#383

Adding cheap and simple solar power to our small outdoor projects (e.g. ESP32, ESP8266, Arduino Pro Mini) removes the need to save energy or recharge batteries. Today we will enhance a simple 50 cents Li-Ion charger board to do exactly that.
I am a proud Patreon of @GreatScott!, @ElectroBOOM , @Electronoobs , @EEVblog , and others. No Docker, No Microsoft Teams, Zoom
Links:
TP4056 boards: s.click.aliexp... or amzn.to/2RqelKK or bit.ly/3eMmZw5
ESP32 WROOM : s.click.aliexp... or amzn.to/3rFVvBc
ESP32 PCB (different color) : s.click.aliexp...
Solar panel 6V : s.click.aliexp...
Solar Panels 5 - 5.5V : s.click.aliexp...
Voltage Regulator HT7333 : s.click.aliexp...
Schottky Diodes (used 1N5819) : s.click.aliexp...
FQP27P06 : s.click.aliexp...
Explanation of Power Path: • Load Sharing for the 1...
The links above usually are affiliate links that support the channel (no additional cost for you).
Supporting Material and Blog Page: www.sensorsiot.org
Github: www.github.com...
My Patreon Page: / andreasspiess
Discord: / discord
If you want to support the channel, please use the links below to start your shopping. No additional charges for you, but I get a commission of your purchases to buy new stuff for the channel
My Amazon.com shop: www.amazon.com...
For Banggood bit.ly/2jAQEf4
For AliExpress: bit.ly/2B0yTLL
For Amazon.de: amzn.to/2r0ZCYI
For Amazon UK: amzn.to/2mxBaJf
For ebay.com: ebay.to/2DuYXBp
www.facebook.c...
/ spiessa
www.instructab...
Please do not try to email me. This communication channel is reserved for my primary job
As an Amazon Associate, I earn from qualifying purchases
#no#midroll#ads

[@davidkempton2894]

Thanks for another interesting video. (BTW it's "Houston" 😊👍 )

[@AndreasSpiess]

Thanks for the correction! My mistake. I pinned your comment that everybody can see it...

[@largepimping]

@@AndreasSpiess It's OK, I just thought you were talking about John Huston 😊 (except he died in 1987)

[@kevinwalters5160]

More Houston confusion: kzbin.info/www/bejne/Y4Orop-CiN6pg9k

[@michaels3003]

@@kevinwalters5160 , well they never said exactly what he quoted, but it is a widely spread myth that is not going away. The sentence was "Houston, we've had a problem." This maybe was not the best way to say it, but this is what the Apollo 13 astronaut said after they were able to communicate with the mission control center again.

[@klausbrinck2137]

In "schwiizerdütsch" (Switzerland dialect of german) it was already correct as "Huston"... ;-)

[@user-yo4re5eb9y]

*I love my new solar **Generater.Systems** . It has the right amount of charging power for my needs and the light weight makes the portability idea.*

[@AndreasSpiess]

I like the solar panels more.

[@Davvg]

Sometimes what you cover is so in line with what I’ve been thinking about that it’s like you’re reading my mind.

[@AndreasSpiess]

:-) Glad it helps.

[@juncusbufonius]

Darn I have my charging boards, now I have to get FETs and some smd caps. You and your making things better. Excellent

[@AndreasSpiess]

Search for FETs with a low Vgs voltage.

[@jdsan6009]

Really good explanation. As a tip from experience: Lion batteries hate heat, and if they are exposed fully charged to high temperatures (say over 60°C) they can swell and leak over time. If your unit is assembled inside a small container, exposed to direct sunshine, it is not rare to reach those temperatures. To mitigate this, keep maximum charging at 80%. Also, if the charger/battery has NTC (temp sensor) it can stop charging over 45°C.

[@AndreasSpiess]

Good point. I measured temperature and, in summer it reached 65 degrees for a few hours. So far I had no casualties. But I expect the battery will not live too long. Charging up to only 80% is not easy with the standard chips. Because of the power path, the battery is not used nor charged during these high temperatures and is only "stored". Which seems to be ok.

[@jdsan6009]

@@AndreasSpiess you are correct. Is always good to be aware of batteries limitations and that is a huge topic. Keep with your great channel, which is personally one of the best in KZbin!

[@whatsupbudbud]

This is an interesting point since batteries don't like charging in negative temperatures as well due to dendrites. So how to best solve this very hot/very cold environment issue with solar charging these small boards?

[@cwa_nylis]

Fun fact: TP4056 has NTC pin, just in these tiny chargers it's not utilised, just pulled to ground.

[@MoritzvonSchweinitz]

Trivia about that Indiana Jones scene: the plan in the script was actually to have a normal "epic fight scene", but Harrison Ford was sick that day, so he recommended just shooting the guy.

[@AndreasSpiess]

Thanks for the additional info. I did not know he was ill. But I took this scene because I expected a long and difficult fighting scene. Like resoldering the small SMD component...

[@stephenkeebler732]

Ford was running a fever between 40-41c and could barely stand up...

[@hansdegroot652]

@@stephenkeebler732 Ha. I never knew that. Alway thought it a brilliant scene. Maybe it is even better now :-) And than you learn that after 43 years on a channel from a guy with a Swiss accent :-)

[@roostertechchan]

Wow! Thanks for the tip about the diode and p-fet. This is exactly what I need for my current solar project!

[@opensiembro7331]

Same.

[@AndreasSpiess]

I am glad it was helpful!

[@arunrajalegend]

@@AndreasSpiess Thanks, this was very helpful video, as we were trying to solve this issue for weeks. one problem i had is, after deep discharge my controller board is not wakes up even there is sun (for whole day). will this solution solves that issue without any external trigger to reset the board or with other components ?!

[@allsmart1441]

​@@arunrajalegendI have the same problem, have you had a chance to try it, can you share the results?

[@McTroyd]

Love that power path circuit! It seems so obvious in hindsight.

[@AndreasSpiess]

Indeed, it is an elegant solution. Not invented by me ;-)

[@TheHaykokalipsis]

The power path concept is very interesting

[@AndreasSpiess]

I agree!

[@TheChopain]

Thank you for a great video! In my projects i also add a temperature sensor and a mosfet to cut off the battery charging if the temperature is too low or too high (charging lithium batteries bellow 0C can causes permanent damage). Alternatively the tp4056 has a pin where one can connect a thermistor. Also if you use lifepo4 not only you don't need the regulator you also don't need the voltage divider to measure voltage. One can use ESP.getVcc() directly. And as a bonus no risk of blowing up 🤯.

[@noddlexxx9161]

wow do you have any guide I can view so I can implement what you did? I want to incorporate it that into my mini electronics project.

[@TheChopain]

@@noddlexxx9161 Sorry i got no guide but it is pretty simple. Whenever the ESP wakes from deepsleep (15 mins) I measure the temperature and depending on the result turn the mosfet between charging board and battery on or off. The project is now running something over 2 years without any manual charging needed.

[@noddlexxx9161]

@@TheChopain oh okay thank you. I'm panning to make like a little backup cam for my car and it would be pretty annoying to switch out the battery you know.

[@ZpeedTube]

​@@noddlexxx9161why not take the power directly from the car? Of course have to step voltage down first, but would not need to have a separate battery.

[@bubbaganoosh1387]

How did you keep the TP4056 off during deep sleep?

[@Davidmc23]

I really appreciate the thought and care you put into your work. Very approachable and always informative.

[@AndreasSpiess]

Thank you for your kind words!

[@SmithyScotland]

I was just waiting for the "we want more!"

[@AndreasSpiess]

You are the first mentioning the sentence!

[@dennis8196]

@@AndreasSpiess We all wait for BigClive to say 'one moment please' and Great Scott to say 'lets get started' everyone gets upset when it doesn't get said. Little details are important to us makers.

[@sebastienl2140]

i just waiting for a "we want less!" solution : lifepo4 (3.0V -> 3.5V supply for esp), mosfet switching the solar to the battery and regulation made in software! Sleep timer near 3.0V

[@dennis8196]

@@sebastienl2140 careful now, that line only comes once every 8-12 videos

[@j.a.seabrook2676]

@@dennis8196 Ha ha so true I watch those two gents also.

[@r5bc]

Sir your channel is one of the best channels on KZbin, providing educative and valuable content, please keep up the good work, YOU ARE DOING A FAVOR TO HUMANITY.

[@AndreasSpiess]

Thank you for your nice words!

[@erikpeterson1989]

I enjoyed this very much! I really appreciate your layered approach to the project, getting the core design together, then looking at further edge cases! This great both for storytelling and for design! Lieber Grüße aus momentan sonnig Seattle

[@AndreasSpiess]

Glad you liked the approach! Here the weather is very bad for the season :-(

[@__3093]

I've never enjoyed youtube videos as much as I enjoy yours

[@AndreasSpiess]

Thank you!

[@herrgerd1684]

Thanks for showing that P channel MOSFET hack. I knew sth like this was used in "zero power off" toggle circuits... But here it is an excellent addition, too!!!

[@AndreasSpiess]

Indeed, this is a very elegant solution.

[@SanthoshLakshmanan92]

Thank you very much @Andreas. I'm working on a wireless project powered by solar. I was stuck on making a powerpath with TP4056, having completed the prototype for the rest of the system. Thank you for the "we want more" section, and explaining the working. Cheers..

[@AndreasSpiess]

Glad it was helpful!

[@vineetwilson5277]

for a small project I did, using esp32, I also changed the R3 resistor (on the TP4056 board) to a more suitable 3.9K SMD resistor, which worked great for the purpose. Thanks for the great video by the way :D

[@AndreasSpiess]

You are right. You can add also a fixed resistor. And the older boards had bigger SMDs...

[@vineetwilson5277]

@@AndreasSpiess yaa, mine had the 0603 ones which are okay to replace if you have an okay eyesight :p P.S. - i also added the ldo and powerpath to my project, didn't know i had become professional too... XD

[@vivekchauhan7468]

I love the power path. Thanks

[@AndreasSpiess]

Me too ;-)

[@neokortexproductions3311]

Thank you sir! this is exactly what I was looking for, your a gentleman and a scholar🤝 I subscribed👍

[@AndreasSpiess]

Welcome aboard the channel!

[@zachreyhelmberger894]

FANTASTIC!!! Thank you for the R&D that went into this!

[@AndreasSpiess]

My pleasure!

[@auronedgevicks7739]

Alternatively, especially if you're more application engineer.. just use the 5v (or USB in) on the ESP32. You will need something like "PFM 600MA Control DC-DC Converter Step Up Boost Module" between the TP4056 battery out and your ESP32. This will boost the signal to a consistent 5V and you can connect that to your board through the USB in. I also highly recommend getting an 18650 with a protection circuit for additional protection. For battery monitor also consider a "I2C 3.7V Li Battery Fuel Gauge" which you can connect to the I2C on the ESP32

[@AndreasSpiess]

This is a good idea if you are. Ot concerned about power consumption.

[@catcam]

Thanks dear Andreas for great, useful and calming content.

[@AndreasSpiess]

You are welcome!

[@tonybell1597]

Thanks Andreas, I’ve been struggling using a tp4056 for a solar project, reckon this is the solution! Perfect…

[@AndreasSpiess]

I hope it will work for you as well.

[@jost459]

What was your problem? To small solar panel? To high charging current?

[@JohnAshleyMckenzie]

The voltage input for the analog input on the raw 8266 is 1v but on the Wemos Mini and other Node MCU boards is 3.3V. Love the video! I learn so much from you thank you!

[@AndreasSpiess]

Yes, you are right.

[@newburypi]

Thank you (yet again) for a clear and informative video. In the past I've started a number of solar projects, but always run into constraints that seem to require expensive solutions. With the information from your video, I will attempt to revive some of them. I may even try to implement them in surface mount technology. I'll let you know if I can fully incorporate your beautiful bodge onto a single board :)

[@AndreasSpiess]

The FET I used for sure was too big! You can use a DMP1045u instead.

[@user-fy8dz7cy7l]

I just think about what you do today. wow!

[@AndreasSpiess]

Good idea!

[@BlueyMcPhluey]

Great video. A lot of the concepts here around lipo batteries and charging were new to me so I had to go off and do a lot of research, but now I have some context I can appreciate the value you have provided

[@AndreasSpiess]

Welcome on board of the channel! Glad you like the content!

[@peterholliday8761]

Please note that if you are building this circuit and using a MOSFET as a switch to either turn the battery path on based on VG=0v, then this circuit may give unintended results with a solar cell. Be careful when choosing the MOSFET as VGS (th) are generally quite low which will cause the MOSFET to turn off when there is a small voltage at the gate, leaving the battery path off and the solar path on for a voltage that may be insufficient to power the ESP (

[@AndreasSpiess]

You are right. With little solar power the FET can be partially on (This is why a low Vgs type is needed). But then the body diode still conducts and we have the same situation as with a diode alone. In all other cases this circuit is better.

[@peterholliday8761]

Yes it is still a good solution. Just need to be careful of how the MOSFET is driven. A better option would be to drive the gate using some other logic and not rely on the solar power path. Great channel by the way.!

[@charlesmarseille123]

​@@peterholliday8761 I'm having unexpected results when the irradiance is small, indeed. the powered esp32 reboots because of brownouts since i only get about 2.6V to it in this situation. My code breaks during brownouts and I need a better gate control circuit. Would you please help?

[@JensHeuschkel]

Maybe we should design a board with your proposed modifications (and a 3.3V regulator) and manufacture it with a PCBA service. I guess many people here in the community would take one or two :)

[@AndreasSpiess]

Let's hope somebody will do it...

[@Giblet535]

Very nice tutorial. The only thing I do differently, I use a 3.3V buck-boost instead of a regulator. I have an ESP8266 that is more than a kilometer from my house, and it has been running m/l continuously for over a year. The battery doesn't do well below 0C... I recommend using the solar panel to shade the other parts from the sun, and mount the parts to an aluminum bar that extends outside the weatherproof enclosure. The bar will remove heat from the enclosure.

[@AndreasSpiess]

Good idea with the shading. I have a few IKEA boxes to try and compare different cooling concepts. I just have to do it...

[@rarenu]

Thank you for this very useful project. You made the already smart ESP32 to a more versatile one by powering it with your smart renewable energy power supply.

[@AndreasSpiess]

Glad it was helpful!

[@das250250]

Excellent summary of the design and well thought out use of common ,cheap parts to bolt on to any project. Well done Andreas

[@AndreasSpiess]

Thank you!

[@joanantonllarchpoyo2837]

The MCP73871 is a complete solar charger with all the bells and whistles needed in your project Andreas... Adafruit has a board with it.

[@Cptnbond]

This is an old board design, at a whopping price of $17.50 + shipping. Adafruit recommends the new board with BQ24074 - cheaper - but out of stock, so Andreas tips is the hackers way of doing it. Have a nice day. ;-)

[@joanantonllarchpoyo2837]

@@Cptnbond Yes....we can go down down to build it with simple transistors...nice learning too. :)

[@AndreasSpiess]

You are right. These chips seem to have all parts on board. Boards with this chips are not so widely available.

[@joanantonllarchpoyo2837]

I've seen at mouser that the MCP73871 cost € 1.59 for € 1.85 for the new BQ24074. The price difference for the two Adafruit boards is only explained for commercial reasons ... nothing to do with the board bom. I have been using the MCP73871 for years and it is possible to solder it by hand with a fine welding tip (the BQ24074 footprint looks similar).

[@grindel80]

is there a cheap chinese version of it?

[@rr974]

Thanks for your video. I was just doing my own research for solar powered ESPs recently and ended up with the TP4056 as the simplest solution as well. But ... there's no point in fiddling with the charging current of the TP4056. The current of solar cells is proportional to the solar radiation and thus is always varying between zero and the maximum current. Your current setting will never be right. But that's not as bad as it sounds. You just want the maximum available current and that's what you get without limiting the TP4056. Chasing the MPP only makes sense with a swtiching charge controller that can transfer energy between different voltage levels.

[@AndreasSpiess]

I agree, we never will find the right current without MPPT. The problem I had was that the panel hardly delivered current when the battery voltage was very low. Then, the TP4056 tried to draw max current and pulled the solar panel to a very bad place on its curve. This is what I wanted to avoid. If you have enough sun and your battery voltage is ok you will not encounter this issue.

[@rr974]

@@AndreasSpiess Thanks for your reply. Are you sure it was not just the Trickle Charge Mode of the TP4056 for a very low battery voltage? The current-voltage diagram can't explain what you observed. For a solar cell the current can only get higher for lower voltages. The power might decrease but the current is constantly increasing for lower voltages.

[@wasserdrucker6227]

That was also my thought. The higher power of the PV results in voltage, not current, but the higher voltage is burned anyway by the charge controller. Concidering the charge controller does not consume current, only burns the voltage above 4,2V, but we charge a battery that has 4V at the moment, you will get more power to the battery when operating the PV with 4 V, then operating it at its MPP at 5V: voltage current power power at battery with 4V 0 0.098 0 0.5 0.097 0.0485 1 0.096 0.096 1.5 0.095 0.1425 2 0.094 0.188 2.5 0.093 0.2325 3 0.092 0.276 3.5 0.091 0.3185 4 0.09 0.36 0.36 4.5 0.089 0.4005 0.356 5 0.085 0.425 0.34 5.5 0.07 0.385 0.28 6 0.03 0.18 0.12 6.5 0 0 0

[@tomasguozden5060]

@@wasserdrucker6227 so if max current is not achieved the module will ask more voltage from the panel, and the equilibrium point would be zero current?

[@tomasguozden5060]

I was wondering, if I had a higher voltage panel and a 5V buck converter, what voltage would the panel set in? I find two equilibrium points

[@zachreyhelmberger894]

WONDERFUL!! And the timing is perfect! We are just starting out making solar remote dataloggers for measuring the water level in our cisterns (we live in a 650 acre earthship community and get our drinking water from the rain and snow that lands on the roof).

[@AndreasSpiess]

Sounds like a perfect fit!

[@ktestable]

oh my, wow! this part 11:56 is the solution to my problem of diy ups. Thanks a lot

[@AndreasSpiess]

You are welcome!

[@vks002]

Yet another brilliant video from you! Thank you so much for such "useful AND interesting" one ! The clarity, detail and practicality in your work is not something that many others can match ! Thanks again. Can you please answer a question about the FET you used - the one you mentioned isn't available here, can you tell what should be the parameters of this FET so that I can choose another one.

[@AndreasSpiess]

You need a P-channel FET with low Vgs voltage (DMP1045u is also good)

[@malou8484]

Houston, Andreas solved the problem ! Thaank you.

[@AndreasSpiess]

:-)) Thank you!

[@TomLeg]

Lovely view of the hills, at the end!

[@AndreasSpiess]

Thank you! It was a nice day when I did my tests...

[@mahudson3547]

Good video. Damn, it arrived a week after I had used an identical circuit with 2 Schottky diodes and missed the pro solution. Still, mine looks nicer! Next time…

[@AndreasSpiess]

I can imagine that the two diodes look better! And the difference is probably not big for this application where we never want to deplete the battery.

[@klausbrinck2137]

6:40 "Never bring an smT-component to an smD-component-fight..."

[@AndreasSpiess]

:-))

[@antiquarian1773]

Love the energy calculation! helps out very much!

[@AndreasSpiess]

Glad it was helpful!

[@YusufBAYLAV]

Thanks Andreas, it is a great idea to convert my esp projects as I am living in Turkey, by the Aegean sea.

[@AndreasSpiess]

So for sure you only need a small solar panel ;-)

[@dorukbu3133]

Yet another great video from Andreas. Thanks! I wonder if it is possible to control TP4056 current setting from the ESP32 instead of using a manual pot. With a DAC maybe ?

[@AndreasSpiess]

You can get electronic potentiometers. But, as said, it is not worth the effort for such small panels.

[@madrian_hello]

Andreas what is the status of Super Power project?

[@AndreasSpiess]

They are testing the prototypes and found some errors. Firmware for the Pi board seems in a good shape.

[@madrian_hello]

@@AndreasSpiess very good news, thank you. 👍

[@grindel80]

@@AndreasSpiess Very nice!!

[@RobSummers993]

As always, another great video! Thank you! I will be using this tech in my garden this summer!

[@AndreasSpiess]

Glad it is helpful for your project!

[@joshportelli]

Great video, content, and delivery. Also love the little hand to point at things!

[@AndreasSpiess]

Thank you! Yes, the little hand seem to have a lot of fans ;-)

[@joey199412]

I would appreciate it if you would add in the amount it cost you at the time to build the total package. I know I can calculate it myself by opening all the aliexpress links and tallying them up but it won't reveal how much you personally paid for the package at the time of making the video.

[@andrewkieran8942]

Having watched Andreas' old mailbag videos, I expect that all of these parts came from his inventory so no new purchases were needed.

[@AndreasSpiess]

Andrew is right. I assume the total cost was around 10 dollars including the battery and the solar panel. Plus 1000 dollars work of a Swiss engineer ;-)

[@Dralsoft]

Does anyone have a Fritzing diagram for the p mosfet and diode components? I quite don't get it :(

[@AndreasSpiess]

I do not use Fritzing :-( Maybe somebody else...

[@squelchstuff]

Excellent Andreas. It seems this came just in time for many of us thinking about solar powered projects. Greetings from sunny-ish UK

[@AndreasSpiess]

Very good, that you have a sunny day! Here I had to wait for quite some time to film outdoors. It is mostly wet these days...

[@davidkempton2894]

@@AndreasSpiess It's mainly wet in the UK too during this month.

[@buzhug35]

Useful as always, dear Andreas! I have to create a system for ventilation of a greenhouse. And that will certainly help me! Thanks for your job! Best regards. Jean-Michel

[@AndreasSpiess]

Glad it was helpful!

[@SteveJones172pilot]

Glad I stumbled upon this video.. Definitely subscribed!! I'm more of a programmer with a little digital electronics experience - This analog stuff is the hard part for me, but this was so well explained, maybe now I can get all these half-built projects back in motion.

[@AndreasSpiess]

Welcome aboard! I am sure you will learn "analog stuff" if you are interested...

[@electrobobuddy6355]

Hands off your knowledge 🙏 sir. Keep delevering , it really helps us. Lots of Love form "India"❤️

[@AndreasSpiess]

Thank you for your nice words!

[@olir.8125]

Das war ein Super interessantes Video, vor allem die MOS-FET Lösung. Bei der Indiana Jones Szene musste ich laut lachen. Danke dafür, Andreas ☺️

[@AndreasSpiess]

Bitte, gern geschehen!

[@yuvarajvelumani7760]

I was looking for such a solution for my overhead water tank. Thank you so much.

[@AndreasSpiess]

Happy to help

[@mig2008pt]

Good video, I’m still waiting for the raspberry pi UPS project video! Something like done in this video, with power path, but connected to a 5v charger instead of a solar panel and all the components adjusted for the RBPi power consumption 👍

[@AndreasSpiess]

It is way more complicated if you want to have automatic shutdown etc. The project team got their prototypes and, during testing, found some errors...

[@mig2008pt]

@@AndreasSpiess yes, I think it is, but the team should try make it work without automatic shutdown, etc Just the basic functionality, to maintain the Pi working in the eventuality of a power outage, in my case, when it happens it’s not long lasting, just minutes. After make that working they can improve it with all the bells and whistles 👍 Keep the good work 😊

[@vu2dji]

Great video! Time to bring down my solar powered sr04t-esp8266 based water tank level monitor. I had simply put the tp4056 board directly without any mods and was wondering why it's not performing well. Now I know exactly what to do! 73 de VU2DJI

[@AndreasSpiess]

Good luck! 73 de HB9BLA

[@JEXUSWAT]

Thank you for the knowledge shared. you earned a new subscriber.

[@AndreasSpiess]

Welcome aboard the channel!

[@johnpipe362]

Thanks. You just beautifully covered another future project of mine. Time to get testing!

[@AndreasSpiess]

Enjoy!

[@meetpatel5142]

love the location you lived at

[@AndreasSpiess]

Thank you!

[@tochimoreno]

Excellent video, thank you. Simple and professional! It is what I need. BTW I also need turn-off the load at night. So I can use the solar panel voltage, or the lack of them, to turn off the load. I need the battery only to supply instant current on LoRa transmit or similar.

[@AndreasSpiess]

Good idea!

[@matheo4934]

I love it so much. Thanks

[@AndreasSpiess]

You're welcome 😊

[@Ed19601]

Like usual, very interesting and practical. I have a few considerations/questions: the powerpath is as you perfectly explain to prevent the battery charging and discharging at the same time, which could basically 'screw up' (my words) the proper charging of the battery. I wonder, if one has an ESP that is in deepsleep most of the time, whether in reality this is a significant problem, as most of the time the discharge would be say around 3uA. I am not sure but i seem to remember that some of the charge boards you tested earlier (#155), e.g. the CN3056 did not have this problem. These boards are also extremely cheap (61 cts in euro). Anyway, you probably know better than I as you have more solar projects than I. I only have one, with a bare ESP8266, fed through a 4056 with protection and no powerpath. The ESP wakes up every 3 hrs, maybe does 5 secs of connection and goes back to sleep. I use a 750mAh battery. Till now, even in the dark days that seems to work, at least i have not been faced with a battery that would remain below cut off. On a side note, as through some lucky finds, i acquired about 100 discarded 'vapes', with lipo cells varying from 500-1200mAh, i will definitely do more solar projects, and will consider the CN3056.

[@AndreasSpiess]

For deep-sleeping devices the power path is needed. And: A good idea to use these lipo batteries for a second life!

[@stasstas206]

Thank Andreas, your lessons are very interesting and kinde of relaxing ))

[@AndreasSpiess]

Glad to hear that! Thanks.

[@psykopat81]

"the finger" is very cute :)

[@AndreasSpiess]

:-)

[@PhG1961]

Very usefull indeed and nicely explained. Pure entertainment on Sunday mornings. What else....

[@AndreasSpiess]

Thank you! Currently, they do not produce a lot of energy. We have bad weather :-(

[@PhG1961]

@@AndreasSpiess Too bad of the weather. Here in Belgium it's reasonable... thanks to global warming and climate change the summers are getting warmer and sunnyer ;-). I'm thinking of buying a few of those solar cells. Also 2nd hand panels with more power are interesting these days.

[@epremeaux]

Make sure the battery has a low voltage cutoff circuit. The voltage divider is constantly drawing current. Even with a large panel, if you put your device in storage for a while, you stand a good chance of having an over-discharged dead battery. The voltage divider is a great battery destroyer. Ive seen lots of lipo devices (not solar charged, just USB charged) go "out of service" due to the voltage divider measurement.

[@AndreasSpiess]

You are right. This is why I use TP4056 boards with cutoff circuitry. You get them also without it.

[@beautifulsmall]

Great video, something I've wanted to do for ages. Very clear explanation.

[@AndreasSpiess]

Glad it was helpful!

[@avrphreak9121]

Excellent stuff! Always enjoying your videos, and learning along the way!

[@AndreasSpiess]

Much appreciated!

[@foxyredsheri]

AMAZING! I AM OBSESSED! THANK YOU! ❤

[@AndreasSpiess]

Glad you like it!

[@cemox007]

thank you, great video

[@AndreasSpiess]

Glad you liked it!

[@MrAbrandao]

Beautiful lesson!! I always learn a lot with your work. Sun is done. Now I look for a 200mA wind generator

[@AndreasSpiess]

Good luck! Fortunately, we do not have a lot of wind here. So I cannot help you.

[@sakisv4138]

Very interesting the opposite connection of the p-mosfet. It is the key to do what you say. The usual way connection simply doesn't work. It cannot conduct when solar voltage gets lower. It conducts only when zero solar voltage.The worst is that the battery will be overcharged by the body diode. Thanks for torturing our brains😂!!!

[@florinadrian4434]

This video arrived at the perfect time for my project. Thank you.

[@AndreasSpiess]

Glad I could help!

[@LarryKapp1]

I have over 6 KW of pv and 13kwh of battery here at my place. I like your simple setup and how you are using those tiny PV and battery - it sure takes up a lot less space than what I have !

[@AndreasSpiess]

But it would be too small to charge your Tesla ;-)

[@MartinPDanz]

Simple and efficient approach. Thanks. Maybe we should use this approach more often.

[@AndreasSpiess]

Absolutely! The film is old and still not fully understood ;-)

[@delinuxspecialist]

So I need to modify my solar powered weather stations and add a "PowerPath" to make them professional :) Thanks Andreas for this great video !

[@AndreasSpiess]

Absolutely! I find it a very elegant solution, especially also the usage of a FET instead of a diode. This concept is used in many other situations (like reverse polarization protection)

[@dd0356]

Power path idea seems interesting and useful. But i think, a fuse could be added where the source is so that in case of short-circuit (its outdoor, rains, high - temperature its 42 C here). and that p-Channel mosfet look giant at 12:06 and nice box

[@AndreasSpiess]

Of course you can use an SMD FET. This one was easy to solder... Concerning the fuse: You are right, a fuse always increases safety.

[@marcoberna78]

@@AndreasSpiess Could you suggest an easy to find smd FET?

[@davidclarkson8735]

@@marcoberna78 Consider an AO6401.

[@andresdans]

For the battery monitor section... when targeting for Deep sleep of the ESP32/8266 I do not like to use voltage dividers... the cannot be put in deep sleep... and they are always draining current (a little... but continuously)... I prefer to let ADC open and measure the internal voltage during the non sleep operation. The drawback is that you realise only when the Lipo starts to drop from 3.47V (3.3+0.17... I mean, from 4.2 to 3.47 you measure always 3.3V)... but as this scenario is normally in low consumption... you have plenty of (run)time (hours or days) to send alarms and go to your device to replace batteries or check what's going on.... as this is the path from from 3.47V to 2.8V of the Lipo...

[@AndreasSpiess]

Good idea! I did not know that you can measure internal voltage.

[@johnbruhling8018]

Very good 👍, I'm actually building a 12-24v wind generator with a 14 amp permanent mag DC motor and appropriate regulator and batt, the breezes here are not so gentle as in Europe! Anyways I saw another video once about inspecting the solar panels with 'near infrared' simply using a digital camera with the infrared filter taken off. They're just LED's and emit this low light when given voltage which also reveals damage.

[@johnbruhling8018]

Lol and I once heard that scene from Raiders wasn't reallt scripted, Harrison was a actually a little petulant and didn't feel like a fight scene so he just fired the blank

[@AndreasSpiess]

I do not think solar panels emit any light. But thermal cameras sees them because their temperature is higher than ambient. Concerning Ford: Another viewer also wrote that he was ill that day...

[@emilalmberg1096]

Wow, you're really thinking, smart!

[@AndreasSpiess]

Thank you!

[@TheUnofficialMaker]

great engineering, love it

[@AndreasSpiess]

Thank you!

[@gaborungvari784]

Thanks a lot Andreas for this very useful video! Useful and interesting, as always!!!

[@AndreasSpiess]

I am glad you liked it!

[@Scrogan]

I’d like to see your take on other energy harvesting methods, like rectennas. Ooh and that PFET circuit is wonderfully elegant. And 90mV dropout on that LDO is excellent!

[@AndreasSpiess]

I once did a video about harvester. It is only for very special applications and you have to use very low power chips...

[@pepethefrog7193]

Changing the charge current resistor is not necessary. Solarpanel will take care of this. Board cuts off the load to prevent overdischarge, no need for alternate current path. I do use this to charge an unprotected 18650 cell. Did not explode yet.

[@AndreasSpiess]

I al glad it works for you. In the special cases I mention you might need these additional parts.

[@mars667]

Excellent and timely, as always. Thanks!

[@AndreasSpiess]

You are welcome!

[@shermanpeabody7065]

I contributed the "Explanation of Power Path" video to my local open source hardware group. Thanks for linking to it. I don't have a solar panel to test with, but have run simulations on this idea, with results posted to my Github (I'm gbhug5a there). What I see is that when the solar panel starts generating current, the mosfet turns off, and all battery current is flowing through the mosfet body diode. If any current is flowing through the Schottky, the voltage on the gate must be higher than on the source, so the mosfet is turned off, and the circuit becomes a two-diode circuit. The only time the mosfet turns on is at night. I'm not sure that does much good. But you can install a second Schottky across the mosfet, in parallel with the body diode, and that greatly reduces the Vf of that path. If the simulations are right, I don't think the mosfet provides much benefit, but a simple two-diode circuit (both Schottky) might work pretty well.

[@AndreasSpiess]

Did you use a Logic Level p-channel MOSFET? Just because you write: "the voltage on the gate must be higher than on the source". P-channel MOSFETS need a negative Vgs voltage for turning on.

[@shermanpeabody7065]

@@AndreasSpiess, I guess I didn't say that very clearly. My point was that even a small current can flow from the panel to the load only if the Schottky is forward biased. But its anode is also the mosfet's gate, and its cathode is also the mosfet's source. So that means whenever any current is flowing through the Schottky, the gate must be higher than the source, and the mosfet must be off. Any current provided by the battery at that point must be flowing through the mosfet's body diode. And yes, I used a logic level mosfet. Anyway, I've just reorganized my Github repo to add another circuit which uses a single opamp to drive the gate. With that circuit the output voltage is never lower than the battery voltage. The opamp runs on 1uA, and common mode inputs extend above the upper rail, so no resistors are needed. So that circuit uses the mosfet, the opamp (DIP) and the Schottky diode. Probably a lot more complicated to work into the charger module, but if that 0.3V drop is critical, the opamp circuit will eliminate it. Thanks very much for your video. You are my main source for new things to work on.

[@steveswoodworking2504]

@@shermanpeabody7065 Hey, I found your design on github, and will use this for my solar powered project! Thanks so much!! I notice you were simulating this, did you or anyone ever build this?

[@GABC2007]

Thanks dear Andreas 🤗🤗

[@AndreasSpiess]

You are welcome!

[@MikeNugget]

I was waiting for this video! Thanks) Will be interesting to see solar equipment for RPi 4 😎

[@AndreasSpiess]

They are still testing...

[@jjanevski]

Really nice and informative video. Thanks for making this. 👍

[@AndreasSpiess]

Glad you liked it!

[@inlywang8157]

very useful and good solar project, talented man👍👍👍

[@AndreasSpiess]

Thank you!

[@Uncle-Duncan-Shack]

Hi Andreas, Comes at just the right time. I have just got some LORA radios to work on ESP32's, it was super exciting yesterday as it's my first time playing with RF. Now I can look at data harvesting using radio. I am testing soil moisture measurement using 950nM infra red by measuring the emissivity of the soil with a buried sensor, it works super well by the way. So now I can add the use of radio to that, and use solar if desired. Very happy with the nice stuff we have to do projects with. As always, I enjoyed the video and the content is useful. Kind regards, South Africa

[@AndreasSpiess]

Can you send me a link for your soil measurement? Sounds interesting!

[@Uncle-Duncan-Shack]

@@AndreasSpiess Hi Andreas, I would have to show you a picture of my sensor and a description of what I did. Do not have it published anywhere, will slip you a message on Twitter?

[@soekarmana]

hahahah i love your "indiana jones" way of soldering thing have done it quite often myself

[@AndreasSpiess]

:-)

[@sebydocky5080]

Thanks for this great video. For all my sun powered project, I am using the mppt charger/balancer PS-BC12111 also including a bay for 3 x 18650.... Works flawless.... and cheap.

[@AndreasSpiess]

This would probably be an overkill. And I am not sure how much of the energy would be used by the charger board itself... But for larger panels a more elaborated board like the PS-BC12111 is needed

[@Peeki]

Nice, do you know of any 2S boards?

[@BerndFelsche]

Nice practical circuit. If one can obtain local insolation data for the deployment location, the panel size and battery capacity can be matched better. Around here, data from the local agricultural research station revealed periods of up to 19 consecutive dull days where insolation over the whole day was less than a tenth of theoretical for the time of year. i.e. long, cloudy periods. Although that was a decadal event, shorter periods of 9 days or so are almost every year, in some cases with relatively clear skies for 3 to 5 days, followed by another 5 heavily overcast days. Note I am in sunny Western Australia! Given the known conditions and assuming that the application was vital, I would size the battery to power the device for 3 weeks without sun, and the panel to fully recharge the battery within 3 days of say 80% sun. If you're monitoring the battery state and panel voltage, then you can schedule a visit if the battery and panel voltages remain inexplicably low... Such as when avian deposits block bright sunlight from the panel.

[@AndreasSpiess]

You are right with your calculations. I did them in the video I mentioned and even gave some links where you find the needed info. I had both my projects out in winter and hardly got below 4 volts even here in Switzerland...

[@max9111]

I guess a solution to the hacking ( more rigorously wacky soldering😀) is to make a PCB. Naively, it does seem so difficult, but maybe there are problems, which I, as a beginner, cannot see. Very useful anyway.

[@AndreasSpiess]

Creating a PCB of course is a perfect thing. Unfortunately, it is time consuming...

[@hgghyxo]

sadly observing there are no more numbering for the videos, it was a great addition, and easy to look back for specific parts.

[@AndreasSpiess]

The about 10 last videos never had numbers. They will get it with age...

[@vaneay]

An overkill solution can be the PS-BC12111 its a MPPT controler plus 3S 18650 BMS integrated

[@vaneay]

store . sure-electronics . com / product : PS-BC12111

[@AndreasSpiess]

This for sure would be an overkill. And maybe it would also use a big part of the solar power of this small panel...

[@vaneay]

@@AndreasSpiess it is for bigger solar panels about 10w to 50w (15v to 24v)

[@ruudb72]

Thanks Andreas! Very good explained and very usefull.

[@AndreasSpiess]

Glad you liked it!