DIY 100W Buck Boost Power Bank Modules And More. Which Ones Work? IP2368 IP5389 SW2303 SC8812A 

You using the original LG M50's? Honestly don't ever really see those. Mostly see the M50LT's nowadays. How do you like them? You're getting really close to the max charge current there at 100W. I typically use a 65W charger to go easier on my cells or if I want to go really easy I'll use a 20W...but it really seems to stress the 20W as its pulling everything it can the whole time its charging...same for the 65W but the charge time is nowhere near as long. Nice to finally hear someone can output 100W from that board. What gauge wire are you using for the main positive and negative? Short length?

@@ericklein5097 I’m using battery holders which I put together soldered together with nickel strips. And a thick short run of wire to the board. Not sure what the equivalent gauge of wire would be but it is definitely overkill for this application.

@@TheWretchedWorld be careful with the battery holders. Besides the high quality Keystone variety with the gold contacts most of the cheaper ones add 30+ mΩ of resistance to each cell in the pack. The ones with springs would be the highest resistance of all with the spring possibly turning into a fuse around 10A. Since you're using a 2P pack the highest current flow you'd see is about 2A from each cell. Jehu Garcia loosely rates the cheaper cell holders as 1C (so I take that as 3-4A) and the Keystone ones as 2C (so maybe 7-10A). I think I measured the Keystone holders to have a resistance of 7mΩ per contact so 14mΩ each....which is nothing in a normal battery pack thats run under 0.5C but for any kind of "power" pack thats running 1C+ you gotta go pure nickel/copper, brass, aluminum, or some combination of one of the top 10 most conductive metals.Definitely hard to spot weld pure nickel with cheap hobby welders (0.15mm thick is about the thickest you can weld under 1000A) but the nickel copper sandwich method works wonders and if necessary you can use nickel plated steel as your top piece that increases the resistance of your weld spot to get a good weld on the copper below. Cell holders work just fine for testing a setup though. I think 16AWG is 1.5mm2 (1.31mm2 more specifically). TZT told me to use 16AWG minimum and I think having a 6' length of skinny wires pushing 8A might have caused enough voltage drop to prevent 100W output. I was able to charge the cells at 100W for some reason but I couldn't output any more than 60W. Happens to me with quite a few boards I buy claiming to be 100W output. My hatred of the SW2303 stems from all 5 I bought being absolute turds. Thanks for the insight. I'll have to give the board another chance now that they make one with a screen.

Thanks for these comments, i have 36pcs 21700, i think il go with 5s version too, il definitely need that high charging amps at 7parallel

@@winmarkesconde5651 The bank is still working flawlessly, the board is a bit quirky. It will sustain 100w out all day but when trying to charge it at 100w it will quickly throttle down to 55w. Still works flawlessly apart from this.

Happy to help people save money when there’s so much junk out there. I need to do an update on the IP2368. Great board but it has one known issue. If the board is disconnected from power it must be plugged into a charger to wake it up every time. I don’t remember seeing this warning on Feb 22 Store in the past but just recently I noticed they are warning people the “old” model with this problem and the “new” model with it fixed are both being shipped out and you might get either. In Chinese BS seller talk that means “you’re going to get the IP2368 with this bug. There might be a new model in the works to fix this” This presents a conundrum: add a switch to cut off the board from the cells and then have to charge before every use or leave it forever connected to a pack and watch the first group of cells get drained down while the other 2/3/4/5 groups don’t lose any voltage. After sitting for a couple weeks I had 3.42, 3.77, 3.77 and 3.77. Manually charged the first group to 3.77 and I’m already down to 3.54 a week or so later. It deserves its own video and trying to see if the 120W module does the same.

I still can't believe that is possible! 150W into a little phone battery even makes me nervous. Would love to see it in person. I wonder how much it heats up.

@@ericklein5097 i know right!! I thought 33w for my oneplus nord was a lot, and it gets pretty warm when charging. IIRC they have a temp target of 50C when charging, so they throttle once they reach that temperature

@@ericklein5097 "realme gt neo 5" have charger 240W 20v 12A

Yes but you will have to buy an IP2368 already configured for 5S or change the resistor yourself. Its a small SMD resistor and not the easiest to change but its just one resistor. I always recommend using power cells or a 2P/3P/4P pack if your primary use is 100W output so using power cells in a 2P is very good. You should have no problem using that parkside battery. The other two pins are likely some sort of communication or power lines are doubled up. I don't know Parkside batteries. I would check for voltage on all 4. Have you taken the battery apart? If you don't change the resistor to 5S who knows what kind of problems you could have. Certainly it would allow the pack to drain far too low and kill your cells. Be careful charging the pack. You have to use a USB charger and the wattage of your charger determines how fast the pack gets charged. A 100W charger is going to send 100W in so 10W to each cell if you have 10. Thats a fairly safe charging rate for most cells. It would be about 2 to 3 amps.

I'd love to try that but my TZT board with IP5389 stopped working! POS (Piece of Shite) Might try one of the other IP5389 boards from Feb 22 Store

Out of curiosity, which IP5389 board was it? The one sold by TZT or the one sold by Feb 22 Store as the 100W bidirectional board? Or was it another board design?

you can configure some boards with the IP5389 to work with 5S

Yes, the IP2368 is a buck boost but I can’t say for certain it will be able to boost to 20V when your alternator isn’t running. A fresh battery in your car should have a 12.7’ish voltage but you can still get around with one that’s on the way out at 12.1V. I know 12.1 is going to be trouble because of voltage drop but 12.7 might be as well depending on how long the line is to your board. I would use the SHORTEST length of wire possible, you want to be right up on the outlet and let your USB C cable do all the reaching for you since it puts out 20.2’ish so even a crappy 6 foot Emark cable will give you 19.6 at 5A You might have to rewire the outlet you choose to run the IP2368 off of with thicker wire. If you want me to I can test exactly how low you can to and still get the 20V PDO to stay active at full load.

I’ve been on the lookout for something that does 28V 5A. No luck. The moment I find one I’ll probably do a video the day I order it then it’ll be another 2-3 weeks until I get it from China. I expect PD 3.1 modules to be on the market in the next 3-6 months. Thankfully current is not increased as that seems to be the major hurdle these designs face. 5A is a lot of heat for a little IC. Let me test my 120W module really quick to see if PPS comes up. I want to say the only module with PPS is the expensive one that guy posted on Reddit but I have no doubt his module is a tank. He seems to be quite knowledgeable when it comes to PCB design and the components he chose.

I’m showing 7 PDO’s the last two being 3.30-21.00V 5A and 3.30-11.00V 5A Not sure if I trust the FNB48’s PD Trigger menu so do some more research to see if Feb 22 Store mentions what the PPS range is. I’ll send them a message too but that’s a crapshoot.

Interesting, thanks for checking. I don’t have a definite use for PPS but I’m imagining that it could be used as a variable lab supply if the you had the right trigger modules. Though I actually haven’t seen any of those on Ali. They’re all 5/9/12/20 fixed, so maybe the PPS is a moot point. Do you know why there’s two PDOs with overlapping ranges. I’m not that knowledgeable on PD specific yet.

@@michalzielinski7504 I haven't seen any boards for PPS either which is strange because I have some QC 3.0 trigger boards that can adjust the voltage up in 20 millivolt increments and 1 volt increments which makes it the perfect bench power supply. The PPS variant would be nearly identical in terms of function and I'm sure programming the board isn't that difficult if they've already programmed boards like the ZY12PDN to cycle through the voltages of regular PD. My guess on why there are two separate PDO's is if you don't provide >12 volts necessary for the buck boost converter to output 20V you are stuck with a buck only module. Perhaps the higher range PPS PDO is only able to be used when you have enough voltage for the boost part of the converter to work and the lower range PDO is for the times when the IC is limited to buck only mode

@@michalzielinski7504 This tiny device can be used as a variable bench power supply but I don't believe its compatible with PPS...they have a tester that is compatible with PPS but I think it lacks the VCC and GND pads that the Tiny PAT has....but honestly the board could be slightly changed to add some space for those pads...but I assume you would need a PC attached at all times to change your variable power supply. Still, its a (relatively) inexpensive board that could be adapted for a lot of uses if the owner of the PCB design is willing to share the gerber file or provide someone with a Tiny PAT with PPS www.crowdsupply.com/usbcee/tiny-pat

No problem gotta stop giving money to scummy sellers that put out shit boards

Ummmm....not bidirectional with wide voltage range....but has to be buck boost? Why not use a buck only if you're that high? SW3518S 1, 2, 4 channel

So I have the SW3516 which is the 60W version plus I think the 18S has some other protocols like Super VOOC. The 16's are amazing. Very sturdy platform. I'm not a huge fan of the buck only modules but lets face it, I'm mostly phone charging so I only need 9V Honestly I've been eyeing up the 4 channel SW3518S. Why get the XY-PDS100 for $10? $20? when the 4 channel (with a PCB mount DC barrel jack on it) for $18.75 shipped? Yeah no aluminum case but I never use the cases really unless traveling.

@@ericklein5097 Correct. My use case is that I have a 25.8V lifepo4 battery setup and would like to charge my laptops only. XY-PDS100 offers me the best bang for buck. I'll check the 4 channel SW3518S. Do you have a link?

@@bariss.3403 just an update, you can see in my most recent Aliexpress haul that I bought the 4 channel for $19. Works great. They are all over Ali if you search SW3518S 4 channel. Don't pay more than $20 USD

Kaiweets..can't remember the model name or number but its relatively new. Saw it on Amazon shortly after it came out and it was only $45. Really not too bad but I'd like to get something that measures inrush current.

Strange. Mine is perfectly happy to get power from RavPower 90W and 65W chargers. Also a no name 65W. I believe it also accepted my brothers M1 MacBook Charger but I would have to double check that. If you have anything below 100W try that. Should max out the charger and maybe push it a little beyond it’s max in my opinion. One limitation is going to be exactly what you are charging. A pack with a BMS might put up a fight if it doesn’t like the fast 6-7A charge rate. I would not recommend charging ANY 1P pack this way unless it is a Molicel P42A or P45B pack. No other cells can take that charge rate without severe risk of a fire. Your typical 2600mah cell that has a max charge rate of 1.3A would very likely go boom if you repeatedly charged it at 65-100W. The cells internal resistance might also be a factor. The charger is anticipating cells with an IR of

I'm having the same problem. it tries to draw more current than it can handle when charging and then it shuts down. I think the chipset got fried as the thing heats up fast. Discharging seems to be ok though

@@sircatsbythe3rd Thats very interesting. There might be a separate set of mosfets for the charging vs discharging and shorting them would lead to your charger shutting off from OCP and the board getting extremely hot. I bet the hot spot will be right on top of some mosfets.

Yes, a 4S2P pack is actually what I would recommend because the 120W module could pull in excess of 8 or 9 amps from your batteries. Splitting that up between two cells would mean you can use just about any 21700, not just a medium or high drain cell.

Been looking at ordering it the past couple days. One of the IC’s is buck boost and the other is buck only so you’d have to provide 20+ volts to get 200W. If you’re going to use a 4S pack to charge a MacBook and an iPhone it kinda makes sense. The buck boost IC can do the MacBook and since your phone only needs 18 or 27 watts it can run the buck only IC just fine on 4S voltage

The amp draw from the cells would be about 9.5 to 10.5 for that application

The one at the 10 minute mark is an absolute turd. Do not touch anything that has an SW2303/PL5501 IC combo. It heats up to 80 Celsius incredibly fast. I have not had any problems with the modules at 3:11 heathing up but I've only tried running them for a maybe 30 minutes or so at 18W. The other option would be to get the four channel version. The larger PCB offers more cooling space unless you are running all 4 channels at 18W

How about sw3516, is it cooler than the module at 3:11 when charged at 18w?

@@stephenshop4946 The SW3518S shouldn't even get warm at 18W. I can test it real quick with an input of 24V and 12V and let you know I've never had a problem with the USB A QC modules. I have 2 of them in a small project box and I used to run my Xtar chargers off solar. If all 4 slots are at 1A it draws 18W and I would have the second board charging a powerbank or something. They are secured down with hot glue and if they got hot the glue would have liquified and thats never happened. Can you post a link to exactly which module you've bought? There's about a dozen on the market. The single SW3518S are available for about $6 on Aliexpress. The SW3516 is much more common on Amazon, I'm not sure if the 18S is on Amazon. I would avoid the 16 simply because why pay the same amount of money for something that has a lower output and less fast charge protocols (18S has Super VOOC and something else IIRC). I think the 3518 without the S is a 65W but I'm not sure. I know the 16 is a 65W because I own some. www.aliexpress.us/item/3256802822997679.html?spm=a2g0o.detail.1000014.21.a25610749ZdS7F&gps-id=pcDetailBottomMoreOtherSeller&scm=1007.40000.326746.0&scm_id=1007.40000.326746.0&scm-url=1007.40000.326746.0&pvid=53dc09c1-f552-4b5e-8b7f-f131084baa9a&_t=gps-id:pcDetailBottomMoreOtherSeller,scm-url:1007.40000.326746.0,pvid:53dc09c1-f552-4b5e-8b7f-f131084baa9a,tpp_buckets:668%232846%238113%23613&pdp_npi=3%40dis%21USD%213.18%212.8%21%21%21%21%21%402103011616876742969407188eda1d%2112000023214459140%21rec%21US%213589144974 Here's a random one I grabbed. Price is ok...you could probably save a dollar, maybe two shopping around

@@stephenshop4946 Oh I see you said 3516...it should be cooler...same as the 18S I think unless there's different fets being used....let me test them

@@stephenshop4946 The new 5V 4A square shaped module is the tits. I'm testing it right now and it does not get hot. I ran one for 30 minutes now at 18.8W and all was good. Just can't use long thin wiring, it'll turn on and off really fast I'm going to post a video in a minute of the new module. Search 5v 4A USB powerbank module. TZT tand one other store sell it for about $3 plus shipping so about $4 each. I"m going to buy 5 more I love them so much.

Like superimpose them as seen on some of the pics Feb 22 store has where they use brass standoffs on four corners? On the old version of the IP2368 board with the purple PCB I think it is possible. If you look at my 120W module you’ll see there are 4 M2 or M3 holes for mounting or using standoffs an the old IP2368 should be the same. Keep in mind it’s usually just the two back that are connected to power, the front two on my 120W module appear to be isolated. Two should be no problem for the brass standoffs but more than that might be an issue of the brass not having the ability to run 400, 500 or 600 watts through them. Make sure you get the polarity right. When looking at my 120W from the front (looking straight into the female C port you plug into) the holes on the back for your brass standoffs are configured as: the left side is positive and the right side is negative. The positive is closest to a row of SMD capacitors

No, I actually never got around to putting heatsinks on any of my SW2303 boards and instead put heatsinks on my working boards. Skip the 2303 and just get an IP2368 or Feb 22 Stores 120W bidirectional fast charging board

@@ericklein5097 I've heard that board melts itself, I want to charge powerbank. Are the xy pds100 boards good??

@@Iinustechtips Sorry just now saw this. I've been wondering what the xy pds100's are SW3518S because its a buck only 100W and that lines up with the SW3518S. The SW3518 is not 100W. You have to get the S. Keep in mind its a buck only module so you have to provide a bit over 20 to get 20V output. The SW3516's have been really solid for me

@@ericklein5097 do you have a link to get it for cheap? ebay seller's want 10$+ for the 3516's

It’s a resistor, R7. I believe it is 18k for 5S. Sep 14 Store provides boards that have already been resistor swapped to your chosen series number and cell chemistry. Decent pricing too

Here’s a link (I uploaded this image to this hosting site) of the tables and where the resistors are ibb.co/4FTPL1b Where did you purchase your module from?

Here’s the other part of the info ibb.co/TBnVWKc

I thought this problem was only affecting the old boards but maybe I’m wrong (the purple PCB) There was a component near the back of the board only rated for something like 16V. It would blow every time for 5 and 6S. One of the reviews on Feb 14 or Sep 14 Store contains this info I believe. Edit you can also use Google translate to read the Chinese datasheet but the Aliexpress listing has all the necessary info for that specific PCB. Specs on the datasheet are for the IC only and may not work with certain components on the PCB

@@ericklein5097 Great thank you!!!

That's pretty impressive. Being slightly short of 100W is pretty common due to the resistance of the cables you use. I spend so much time looking for good deals on SHORT USB C cables so that I can properly test these devices. I no longer buy 6 foot cables unless it is for day to day usage where maximum output isn't a concern. You make me really want to buy one of those 200W modules now but I'm diverting every dime I have to A123 modules.

do you have a link? i cant find any modules only the ic's lol

@@Iinustechtips it’s on Feb 22 Store

@@ericklein5097 What about the USB C laptop cables rated specifically for 100 watts? I was hoping a laptop USB C to 5.5*2.5 cable would be able to carry 100 watts as rated, because they make most of them 6 feet or 2 meters. The length seems to be the biggest problem as mentioned, but I wonder if they're better than generic C cables since they're specifically rated for 100-watt charging.

@@beefstickswellington1203 so the key is that a cable capable of communicating that it can handle the 5A is needed and that requires an Emarker chip inside the USB C side of the cable. If there is no Emarker, then its a 3A max....or at least it can only allow the protocol to end 3A. You can force 5A through non Emarker cables but not with PD. Length only matters for resistance which affects your voltage drop. You want to stay with the shortest cable possible. Wire gauge also matters but most are 20 to 24AWG. 20AWG is preferred so if you see a cable advertising 20AWG it will have less drop than most. I don't think I've seen 18AWG. Aliexpress is loaded with C to DC cables that are 100W capable.

Mine never gets that hot. Is this during 100W-120W output? Normal inductor temp for me is around 70 maybe 80 Celsius. Let me know what your output wattage is and i will simulate. Mine is the same value with 2028AY below it. It’s kinda faded so I might have gotten one character wrong.

​@@ericklein5097 Hi my temp at 120w is 120 and at 60w temp is ower 90

@@bc-nr5tj Mine never went over 70 even at 120W for a couple minutes. Ran it at 100W for 10 minutes and stayed around 60. It was the IC's and the mosfets getting up into the 70's that worried me I think you might have a problem with your board.

​@@ericklein5097 i add 2 2.2uh coil in series and temperature was lower. I can't test at max 120w, coils were smaller. I ordered 6.8uh to replace my orginal. If they arrive I will leave ubdate.

@@bc-nr5tj you might want to check various aliexpress listings to see if any of them mention upgrades or different versions of the boards. I know the IP2368 has multiple versions and the IP2368 Pro has at least one revision. If you find any mentioning of the inductor being upgraded then it was likely an issue in the past. I think I ordered mine in the summer of last year. I'd say right around this time

Did I say 5318 in the video? Must be mixing it up with the 5389 when I say things fast. I have never tested any of them with small loads. I know quite a few off the shelf powerbanks come with that feature usually obtained by holding down the output button for 3 seconds. My PowerAdd powerbanks do it and I think most of the kit powerbanks sold by Feb 22 store have that feature (I don’t think the 45W one does but I believe the 20W one that many other stores sell might have that feature) I just jump straight to the max output since that’s where all the bullshit is. I definitely advise anyone looking at the 35XX boards to skip the 3516 and 3518 and just go for the 3518S since they are available as a single channel, dual channel and four channel unit on one PCB. The 4 channel is $19 shipped so less than $5 a channel. Powering it as if you were going to use all of its capabilities can be tricky and most people can get away with just using a 24V 10A power adapter but 400W (450W to be safe) is going to need 24V 17A+ power supply. You can go the cheapo route with a Chinese power supply or grab two 460W HP DPS server supplies off eBay for $20 and put them in series (requires minor mods)

​ @ericklein5097 Nah! My slip up, you said it right, just fixed my comment. 😇 The “holding down the output button for 3 seconds” advice is great, thanks, but I have some DIY power banks made with inexpensive modules just a few years ago which I honestly don't know if they are going to cooperate in that regard (I doubt 😒 it). The issue here is that manufacturers should be more aware of this requirement, because power banks should also serve to power up low power devices without problems. I remember that when using my USB soldering iron with a power bank the thing got really annoying due to the power bank shutting down every time the soldering iron went into standby, something which happened rather often due to its short time lapse before going into standby (around 24 seconds!).

@@SalvaBarkuti This is an overly complicated option but...if you were to also put a power resistor in the circuit that drew say 250mA....that would keep the powerbank awake without taking too much current. The easiest way I could imagine doing it is to find a cable that has multiple plugs like those 3 in 1 USB C, lightning, and micro cables. Chop off the one plug you'll never use (usually the micro but if you don't have apple then chop the lightning off) and attach a power resistor to that. I think a 16 or 20 ohm 5 or 10 watt resistor would do the trick. YOu probably want to slightly upsize your resistor (that's why I said 5 or 10 cuz you could potentially go smaller) just so that it doesn't get insanely hot just incase you touch it. 200 degrees celsius is serious burn territory....80 celsius you can get away without really burning yourself if you get your skin off of it quick

sorry to spam here but youtube always removes my comments from greatscott :( What i replied there was: My voltmeter/tester is right on the output of the C port so the cables are not interfering for me and mine is not shutting down around 90W but instead changing to 7V so that is why i was thinking that the issue is somewhere else. My step-down controller was giving a high pinch voice as i was reaching 90W (even though it could do 300W on paper). I was trying to hold it around 87W for a few minutes and there was no issue...but the temperature of the pcb was around 70-75 in a 20 C room so in the summer it is almost 100 if i would put it in a case. I was thinking to add a little cooler or something like that. But still my main issue is that i can't use it to charge from either of the ports. Do you have any suggestions about that? I can see that after connecting the charger then the pcb is realizing that it should charge and the percentage of the battery starts to rise very fast...but in reality, nothing is happening. I asked the seller about this but i already know it is a dead end. Thanks for the other info, I will check them. Update: I just checked and my USB tester was not directly connected to the PCB so could be...it is a 2$ essager 100W cable to it is possible that 10W went missing in 1M I don't have a spot welder yet and I didn't want to solder directly to the batteries so i Just used tapes for the moment. I just uploaded the video of my testing. Please don't notice the not so professional methods :)

@@cinkoxid you still can’t get charging output? Or input?

@@ericklein5097 output is okay but input is still not working. I tried every port but no luck. Maybe I will measure every resistor, etc on the pcb to find what could be wrong. The aliexpress seller after a few days of talking was finally able to understand that I sent him a url with a video of the issue so now I'm waiting for the support's reply.

@@ericklein5097 today aliexpress seller relied from the support. They are saying i should try a thicker wire and the charging will be good..... what a joke and i was waiting for a week for that answer. I thought they will tell me to change a resistor or something like that but no...the wire must be the problem😒😒

@@cinkoxid they have no clue what they are selling. Every single response I have gotten for a technical question has been "use thicker wire" "use quality cells" or some other nonsense