Jesus… what an in-depth observation/analysis & brilliant resolution… when I say “I WISH THIS MAN WAS MY UNCLE OR NEXT DOOR NEIGHBOR” you have no clue how serious I am…. I’m so frustrated trying to learn all of this DC power & lithium battery harvesting stuff and don’t know anyone who can help. I think it’s incredible that you have this knowledge. I’m subscribed!
Thanks so much for taking the time to explain this board. I recently discovered it, thanks to Ralph Bacon and Andreas Spiess. I've already discovered its "flaw" in reversing the battery. Oops!
I modified the 1850 shield with success. Very fine explanation. However, the mod is not for 3-thumbed technicians like me. Soldering the tiny MosFet was quite difficult and scraping the 4 traces requires great care to avoid scraping the ground plane as well. I overscraped and had various short circuits to ground when soldereing the new wires. Thank you kindly for your fantastic explanation!!!
Excellent Video. You took a very complex topic and presented it in a very logic and simple manner building on each subtopic. Thanks so much for your contribution.
AGREED! However, I just subbed and browsed his channel in excitement for more only to find that this was his LAST VIDEO 1 year ago and I don’t see any activity since… I wanted to find contact info to request topics & ask questions but I’m getting a funny feeling something may have happened to him. If you’re someone who subbed to him and “knew him” do you know if he’s ok?
I do like the methodical, clear way you show a potential solution to the problems in this device. I have one of these devices, and have shown my viewers it too, but as I've never charged it whilst using it I was unaware of the underlying issues. I just wonder how many other PSU-style designs have similar flaws? I'm thinking it would be quite straightforward to design a PCB for a standard 18650 holder plus all the mods you mention alongside the original circuitry, plus some battery reverse-polarity protection in the form of another P-channel MOSFET, perhaps. Then it would be quite a good design! I wonder how much that would cost? You got me thinking!
Ralph, I was thinking the exact same thing! I know virtually nothing about electronics and circuitry, and am only "fair" when it comes to EasyEDA, but if I had a circuit diagram, I'll bet I could come up with PCB.
Fantastic.! I was looking for this for some time. I was connecting this module to 6V panel.and was wondering why i always have to remove battery everytime once its fully discharged...!😊
Wow amazing investigation, thanks for sharing, those design flaws were driving me mad, I will try your fix. I have used this module for powering a Raspberry Pi and was quite perplexed because it worked fine for some time but then I connected some USB devices and after several hours it stopped working. As you comment that's because the charging IC was unable to deliver enough current and the battery started discharging until it was depleted, then it will never charge again because at this point it only provides 60mA and Raspberry Pi alone requires more for booting. Also used it for a webcam that draws less power and 6 months later it happened what you explain as other fault, always charging the battery destroyed it, the initial 2000mAh battery capacity was reduced considerably to the point it was no longer usable.
I have one of those. It charges the 18650 to 4.2 volts. I unhook the input power. I then hook a 3mm led to it with resistor to the pins on the side. The led will work for about 2 or 3 hours. The shield shuts power off at 4 volts. I kind of thought it would shut off at a much lower voltage. I think the stock version of this is useless. Or maybe I was shipped a defective one. Thanks for the video.
Wow! Excellent video, technical detail and information in general. I was wondering about these and that charge controller observation regarding the voltage when charging begins. I was thinking using a solar cell for the input, with added input circuit protection with low voltage/current shutoff and high voltage/current shutoff. Otherwise, I felt confident using to power my weather station. Now, after watching this, seems easiest just using a wall wart to power and perform your upgrades so is more a true UPS and that's about all due to the charge controller situation not letting the battery voltage drop to the typical low voltage of the Li-Ion battery of 3.0V or even like 3.2V and even stranger is sounds like doesn't even go below the nominal of 3.7V before charging. Wondering if that's to improve battery life or something or what? Also, wondering about somehow improving the charge controlling circuit so the input power supply is only used when the battery voltage drops below like 3.2V or whatever that lower voltage is set to? Anyways, thanks for sharing. Liked and Subscribed.
Great explanation. Do I need the mosfet? I prefer your first 2-diode solution. I only need 5V for the Vin of my Arduino, ESP32 or Teensy. Also, I don't need the switch mod as they will always be in use. Thanks for your informative video!!!
So, is there a safe, 18650/16340-based alternative that doesn't require any fiddling to power a small project, eg. TTGO controller and a 5V sensor? Converserly, ""the V3 lacks a load sharing circuit, so it may not terminate charging properly if the load current is too high" : What if I use a 1A charger? Would charging be safe without modifying the board?
Hi, very informative video. I have just bought the V9 version (4 battery slots) and it uses the TP5602 chip in place of the TP4056. Does this then mean that your modification does not apply to this new version?? Regards Don....
Demn i bought a dual one of these to hook up to a small solar panel and supply a wyze camera. Got right to the end when you said not for solar, sheet! Edit: could I just put a cap on the solar input?
I believe you have your Mosfet S and D the wrong way around. The reason why your circuit seems to work is because the Bulk Diode is conducting when the battery voltage is greater than the 5V supply voltage. In fact based upon this you could have just used two diodes instead of a diode and a Mosfet.
PLEASE advise!!! Why drain and source here is in reverse? It looks like in P channel Mosfet Source is usually a positive (+) and Drain is what goes further to circuit. I.e. when we replace a diode that goes from battery we should place a Mosfet source where anode of the diode was.
I did this modification a year ago, and I still think there is some problem in the circuit. I installed a new battery one year ago and I tested that it can power my RasberryPi for 11h hours with the wall charger disconnected. I left it there working for a year and on the last power outage it lasted 2 minutes, I measured the battery capacity and it was 50mAh (I think it was like 2000mAh when new). I had this shield working for 5 years and before the mod it had the same problem, it ruins batteries in few months. I think the problem may be that the battery slowly discharges and charges very often and this degrades the battery. I have an INA219 monitoring the input voltage and current to send alerts when there is a power outage, I will try to connect an other to the battery to verify my suspicions.
I bought this shield in a tech shop, and the said to me that I should charge the 18650 battery with 5v 1amp or less, is this wrong or right? Is there any problema if I actually use 5v 3amp?
I tried making this load sharing circuit and with 5v applied to Vin, I'm getting 4.5v on the battery connections. That doesn't sound right to me. Can you please help?
He explains that early in the video… this module has an integrated converter for 3.3v & 5v and he shows exactly where they are.. I guess you skipped that part.
Thank you very much. I made a similar circuit yesterday using diodes. But, the problem is, if the load needs more than 5v 1A, the boost converter is taking 4 amps. I am using XL6009. Don't know why this is happening
I have another problem with 2 and 4 battery shield. What about auto shut down? I have project which draws few milliamps, but shield to work needs something like 100mA. 1 battery shield, don't have that issue, and don't have that auto shut down, which my another project can work with monts. How I can over work that problem to get 5V output and draw minimum 3mA current from that pins, and don't have issue with auto of?
Why can’t we use the same SS14 Schottky diode instead of SB220, are there any differences in performance other than ampere rating. SS14 is Schottky Barrier Rectifier and SB220 is Schottky Barrier Plastic Rectifier. Does the term ‘plastic’ has any significant meaning?
The TP4056 is painful when it charges because it get my 3A 18650 to 4v toot sweet time then the remainder of the 0.2v it takes almost as long as it did for the first 4v. Insane. Do you know if there is a way to help this a bit as I just know they are being too overly safe. Probably would be fine on 1A or 2A from when this 4056 was originally designed, but it is really too long for 2h to go up 0.2v. It was so long I originally thought I received some bad board then I thought about it, and decided I would see if it ever turned on the blue diode (blue on mine not green). Sure enough 2-3h later I was about to go to bed and it turned blue. Tried a few more since with the same amount of time as I measured it. VERY slow after about 4.0x volts.
This is an extremely high quality video. Thank you for taking the time to do this so thoroughly.
TOTALLY AGREED. Only 100 subs!!!!
What a crying shame
Jesus… what an in-depth observation/analysis & brilliant resolution… when I say “I WISH THIS MAN WAS MY UNCLE OR NEXT DOOR NEIGHBOR” you have no clue how serious I am…. I’m so frustrated trying to learn all of this DC power & lithium battery harvesting stuff and don’t know anyone who can help. I think it’s incredible that you have this knowledge. I’m subscribed!
Thank you man! You saved me from scrapping my project.
Thanks so much for taking the time to explain this board. I recently discovered it, thanks to Ralph Bacon and Andreas Spiess. I've already discovered its "flaw" in reversing the battery. Oops!
I modified the 1850 shield with success. Very fine explanation. However, the mod is not for 3-thumbed technicians like me. Soldering the tiny MosFet was quite difficult and scraping the 4 traces requires great care to avoid scraping the ground plane as well. I overscraped and had various short circuits to ground when soldereing the new wires. Thank you kindly for your fantastic explanation!!!
I got the same board in 2024 and observed the same recharge cycle problem, then found this amazing video 😮
omg such a clear explanation from all the junk out there on the internet. thanks
This is a really useful video; very clear and practical. Definitely going to check out any other videos from this channel.
Excellent Video. You took a very complex topic and presented it in a very logic and simple manner building on each subtopic. Thanks so much for your contribution.
AGREED! However, I just subbed and browsed his channel in excitement for more only to find that this was his LAST VIDEO 1 year ago and I don’t see any activity since… I wanted to find contact info to request topics & ask questions but I’m getting a funny feeling something may have happened to him.
If you’re someone who subbed to him and “knew him” do you know if he’s ok?
WOW, OMG, Thank you for the time it took and the knowledge you shared getting all this done!
Excellent video and very well explained !
Thank you for this video. It was incredibly well explained.
Thanks for the shield explaning and... they have to use this accent to teach students speak english :) Clear as bottled water :) thx
I do like the methodical, clear way you show a potential solution to the problems in this device. I have one of these devices, and have shown my viewers it too, but as I've never charged it whilst using it I was unaware of the underlying issues. I just wonder how many other PSU-style designs have similar flaws?
I'm thinking it would be quite straightforward to design a PCB for a standard 18650 holder plus all the mods you mention alongside the original circuitry, plus some battery reverse-polarity protection in the form of another P-channel MOSFET, perhaps. Then it would be quite a good design! I wonder how much that would cost? You got me thinking!
Ralph, I was thinking the exact same thing! I know virtually nothing about electronics and circuitry, and am only "fair" when it comes to EasyEDA, but if I had a circuit diagram, I'll bet I could come up with PCB.
Yes, Reverse Polarity is a must.I fried one module just by putting the battery in a hurry 😢
Hi Ralph
Good video, very helpful to cover the short Cummings for China crap that there is a lot off.
Fantastic.!
I was looking for this for some time.
I was connecting this module to 6V panel.and was wondering why i always have to remove battery everytime once its fully discharged...!😊
Amazing! Thank you so much!
Wow amazing investigation, thanks for sharing, those design flaws were driving me mad, I will try your fix. I have used this module for powering a Raspberry Pi and was quite perplexed because it worked fine for some time but then I connected some USB devices and after several hours it stopped working. As you comment that's because the charging IC was unable to deliver enough current and the battery started discharging until it was depleted, then it will never charge again because at this point it only provides 60mA and Raspberry Pi alone requires more for booting.
Also used it for a webcam that draws less power and 6 months later it happened what you explain as other fault, always charging the battery destroyed it, the initial 2000mAh battery capacity was reduced considerably to the point it was no longer usable.
I have been looking at these boards for a while, so thank you for this video.
Will this be the same for 2, 3, and 4 battery holders of the same type?
I have one of those. It charges the 18650 to 4.2 volts. I unhook the input power. I then hook a 3mm led to it with resistor to the pins on the side. The led will work for about 2 or 3 hours. The shield shuts power off at 4 volts. I kind of thought it would shut off at a much lower voltage. I think the stock version of this is useless. Or maybe I was shipped a defective one. Thanks for the video.
Great video - thanks for sharing.
Wow! Excellent video, technical detail and information in general. I was wondering about these and that charge controller observation regarding the voltage when charging begins. I was thinking using a solar cell for the input, with added input circuit protection with low voltage/current shutoff and high voltage/current shutoff. Otherwise, I felt confident using to power my weather station.
Now, after watching this, seems easiest just using a wall wart to power and perform your upgrades so is more a true UPS and that's about all due to the charge controller situation not letting the battery voltage drop to the typical low voltage of the Li-Ion battery of 3.0V or even like 3.2V and even stranger is sounds like doesn't even go below the nominal of 3.7V before charging. Wondering if that's to improve battery life or something or what?
Also, wondering about somehow improving the charge controlling circuit so the input power supply is only used when the battery voltage drops below like 3.2V or whatever that lower voltage is set to?
Anyways, thanks for sharing. Liked and Subscribed.
Great explanation. Do I need the mosfet? I prefer your first 2-diode solution. I only need 5V for the Vin of my Arduino, ESP32 or Teensy. Also, I don't need the switch mod as they will always be in use. Thanks for your informative video!!!
So, is there a safe, 18650/16340-based alternative that doesn't require any fiddling to power a small project, eg. TTGO controller and a 5V sensor?
Converserly, ""the V3 lacks a load sharing circuit, so it may not terminate charging properly if the load current is too high" : What if I use a 1A charger? Would charging be safe without modifying the board?
Hi, very informative video. I have just bought the V9 version (4 battery slots) and it uses the TP5602 chip in place of the TP4056. Does this then mean that your modification does not apply to this new version?? Regards Don....
Demn i bought a dual one of these to hook up to a small solar panel and supply a wyze camera.
Got right to the end when you said not for solar, sheet!
Edit: could I just put a cap on the solar input?
I believe you have your Mosfet S and D the wrong way around. The reason why your circuit seems to work is because the Bulk Diode is conducting when the battery voltage is greater than the 5V supply voltage. In fact based upon this you could have just used two diodes instead of a diode and a Mosfet.
PLEASE advise!!! Why drain and source here is in reverse? It looks like in P channel Mosfet Source is usually a positive (+) and Drain is what goes further to circuit. I.e. when we replace a diode that goes from battery we should place a Mosfet source where anode of the diode was.
found an answer, it is actually so that 5V from micro usb would not pass to 4,2V battery
I did this modification a year ago, and I still think there is some problem in the circuit. I installed a new battery one year ago and I tested that it can power my RasberryPi for 11h hours with the wall charger disconnected. I left it there working for a year and on the last power outage it lasted 2 minutes, I measured the battery capacity and it was 50mAh (I think it was like 2000mAh when new). I had this shield working for 5 years and before the mod it had the same problem, it ruins batteries in few months. I think the problem may be that the battery slowly discharges and charges very often and this degrades the battery. I have an INA219 monitoring the input voltage and current to send alerts when there is a power outage, I will try to connect an other to the battery to verify my suspicions.
I wish I could purchase a unit that does all this. I'm not good doing all that fiddly work.
Sir, thank you so much!
Wonderful mod
I bought this shield in a tech shop, and the said to me that I should charge the 18650 battery with 5v 1amp or less, is this wrong or right? Is there any problema if I actually use 5v 3amp?
I tried making this load sharing circuit and with 5v applied to Vin, I'm getting 4.5v on the battery connections. That doesn't sound right to me. Can you please help?
How they are getting 5v output from a 3.7v battery? Are they using buck converter? If so won’t it drain the battery quickly?
He explains that early in the video… this module has an integrated converter for 3.3v & 5v and he shows exactly where they are.. I guess you skipped that part.
You need a boost converter to get 5v output from a 3.7v battery....not a buck converter....
Thank you very much. I made a similar circuit yesterday using diodes. But, the problem is, if the load needs more than 5v 1A, the boost converter is taking 4 amps. I am using XL6009. Don't know why this is happening
I have another problem with 2 and 4 battery shield. What about auto shut down? I have project which draws few milliamps, but shield to work needs something like 100mA. 1 battery shield, don't have that issue, and don't have that auto shut down, which my another project can work with monts. How I can over work that problem to get 5V output and draw minimum 3mA current from that pins, and don't have issue with auto of?
could you use it and in the same time charge it with solar power or other source
I think DMP1045U won't work. the circuit needs a depletion mode P-MOSFET.
good
can i use this module to powered my rc car ? my rc car runs 4.8v battery , and my plan to use this module with 18650 battery.....
Why can’t we use the same SS14 Schottky diode instead of SB220, are there any differences in performance other than ampere rating. SS14 is Schottky Barrier Rectifier and SB220 is Schottky Barrier Plastic Rectifier. Does the term ‘plastic’ has any significant meaning?
Schottky is Schottky - just have a look at the current rating. If that matches or your Diode can handle a bit more current, then you are fine.
Does anyone know if the V8 with two batteries have the same issue?
"Becareful". I just ordered one that says the same thing...
The TP4056 is painful when it charges because it get my 3A 18650 to 4v toot sweet time then the remainder of the 0.2v it takes almost as long as it did for the first 4v. Insane. Do you know if there is a way to help this a bit as I just know they are being too overly safe. Probably would be fine on 1A or 2A from when this 4056 was originally designed, but it is really too long for 2h to go up 0.2v. It was so long I originally thought I received some bad board then I thought about it, and decided I would see if it ever turned on the blue diode (blue on mine not green). Sure enough 2-3h later I was about to go to bed and it turned blue. Tried a few more since with the same amount of time as I measured it. VERY slow after about 4.0x volts.
How amper output
can i use this module to powered my rc car ? my rc car runs 4.8v battery , and my plan to use this module with 18650 battery.....