering on them is not trivial at all, it could be done of course, but you really have to understand what you are doing. 18650's gurus like this guy actually prefer to solder the cells instad of using a spot welder, and they know what they are doing. I'm quite sure that a cells' soldering tutorial is available in pretty much every language, so please spend few minutes to understand the process and the risks in your mother tongue (or whichever language you are fluent in). For instance, pretend to have a drop of tin falling in the small gap near the positive pole, shorting the cell and impossible to remove within seconds. Well, that's why i do my stuff with a bucket filled with water near me. Never actually used, but better safe than sorry, right?
Maybe someone is guessing why the OP used two cells. Not one, not three, two. Well, i'm guessing too but i have few info that i'd like to share. The advantages of two cells instead of one are quite easy to understand: you have two cell in parallel, therefore additional capacity and you could use also some old and low performing cells. The TP4056 is set to protect a cell with nice performance, and old worn out cell might be stressed by that. The standard TP4056 have an output current limit of 1.2A, not a lot but it could cause heat generation if you use cells with high internal resistance (aka: old and/or worn out). With two cells in parallel the 1.2A current will be splitted. It's not suggested to use old cells, of course, but let's say that's a smaller issue if they are two.
Following this reasoning, why not three or four cells? Well, two cells it the suggested maximum cell number to be connected to a single TP4056. Why? Because of charge safety. Not to mention the eternity that would be required to charge 4 3000mAh cells in parallel with no more than 1A of current. While time might not be relevant, safety is! The issue here is the mechanism that will stop the charge. After the constant current (CC) phase the TP4056 will switch to a constant voltage (CV) phase, and it will check the current consumption to determine whether the charge should proceed or not. there's has a threshold current, let's say 25mA, which is the limit beetween a fully charged cell and a cell that can accept other energy. The cell won't stop to accept current by itself, that's why this threshold current must be monitored. If you add to the equation that the supplied voltage might not be exaclty 4.20V, maybe it's a little more as per the TP4056 data sheet, it will be even more clear that you can't just apply a stable voltage to a lithium cell. Now, how can the TP4056 determine if it's charging one or ten cell? Easy, it can't! Therefore it won't have any clue if it's overcharging -for instance- five cells connected in parallel or correctly charging a single cell. As long as the five hypothetical cells adsorb 5mA each, which is way less than the 25mA rated as safe, for the TP module everything is proceeding in the proper way but it's not!
Step 5: Performance
It's not like a portable powerwall but...
I've used two reclaimed 1700mAh for this prototype (yes, i'm wondering if I should design a V2.0), therefore i have 3400mAh @ 3.7V. The maximum current for this cells is constrained by the TP4056 overcurrent protection, which is 1.2A for most part of the modules. If you use reclaimed cells you should check that both can more or less handle 1.2A, just to be sure. New cells from reputable suppliers will be ok. Do you know how to perform a discharge test? Well, google it, you'll need few other tools.
Since the power supply output is variable it's a nonsense to ask "how many amps can i have?" because we should talk about how many watts it can provide. To stay safe we can take 4W on average but when the cells are drained the limit will likely be 3W: max current times minimum voltage 1.2A x 2.5V = 3W. Please note that this doesn't depend on the cell's quality but on the TP4056 capability instead: Better cell will only have this power drop later in time, which is really nice, but it can't avoid it.
With my cell this means that i can have a 5V 800mA output for a couple hours at least, with the proposed ones you should have the same output watts but for 6-7h.
Step 6: Case Assembly
On the OP page there's a video that explain the physical assembly of the case, so it's not really useful to describe the same process by words, i'll only provide some additional hints and photos.
First hint: the hole for the step-up trimmer (the one that will change the output voltage) could end in a wrong place. This is not an OP fault, i've checked and two batches of step-up modules from the same supplier have slightly different trimmer position. Anyway, not a big deal, you can easily adapt the case with a cutter or a drill bit.
Second hint: the OP didn't explain a lot about the switch, in the photo you can see what i found. It's similar, it works, but i had to slightly adapt its site with a cutter. Like the first hint, actually an easy mod.
Third hint: i choose to mount the voltmeter upside down, to reach the voltage trimmer with my index finger having my inch strongly grasping the device. Totally unnecessay, in the end i've decided not to have the trimmer plastic cap mounted. It's difficult to glue and i don't want to have the trimmer reachable from the outside of the case, exposed to bad handling. A small flat screwdriver is a better (safer) option for me but the choice is up to you.
Now power up the device, test that everything is working properly and proceed to seal the case with not-so-rapid glue (you will appreciate the chance to properly align everthing). Few rubberbands and a small clamp are perfect for this task.
That's all,enjoy your new variable portable power supply!
Step 7: Drones and Other Useful Tricks
Since i've added a standard DC connector it's really easy to assembly different adapters, the first one can be connected to the mostly used RC drones' battery connectors.
In fact, the best use for this device i've find so far is to test and power up for the first time my drones. You really don't want to use a lipo battery (or anything that can provide a lot of amps) to test your last drone build: if there's something wrong, chances are that a lot of things will be damaged, if not literally blown up. With a power supply that can provide only few watts even some wrong connection "could" be recovered, and that's comfortable.
Not only, since it's variable i can test (even at the field) if every safety sistem is working properly, one more step to do in the pre-flight check list that could save the day: if the battery voltage defined threshold was properly set and the proper warning signal was tested A LOT of expensive drones, piloted by skilled people, would have come back home.
As you can see i've also used a 9V clip adapter to power up a multimeter with a drain battery, it can measure it's own former battery so it's clear that this device could help in some emergency case.
That's all, happy portable variable power supply, folks!
ps: if you find this -ible interesting or it helped you please consider a little paypal donation, it will be appreciated.