Hello Andreas, just discovered your channel. How great it is that you are giving all this theoretical background before trying to build something. This is what more people should do. Good work, subscribed.
I have almost certainly watched this video before but I didn't have any specific interest in solar power then. Now I do and I searched your channel for a video on exactly this topic. As is so often the case, your video gave me exactly the information I needed, avoided unnecessary and irrelevant detail and explained everything clearly. As we say in England, it was spot on. Thank you.
Hi Andreas, first thanks a lot for all the good videos! For a cheap MPP tracker I can recommend a simple TP 4056 Lipo charging board. My setup works as follows: 5V Solar Panel, the charging board (w/ cell protection) and a LiPo battery. As long as the light is very dim, the voltage from the panel is too low for the charging board and no current flows. When there’s more light, the PV voltage climbs over the battery voltage and the charging board starts to charge the battery. If the charging current is too high, the PV voltage drops and so does the current - the PV voltage can climb again. So the charge controller will always keep the PV voltage at the point where it can draw the maximum power from the panel. Two things have to be taken care of: - . - To keep the system from oscillating, I put a capacitor parallel to the panel- The maximum input voltage of the TP4056 is 8V. When the battery is fully charged, no more current is drawn from the panel and the PV voltage will rise to the open circuit voltage. This should not be higher than these 8V
This is my favorite video of yours to date. The fact that someone was kind enough to translate it to Portuguese also means I can scare my father (who thinks solar panels probably aren't much harder than connecting panel, battery and an LED lamp in parallel) all under the guise of a friendly 15min video. Not gonna lie I still don't get why MPPT is a $100 device instead of a little board that you buy in packs of 50.
A trivial typo at 2:15 minutes into the video. "×" used instead of "=". Your video has read "the question" in the minds of so many and answered excellently. Congratulations and thank you.
"Because this is boring we do some tests outside..."... Cool, wirklich nicht schlecht eingebaut. Der Beitrag war aber so oder so sehr belehrend, vor allem die Fakten über die Sonneneinstrahlung und wieviel Energie man nun effektiv abzapfen kann. Besten Dank.
Great video, thanks. I've been running an ESP-07 (all LED's removed) for more than a year now here on the Prairies in Canada where temperatures during the winter drop quite often below -35C. I used trial and error to figure out how long to put the ESP to sleep such that it wouldn't kill the battery during a worse case winter scenario. I will have to watch your video again and then apply these new learnings to see how much margin I was playing with. :) I am using a 1800mAh Li-on battery from a Sony PS3 Controller (rated at max charging current of 400mA and max charging voltage of 4.2 volts) and a 5V- 200mW 4.3cm X 3.4cm solar panel. When I tested the solar panel the maximum voltage I could get was 40mA and about 4.5V so I took a chance and connected it directly to the battery (so far so good and it's been more than a year!). I use only a diode to drop the voltage to the ESP and it has never gone above 3.4V. The ESP is on a 200s sleep cycle. As I said it has been working flawlessly for over a year but tomorrow I want to run this through your math. Thanks again for another wonderful and informative tutorial!
Thank you for your feedback. I think, physical laws are similar, even in the Prairies;-) Maybe this page helps for calculation of the power requirements of your ESP: battery-life.of-things.de/battery-life-calculator.php
Like your videos. They always start out with what you are trying to solve. That is very important. i often get lost in the weeds and forget about the original intent. I often don't need the projects you are working on BUT I always listen to learn. Great teacher!
Thank you for yet another interesting video. Some things you might want to look into, if you don't get the expected efficiency are: Angle of the panel (I noticed your panel laid flat on the table, which is not optimal), and temperature (cool panels are more efficient than hot panels).
I tried with various angles, but did not get big differences (maybe because it is summer and I did the measurements in the early afternoon. Then, I took the easy(flat) way. But you are absolutely right concerning the angle. If you have a close look at the diagrams in the video, you see the optimal angle for a whole year. Concerning temperature I cannot do a lot, I think, because the place is defined by the purpose. But, because the most critical time is anyway in December/January, the temperature maybe is not a big issue (in summer, I will have too much energy).
One of the best explanation in the whole youtube. Giving easy explanations to easy concept is an uncommon skill. This is real engineering: i'm sure the beer helps! Greetings from Italy
Andreas, Your videos, and this one in particular, is excellent. Pitched at just the right pace, and level for me anyway. I like the structure, and the order. These are some of the most informative, and useful, videos I have seen. Grateful thanks from SanFran.
I look forward to your controller video. Many of them have a USB connector. They claim up to 5V @ 2A, which is enough to power a Raspberry Pi 3. Of course there is no need to start a discussion on how big the batteries would need to be for that application, since we are talking about an ESP8266 & 18650 batteries. Possibly/probably a small single board controller would be sufficient. I'm now going to watch your video #64 again, about batteries appropriate for the ESP8266. Thanks for your videos.
I think, we also can power a Raspberry Pi. The same formulas would apply, just the panel and the battery would be bigger (about 3 times). And i do not know of a "sleep mode" for the Raspberry
Ja ist denn schon wieder Sonntag :-). Vielen Dank Andreas für die sehr gut strukturierte Vorgehensweise, sowie Deinen Berechnungen und Erklärungen. Freue mich schon auf die Fortsetzung... :-)
I'm doing something similar. I switched from the ESP8266 to the ESP32 so I can more easily monitor battery voltage and sensor voltages. One thing I noticed though was that the MPP charge controllers are expensive. So expensive, I just opted to buy more solar cells, which are comparatively cheap, and use a TP.4056 charge module (very cheap) It would have been a better idea for me to have done a proper economic analysis though. Looking forward to your analysis! Keep up the great work!
Here's a graph of the past few days of data logging to google docs. You'll notice gaps in data which is when my ESP32 has been hanging and causing my battery to drain completely. I hope I've solved the problem, but I've found the sample code for things like the DS18B20, Wifi connections, and so forth need some improving. Unless you have some good watch dog timers, you need to make sure your code is really bulletproof. docs.google.com/spreadsheets/d/1ni20axRWzek2f--lTF_YY7c1_GgZjzM9ziM6vrI6V38/pubchart?oid=108740660&format=interactive
Thank You for your time. I was also making some tests. Here we have more sun. 10W 5V panel + TP4506 + one 18650 + ESP8266 with deep sleep. Not that efficient but is working well so far. Looking forward for the next video.
Cool, great topic! Interesting info and very informative video! I have a solar-powered network node laying in my window sill for about a year now. It's a 6V 1W panel connected to a ESP-07 module via a 650 mAh li-ion battery with a charger/protection module. It wakes up for ten seconds every 15 minutes, day and night. It survived summer, but stopped in the winter because it didn't get enough sun. It's only for testing/learning now, but in the future I want to make a solar-powered node for my automated garden project, so plenty work left.
Thanks for the info. Maybe you make the calculations I did for your device and compare the results. Then, we would already have one confirmation (or not) of the theory...
Oh Andreas😀 A superb video, full of useful information and there is going to be a follow up too🙏 It's wonderful that you explain the math and importantly your thought process. I believe you managed to link together a number of things that I thought I understood. So very happy to see things more clearly. As always I am very appreciative of your time and efforts and just how concise you are. The project I am about to start is solar 2 axis for a 160 x 160 solar panel. So this is a great help. Bless you Andreas 👐
Thanks, Dean. Änd of course, we are interested in the results of your project, especially the efficiency (more energy because of good positioning vs energy consumption of motors)!
Andreas Spiess noted thank you. I will set up another INA219 to monitor the motor power requirements and as you say, work out the efficiency. From my research you only need to be within + or - 5 degrees. You will still be 99.6% efficient within this range. So had planned to only adjust at 6 degrees. I assume this is what you were reminding me to think about? Many thanks
There is certainly are lot of issues to consider about solar tech. Sadly more often than not the green desire and market hype clouds peoples judgement when investing in solar as the infamous "solar roadways" and Dave Jones have shown. One thing often not mentioned is how the solar panel heats up when exposed to the sun. The efficiency is affected by the panels temperature. If they are incorrectly mounted on a stone tile roof for example the stone tiles with high thermal mass storage capacity might also help heat up and prevent them from cooling. Another thing is the panels angle to the sun. Most often they get fix mounted. Solar tracking is expensive but might be how the supplier calculate their yield. Furthermore there are always losses when using converters. Storing the energy also mostly entails converting and self discharge losses. And then there is the matter of distributing the energy. Keeping them clean from snow and dust and wet leaf is also something that is often ignored. Thank you Andreas for your excellent video that 'shines light' on some of the issues with solar tech.
I think, there are two distinct solar applications: One is to use solar to get maximum energy out. This is where you find a lot of information. And your issue is also located in this application (btw: look for a comment in this video where a viewer reports, that he looses 20% in hot situations) The other application is the one I try to deal with: Get enough energy during the whole year. Here, We have to solve other problems like efficiency in low power conditions, deal with low voltage panels etc. Here, I did not find a lot of information. And heat is not so important, because we anyway will have too much energy in summer.
Simply excellent. By the way, I wonder about the reason why there are people who have voted 'I do not like'And I live in Spain,... you can imagine that we have sun, sun and sun
Fantastic! It is as if you could see that I was experimenting with a solar panel and Li-On battery today & yesterday way on the other side of the planet. Plus I was trying to calculate panel and cell sizing too. Perfect timing and great spot-on content. Thanks. AND tomorrow is the Summer Solstice - optimal conditions for experimenting. I'm addicted to your videos.
Hi Andreas, very interesting topic as always. Your main concern should be the min max temperatures which the device (including charging and battery modules) will "suffer". Chemical batteries most likely loose as much as 70% capacity when cold until they refuse to charge at all. Also due to reduced molecule movement the voltage is reduced when the battery is connected to small loads (few mA). also batteries can be permanently damaged when they get too hot (fast charging and heat from the sun). Think about the Rosetta Research Lander which suffered those problems on the comet it landed on, somewhat extreme example but you will experience similar problems. Lead gel batteries also often used for off the grid solar systems are best but very bulky... when reducing the max load and total power consumption a powercap can be used.
Thank you for your input. You are right, batteries (and temperature) are an additional issue we have to consider. Another possibility could be to use a big solar panel and super caps. We will see...
Composite systems are quite good at this implementation; There are two power sources, one is solar and one is backup power from the mains electricity. It works just like you expect when conditions are alright; charge the battery in daytime and use it in nighttime. But when solar panel voltage drops too much and battery voltage is below threshold, the device turns to backup power so the device doesn't die.
To have grid power is always a good thing and reduces the size/cost of a project considerably. But for me, I want a completely independent device. So, unfortunately, I have to go the hard way (or change where I live...)
Andreas, I really enjoy your videos, you obviously put a lot of thought into their structure and order. This makes for some outstandingly informative stuff!!! Many thanks from this very grateful Welshman.
Great video. Really simple common sense calculations. I think anyone dabbling should be able to figure out this themselves but if not you would have helped. What I like best about your channel is the clear straight engineering without any BS. No lies or faked stuff. Too much of the internet and social media is being invaded by persons peddling fake information. Please do keep up the good work. I will keep watching.
"In the middle I had to stop to drink a beer". This made my day! I really enjoy your videos/tutorials. They are always very informative and full of very useful knowledge as well as great results. Thanks for sharing!
Thank you. I have been looking at videos with similar content for the past few days and I yours has been the most helpful. I hope you continue to provide more. Wish I could give you two thumbs up
Great tutorial. This open up so many applications. Im more I treated in the charging units. If there is bigger, smaller and also if you can order them online
I would like to suggest that you spend considerable time on the solar panel and the sun's path in Basel. On June 21 at solar noon here in Phoenix it is ~82 above the horizon; yours is ~50'. I noticed that your test panels all seem to have a reflective surfaces. At 50' the sun at solar noon is quite low compared to Phoenix. The most efficient solar angle to the plane is 90' so with your cell lying flat in your picture you are off by 40'. With the reflective surface most of the light is going to bounce right off just like a mirror. So now the question is does tilting the plane make any sense or a tracking panel even better. Here in Phoenix the panels around town all are pointed north and at an angle off perpendicular determined by some formula for year long maximum efficiency. I am old enough to remember seeing the first solar cell back in '56 made by Bell Labs; tiny little thing. Hope this will help you in your decision making. Out here we have been experimenting with the sun for a very long time. APS has a solar farm experimental station at Mcclintock and University in Tempe just down the street from ASU. Try a Google Earth look see. Heat can be a problem for the panel so let it breath. 106' and the sun has set.
I tried to tilt it to the maximum, but the difference was not big. So, I decided to do it flat to be easier comparable between all panels. After your post, I looked into solarelectricityhandbook.com/solar-irradiance.html and compared the two values for Basel/June. Flat, the factor is 5.7 and at 57 degrees (best for summer) it was 5.43. So, also here, the difference does not seem very big. In this application, It would be advisable to chose the angle for the worst month of the year as we will have too much energy in summer. I hope, that the surface of solar panels is penetrable for sunlight and does not act as a mirror, but of course, I cannot test it. BTW: I loved Phoenix and its surroundings when I was there. I flew also to the Sedona airport (with a Cessna) and will never forget that...
mini solar produce constant power you can use( buck converter) or( boost converter) first then you storage power to battery . the milliamps and voltage change all the time when you produce power and you can change it at any ratio you like , to make it more efficient.
That's an interesting project! One thing though, since these small solar powered devices will stay outside for some years ideally: Did you account for the degradation of the 18650 when fully charged in summer? The cell will degrade much faster then it has to if the temperatures are up and it's fully charged, even though it doesn't need the additional capacity in these conditions. You could even reduce the charge limit dynamically based on the mean temperature to get maximum lifespan. Thanks for your informative videos!
You mention another complication to the project. It will depend on the way I will implement MPPT. If I use a micro controller, I can control all these parameters. With normal MPPT controllers, this is not possible (I assume).
I like the pace and thoroughness of your video. I'm currently on 11:29 and trying to figure out with no results where you obtained the "Solar irradiance figures" presented on the left-hand side of the screen. Thank you.
Thank you! Very interesting. I've been thinking of doing some solar energy experiments myself, sadly northern Sweden is lacking in sun hours most of the year. :)
I have ordered some TP4056 from AliExpress. These connects great to the 18650 battery and deliver the right start and stop for charging. Additionally these are incredible cheap. Looking forward to get them delivered. If they support the MPP I do not know. I assume not.
Marketing spam. Social marketing bots are spamming TH-cam videos from people like you to make money from subscriptions. Middle-women/men nobody needs. I prefer your Swiss attitude, because it is ethical.
Very good. Very very good. I live in Brasil, Mato Grosso state, very warm and have very sun, and was thinking about and esp8266 for solar weather station, and this video will help me very much, thank you!
Looking forward very much to the next video especially cutting off charging the battery at 4.2v and at the low voltage point. This should solve issues I've been having using the piZero and the sparkfun sunny buddy MPPT controller.
I think too, that it could be an interesting part. But I also think (or fear), that we have to keep the number of parts as low as possible because they will use energy all the time, also when there is no sun...
Nice vid and a fresh view on the sizing question. But I think you for got one big problem. How to survive freezing conditions with lithium type batteries. For me the fun part of solar is to run something standalone where there is no power like a shed or maybe a bird house or something. Been looking at many solar projects but almost all ignore freezing conditions. The only solution so far (for small devices) is to protect the battery by no charging it. But what is it freezes for 3 or 4 weeks. than the light or sensors will stop working for sure or the wifi on the eps8266 has drained the single 18650. Heating pads require loads of power so back to lead acid than? It all feels very disappointing. How was the beer? :)
I have no long experience. But most documents I read about Li-Ion batteries in the cold are if you charge them with high currents. That is not what we do. And I had viewers reporting that their batteries survived winter. So maybe it is time for a test in your situation? The loss would not be too big, I assume. And you have a fair chance to succeed.
Andreas very useful video. I have been trying to do a similar thing for my esp8266 could not decide on a charge controller. Looking forward to the next episode.
Sunny Buddy - MPPT Solar Charger from sparkfun and Adafruit USB / DC / Solar Lithium Ion/Polymer charger - v2 both quite expensive and then there was also this one thepihut.com/products/solar-lipo-charger-3-7v?ref=isp_rel_prd&isp_ref_pos=4
Very good video for the right reasons. I have a question: which would be the algorythm to calculate needed panel for chargins a 18650 battery?? Thanks in advance
Hi Andreas, good information, thank you. I am running an arduino based weather station with solar power in Zurich. I made good experience with a charger module. This is the one: 03962a / TP4056. but there are two versions, one with dedicated battery connectors and one with shared ones. I prefer the newer version with dedicated connectors. Cheers SJ
Nice explanation. But an additional factor that has to be concidered depending on location ist the dust! E.g. in Dubai the solar panel would get dirty/dusty after some time and thus convert less energy. This can happen without a dusty desert ... some air polution + dirty rain or snow can destroy the calculation also.
Awesome video. I would only start from the other end. You know how many days/hours you need to run the device from a battery, then calculate the size of solar panel. If you need a week of battery life and power consumption of your device, you have a size of a battery. Then you can calculate size of solar panel to charge the battery within a day or two
@@AndreasSpiess thanks for the reply. Would you consider using solar to power up motors? I would love to do rover-like robot. Or powering such device from solar is not feasible?
Nice work. However, i was wondering, wouldnt it be possible to instead of relying on solar panel power in the worst month (The way you do calculations now), incorporate a larger battery to compensate for the winter months? i.e. Store surplus power generated in the summer and use it during the winter? Much like you would for nights and bad weather periods? Or would be battery become too large? Well, if you need 14 batteries to surive 14 days, and you could reduce your power consumption to 10%, then the 14 battery setup should survive 140 days which should be plenty to carry you through the winter. Of course this will increase weight, but the sun-exposed-footprint of the device will be smaller which might make it more convenient. There is probably some room for cost-optimalisation though.
I think, there are many scenarios possible. The formulas will be the same. Maybe it is even possible to "simulate" different scenarios, compare the results, and chose the right one for your power need and location.
I bought a small solar panel from AliExpress sold as "5V 400mA". I measured the open circuit voltage at 6.5V and short circuit current at close to 150mA. I determined the MPPT of the solar panel to be about 300mW at 5.8V and 50mA (measured while adjusting a 10k rheostat in series with the panel). Does this mean I should replaced the resistor on the TP4056 to only draw 50mA?
Optimal power point tracking may also be relevant to mechanical generators like wind generators and bicycle dynamo. That being said, is wind a viable way to support a wireless sensor? If it's not viable on its own, it may provide just the right supplementary power just for the bad weather days, if your climate happens to make rainy days also windy.
Excellent video. Perhaps we need to consider other ways of generating power, such as small wind generators? I haven't seen anything suitable yet, but it must now be possible to make anemometer-style generators that can provide a small but usable charging current to supplement solar panels, or even replace them in areas where wind is more common than sun?
For sure there are other possibilities, but the sun is very handy and everywhere available. Wind, for example, is not available where I live (at least not in quantities).
Well, your calculations may have left enough margin for error there. But I feel like it's worth mentioning, that if your battery is large to get you through the night (not the winter), you should design things to work in the worst case scenario and not on average across the whole year. In the summer you will be getting more power than the average and in the winter less. Therefore if your solar panel produces enough power on average over the year, that doesn't mean it will produce enough power in the winter. Especially not on a unusually cloudy day after your solar panel has collected some dust etc. Personally I would oversize the solar panel just a little bit more to be safe.
+Andreas, thanks for this great video! One question: What is the 10% you mention at 11.25? Is that solar panel efficiency? Because earlier you were using 15% as the efficiency multiplier. Also, at ~3 minutes in the video you're making calculations based on the year-around solar radiation, but you take into account 15% efficiency for the solar panel, and then another 33% for the charging circuit. Then at end you make the same calculations again using the data for December instead of year-round, however here you only take 10% efficiency into account, so why 10 instead of 15, and what happened to the 33% of the charger circuit, I think we should still take it into account, right?
The 10% take also the losses of battery and charging device into account. The solar cell itself has around 15% and you lose some energy if you charge and discharge a battery
Thanks, I realize now I misunderstood the formula (possibly related to the fact that I was watching it at 2AM :)) I thought you only deducted 10% for the efficiency, but in fact you're deducting 90%
Thanks Andreas you saves my times and moneys . Very helpful and experimental video Great job.I hope God gives you a more healthy and long life.Best Regards
Thanks for sharing I learned a lot from this and your videos. I am building a weather station using ESP8266 and this is spot for my power requirements. I live in Malta so solar irradiance should be similar to that of Dubai. Many thanks.
I'm just starting to look into this, so I don't have your hands-on experience, but I was thinking about how one can monitor solar intensity with respect to solar panels. A Lux meter is not too bad, because most panels use the 380-750nm range, but it's not a full correlation with Lumens. The cheaper Lumens meters use two photo-diodes, and calculate Lumens based on their outputs. But their output is not linear, although it does roughly correspond the the human eye. A solar panel has a much more linear response to wavelength, in the visible region. Maybe you could just use a standard solar panel of the same type as you use in your tests, and record the power generated from it during each test. Then that would automatically take into account the effects of atmospheric moisture, dust etc on solar radiation intensity. You could even use that to normalize your test data by using the standard power output for a given series of tests: just record it as you do each test. Then if something happens during a specific test you will both know it and be able to compensate for it.
I think, I will only try to monitor the energy which really came from the panel, because I am not interested in the phases with energy in ambundance, only for the critical phases. This is different to most applications where you want to get must out of your panels.
That's not what I mean't. You tried to use a Lux reading to compare readings taken at various times for your system, and my point is that Lux is not an ideal measurement for solar power: the wavelengths and weightings over these wavelength ranges are different for Lux and solar radiation. I understand that you want to concentrate on the low energy times: Winter basically. So you need some way of making a quantitative measurement of how well a specific system behaves for a given amount of solar energy. So you need to be able to measure solar energy.
Very good analysis. I tried using 3W Poly crystalline panel in India to power the NODEMCU without a battery (using a voltage regulator) but wasn't successful. Can't wait for your next video to see your Battery size recommendation..thank you
Andreas Spiess yes, got it thanks..will watch out for next video on how to charge..will it be a charge controller or charging module for LiPo..will wait and see
Interesting video and very nice explanation. For the next episode for the mppt, pwm, etc chargers please take into account that mppt chargers usually draw much more current than the other ones and that it might not be as good as a pwm one especially for low power situations.
Most of these chargers are not for this scenario They use quite high voltages and are rated for many amperes. I did not find many of them suited for low power applications so far.
So so true. Seems we need to make it ourselves or modify . I have a couple of these but with the display, which can be turned off.. www.ebay.co.uk/itm/1PCS-5A-MPPT-Solar-Panel-Controller-Voltage-Step-down-Module-Constant-Current-/292153107129 You need to make sure to use a Schottky diode (place for that on the PCB) and most of these mppt boards since when you have no solar the power from the batt goes back into the board. But this makes it a bit more difficult since the tolerance of all components might make the batt charge over 4.2 Volt. So if I would use these kind of things to make a L-Ion charger I would prefer 2 instead of 1 cells but to make it charge only to 4 Volts which is better for Li-Ion cells anyway. Already looking fwd to next Sunday.
I love Your videos. You are encountering and describing with easy words (and tests!) many of problems with building IoT devices. Not only simple "Hello World" projects. :)
Great video. Result ( with deep sleep for 10 seconds ) is panel 10x10cm and one 18650 battery. The information about requested capacity of battery is missing. 3000mAh?
I think one thing not taken into consideration is low temperatures affecting battery power and life. That can be a difficult obstacle to overcome in higher latitudes. A housing that can absorb some heat from the sun will help, but obviously not when there is no sun. Discharging can also be an issue at lower temperatures, even if it is less of an issue. However, keeping a device running for long periods of time when batteries are cold will be an issue. It looks like Basel doesn't experience temperatures much lower than zero in December and January, but where temperatures are near to or below zero charging batteries can be a problem. Some Lithium batteries will simply not charge below freezing, and will have issues charging below 5 degrees, only allowing charging at very low currents. I wonder if you have been forced to take this into consideration for your projects?
Hi Andreas. I wonder if Amorphous silicon cells could offer any advantages during cloudy winter months. They have a much lower overall efficiency, but better low-light efficiency.
...and i do have another hint. solarcells r working way more efficiently if they are directly pointed to the sun. this means 254,5° movement horizontal, and vertical can be fixed somewhere at the middle between 0 and 65,9° azimuth (also based on basel). 22° az or little more will be optimized for winterseason. it wouldnt be that hard to make the cell moveable, and combined with a solartracker it would be a very nice and interesting topic for your channel :-) i was up to build such, but... yah. didnt.
I did a solar tracker with 4 photo resistors and a dc motor for a school project. it is really simple to make. I bet it would be even easier with a servo motor I could upload it if you don't mind that is all written in spanish.
Patchoulino Don't think so. Servos usually rotate 180deg i just read. As calculatet more as 250deg is needet. And exact positioning isn't needed. Follow up, and endswitches should be good enough.
I am sure that this is great for high power applications. But it would be interesting to fin out for small panels, if the energy used for the servos can be made up with the additional energy you get...
Great thinking. The thrill of building variable vector control would justify doing this. However, the practical consideration of the added equipment (and load) to extract nominal gains would perhaps be defeated by what would more easily be achieved by adding one more cell. In a well balanced system, much of the power is not used anyway to protect overcharge, right? But, yes, someday I'd like to try the servo control anyway. If a 180 servo is centered on the 254 horizontal sweep, only 37 degrees would be missing on each side. Better yet, orient the sweep to the most critical conditions knowing that maximum voltages will be clipped by the battery regulator anyway.
Another very useful video, Andreas! I assume it gets below freezing many days in the Swiss winter, true? I think that Lithium batteries should not be charged below freezing, true? What type of battery would be good to use for a solar project that will be below freezing (say down to -10C) some days?
Hi thanks for the video. Yes designing a 24/24/365 solar powered system is quite different than to design a device to maximize solar harvesting during sunny days. Once you design it to gather enough power to pass bad winter days, you will discover that the panels will deliver > 500 times too much energy during hot sunny days. So one thing to consider is what to do with that energy? My current approach is to use that energy to power fans, that will cool down the whole system.
Ups! Of course >500 was exaggerated. I meant factor ~50. I have tested my 5W panel to deliver 15-20mA on rainy days compared to 800mA maximum in bright sun. Anyhow these little 20mA are reducing the energy deficit by ~25%. I will activate the fans by the signal "fully charged" of my charger modules, essentially to avoid a source voltage increase above the specs of the regulator when it stops dragging current. Additionally the Arduino will be able to activate them temporarily as well, when its own temperature exceeds a threshold, that will improve the panel efficiency.
I didn't close my eyes one single time. Great video. About the battery size for 14 days of bad weather. Wouldn't this assume that there is absolutely no sun for 14 days? Even if the weather is bad and it is pretty cloudy, the solar panels should convert at least some energy. Or do panels not work at all when it's really cloudy?
Hi, Andreas, very interesting tutorial (as usual). I've got a question: To calculate the watt per m2 available, is it realy correct to divide the annual solar radiation by the number hours in one year (8760) ? Since there is no radiation at night, wouldn't it be exact to divide this annual solar radiation by the number of hours of sunshine each year ?
One important variable that should be considered also is the environment temperature. I have a 500W solar panel that reduces ~20% when we have hot weather. So, with clear direct face of the sun (and perfect inclination) i have almost have full power in the winter (498W was my best measure) and with around 30ºC or more, i have around 400W
+Nuno Justo Thanks for your info. This is more important for the usual setup. For this setup it is even an advantage, because we need every Watt in winter and we have to much in summer.
yeah, definately. Soon i will do my first project with a 100x60 cm panel. This will power up my new toy omega 2 plus. you should take a look at it. onion.io/omega2/
A Review of a full setup would be nice. E.g. OpenHAB2, MQTT, ESP32, Mongoose OS, powermanagement as well as solar and lipo setup. This would be interresting for building wireless sensor devices and integrate them into a free home automation system.
Vielen Dank für das sehr interessante und lehrreiche Video. bin in diesem Thema ganz neu unterwegs und habe folgende Frage: Muss das Solarpanel zwingend 5V haben wenn es an eine Batterie angeschlossen wird oder können die Volt auch höher sein?
Die Spannungen müssen immer inn erhalbden Spezifikationen sein. Die findest du normalerweise dort, we do das Bord kaufst. Oft sind sie für leicht höhere Spannungen ausgelegt.
Hi Andreas, thank you for this video. Do you think the WeMos D1 Mini can resolve some of these issues? Eg. Charge Battery Through Solar Panel, Detect if power goes above or below 4.2v or 3.3v?
+Anton Chan The Wemos are similar to the other NodeMCU boards and have many components on board. For low power applications, I prefer plain ESP-12 modules
sir, you have summarized the learning of multiple weeks to minutes . great video
Thanks!
Hello Andreas, just discovered your channel. How great it is that you are giving all this theoretical background before trying to build something. This is what more people should do. Good work, subscribed.
Welcome to the channel!
I have almost certainly watched this video before but I didn't have any specific interest in solar power then. Now I do and I searched your channel for a video on exactly this topic. As is so often the case, your video gave me exactly the information I needed, avoided unnecessary and irrelevant detail and explained everything clearly. As we say in England, it was spot on. Thank you.
Glad it was helpful! I try to produce the videos with my most important questions in mind...
Hi Andreas,
first thanks a lot for all the good videos!
For a cheap MPP tracker I can recommend a simple TP 4056 Lipo charging board. My setup works as follows: 5V Solar Panel, the charging board (w/ cell protection) and a LiPo battery. As long as the light is very dim, the voltage from the panel is too low for the charging board and no current flows. When there’s more light, the PV voltage climbs over the battery voltage and the charging board starts to charge the battery. If the charging current is too high, the PV voltage drops and so does the current - the PV voltage can climb again. So the charge controller will always keep the PV voltage at the point where it can draw the
maximum power from the panel.
Two things have to be taken care of:
- . - To keep the system from oscillating, I put a capacitor parallel to the panel- The maximum input voltage of the TP4056 is 8V. When the battery is fully charged, no more current is drawn from the panel and the PV voltage will rise to the open circuit voltage. This should not be higher than these 8V
+Frank Schmidt Sounds very promising... have to try it. Thanks for the tip!
Very informative and well organized video. Thanks Andreas for the offering!
You are welcome!
This is my favorite video of yours to date. The fact that someone was kind enough to translate it to Portuguese also means I can scare my father (who thinks solar panels probably aren't much harder than connecting panel, battery and an LED lamp in parallel) all under the guise of a friendly 15min video.
Not gonna lie I still don't get why MPPT is a $100 device instead of a little board that you buy in packs of 50.
Thanks for the nice words!
A trivial typo at 2:15 minutes into the video. "×" used instead of "=".
Your video has read "the question" in the minds of so many and answered excellently. Congratulations and thank you.
You are right! But unfortunately, I cannot correct it anymore...
"Because this is boring we do some tests outside..."... Cool, wirklich nicht schlecht eingebaut. Der Beitrag war aber so oder so sehr belehrend, vor allem die Fakten über die Sonneneinstrahlung und wieviel Energie man nun effektiv abzapfen kann. Besten Dank.
Gern geschehen!
Great video, thanks. I've been running an ESP-07 (all LED's removed) for more than a year now here on the Prairies in Canada where temperatures during the winter drop quite often below -35C. I used trial and error to figure out how long to put the ESP to sleep such that it wouldn't kill the battery during a worse case winter scenario. I will have to watch your video again and then apply these new learnings to see how much margin I was playing with. :) I am using a 1800mAh Li-on battery from a Sony PS3 Controller (rated at max charging current of 400mA and max charging voltage of 4.2 volts) and a 5V- 200mW 4.3cm X 3.4cm solar panel. When I tested the solar panel the maximum voltage I could get was 40mA and about 4.5V so I took a chance and connected it directly to the battery (so far so good and it's been more than a year!). I use only a diode to drop the voltage to the ESP and it has never gone above 3.4V. The ESP is on a 200s sleep cycle. As I said it has been working flawlessly for over a year but tomorrow I want to run this through your math. Thanks again for another wonderful and informative tutorial!
Thank you for your feedback. I think, physical laws are similar, even in the Prairies;-)
Maybe this page helps for calculation of the power requirements of your ESP: battery-life.of-things.de/battery-life-calculator.php
Like your videos. They always start out with what you are trying to solve. That is very important. i often get lost in the weeds and forget about the original intent. I often don't need the projects you are working on BUT I always listen to learn. Great teacher!
Thank you!
Thank you for yet another interesting video. Some things you might want to look into, if you don't get the expected efficiency are: Angle of the panel (I noticed your panel laid flat on the table, which is not optimal), and temperature (cool panels are more efficient than hot panels).
I tried with various angles, but did not get big differences (maybe because it is summer and I did the measurements in the early afternoon. Then, I took the easy(flat) way. But you are absolutely right concerning the angle. If you have a close look at the diagrams in the video, you see the optimal angle for a whole year.
Concerning temperature I cannot do a lot, I think, because the place is defined by the purpose. But, because the most critical time is anyway in December/January, the temperature maybe is not a big issue (in summer, I will have too much energy).
Now I remember. I will have a look at it. Thanks for the tip!
One of the best explanation in the whole youtube. Giving easy explanations to easy concept is an uncommon skill. This is real engineering: i'm sure the beer helps! Greetings from Italy
Thank you very much for your nice words. I just returned from a trip to Italy. It is a beautiful country
Andreas, Your videos, and this one in particular, is excellent. Pitched at just the right pace, and level for me anyway. I like the structure, and the order. These are some of the most informative, and useful, videos I have seen. Grateful thanks from SanFran.
Thank you for your nice words! Nice city, BTW.
I look forward to your controller video. Many of them have a USB connector. They claim up to 5V @ 2A, which is enough to power a Raspberry Pi 3. Of course there is no need to start a discussion on how big the batteries would need to be for that application, since we are talking about an ESP8266 & 18650 batteries. Possibly/probably a small single board controller would be sufficient. I'm now going to watch your video #64 again, about batteries appropriate for the ESP8266. Thanks for your videos.
I think, we also can power a Raspberry Pi. The same formulas would apply, just the panel and the battery would be bigger (about 3 times). And i do not know of a "sleep mode" for the Raspberry
Ja ist denn schon wieder Sonntag :-). Vielen Dank Andreas für die sehr gut strukturierte Vorgehensweise, sowie Deinen Berechnungen und Erklärungen. Freue mich schon auf die Fortsetzung... :-)
Bitte, gern geschehen!
I'm doing something similar. I switched from the ESP8266 to the ESP32 so I can more easily monitor battery voltage and sensor voltages.
One thing I noticed though was that the MPP charge controllers are expensive. So expensive, I just opted to buy more solar cells, which are comparatively cheap, and use a TP.4056 charge module (very cheap) It would have been a better idea for me to have done a proper economic analysis though.
Looking forward to your analysis! Keep up the great work!
The TP.4056 was suggested by other viewers, too. So, I will include it into my investigations.
Here's a graph of the past few days of data logging to google docs. You'll notice gaps in data which is when my ESP32 has been hanging and causing my battery to drain completely. I hope I've solved the problem, but I've found the sample code for things like the DS18B20, Wifi connections, and so forth need some improving. Unless you have some good watch dog timers, you need to make sure your code is really bulletproof.
docs.google.com/spreadsheets/d/1ni20axRWzek2f--lTF_YY7c1_GgZjzM9ziM6vrI6V38/pubchart?oid=108740660&format=interactive
Thank You for your time. I was also making some tests. Here we have more sun. 10W 5V panel + TP4506 + one 18650 + ESP8266 with deep sleep. Not that efficient but is working well so far. Looking forward for the next video.
+Cristiano Sulzbach Thanks for your feedback. I will try the TP4506.
Cool, great topic! Interesting info and very informative video! I have a solar-powered network node laying in my window sill for about a year now. It's a 6V 1W panel connected to a ESP-07 module via a 650 mAh li-ion battery with a charger/protection module. It wakes up for ten seconds every 15 minutes, day and night. It survived summer, but stopped in the winter because it didn't get enough sun. It's only for testing/learning now, but in the future I want to make a solar-powered node for my automated garden project, so plenty work left.
Thanks for the info. Maybe you make the calculations I did for your device and compare the results. Then, we would already have one confirmation (or not) of the theory...
Oh Andreas😀
A superb video, full of useful information and there is going to be a follow up too🙏
It's wonderful that you explain the math and importantly your thought process.
I believe you managed to link together a number of things that I thought I understood. So very happy to see things more clearly.
As always I am very appreciative of your time and efforts and just how concise you are.
The project I am about to start is solar 2 axis for a 160 x 160 solar panel. So this is a great help.
Bless you Andreas 👐
Thanks, Dean. Änd of course, we are interested in the results of your project, especially the efficiency (more energy because of good positioning vs energy consumption of motors)!
Andreas Spiess noted thank you. I will set up another INA219 to monitor the motor power requirements and as you say, work out the efficiency.
From my research you only need to be within + or - 5 degrees. You will still be 99.6% efficient within this range. So had planned to only adjust at 6 degrees.
I assume this is what you were reminding me to think about?
Many thanks
5 years have elapsed. What is your experience with the subject?
@@TishSerg solar big size
There is certainly are lot of issues to consider about solar tech. Sadly more often than not the green desire and market hype clouds peoples judgement when investing in solar as the infamous "solar roadways" and Dave Jones have shown. One thing often not mentioned is how the solar panel heats up when exposed to the sun. The efficiency is affected by the panels temperature. If they are incorrectly mounted on a stone tile roof for example the stone tiles with high thermal mass storage capacity might also help heat up and prevent them from cooling. Another thing is the panels angle to the sun. Most often they get fix mounted. Solar tracking is expensive but might be how the supplier calculate their yield. Furthermore there are always losses when using converters. Storing the energy also mostly entails converting and self discharge losses. And then there is the matter of distributing the energy. Keeping them clean from snow and dust and wet leaf is also something that is often ignored. Thank you Andreas for your excellent video that 'shines light' on some of the issues with solar tech.
I think, there are two distinct solar applications: One is to use solar to get maximum energy out. This is where you find a lot of information. And your issue is also located in this application (btw: look for a comment in this video where a viewer reports, that he looses 20% in hot situations)
The other application is the one I try to deal with: Get enough energy during the whole year. Here, We have to solve other problems like efficiency in low power conditions, deal with low voltage panels etc. Here, I did not find a lot of information. And heat is not so important, because we anyway will have too much energy in summer.
Simply excellent. By the way, I wonder about the reason why there are people who have voted 'I do not like'And I live in Spain,... you can imagine that we have sun, sun and sun
Nobody really knows anything about the "don't like" voters...
Fantastic! It is as if you could see that I was experimenting with a solar panel and Li-On battery today & yesterday way on the other side of the planet. Plus I was trying to calculate panel and cell sizing too. Perfect timing and great spot-on content. Thanks.
AND tomorrow is the Summer Solstice - optimal conditions for experimenting. I'm addicted to your videos.
This is the power of the internet, and English as a global language, I think. Really a beautiful world (at least, this part)...
😮how could i use solar only no battery? "On" to Brown out causes a hang .
Hi Andreas, very interesting topic as always. Your main concern should be the min max temperatures which the device (including charging and battery modules) will "suffer". Chemical batteries most likely loose as much as 70% capacity when cold until they refuse to charge at all. Also due to reduced molecule movement the voltage is reduced when the battery is connected to small loads (few mA). also batteries can be permanently damaged when they get too hot (fast charging and heat from the sun). Think about the Rosetta Research Lander which suffered those problems on the comet it landed on, somewhat extreme example but you will experience similar problems. Lead gel batteries also often used for off the grid solar systems are best but very bulky... when reducing the max load and total power consumption a powercap can be used.
Thank you for your input. You are right, batteries (and temperature) are an additional issue we have to consider.
Another possibility could be to use a big solar panel and super caps. We will see...
Composite systems are quite good at this implementation; There are two power sources, one is solar and one is backup power from the mains electricity.
It works just like you expect when conditions are alright; charge the battery in daytime and use it in nighttime.
But when solar panel voltage drops too much and battery voltage is below threshold, the device turns to backup power so the device doesn't die.
To have grid power is always a good thing and reduces the size/cost of a project considerably. But for me, I want a completely independent device. So, unfortunately, I have to go the hard way (or change where I live...)
Andreas has made science out of what I would have done by trial and error. That is some research!
I also use trial and especially error a lot. But hidden from the channel ;-)
Andreas, your videos just keep getting better!! Great info and fun.
:-)
Andreas, I really enjoy your videos, you obviously put a lot of thought into their structure and order. This makes for some outstandingly informative stuff!!! Many thanks from this very grateful Welshman.
You are welcome. And thank you for your nice words!
Great video. Really simple common sense calculations. I think anyone dabbling should be able to figure out this themselves but if not you would have helped.
What I like best about your channel is the clear straight engineering without any BS. No lies or faked stuff. Too much of the internet and social media is being invaded by persons peddling fake information.
Please do keep up the good work. I will keep watching.
Thanks for your support!
"In the middle I had to stop to drink a beer". This made my day! I really enjoy your videos/tutorials. They are always very informative and full of very useful knowledge as well as great results. Thanks for sharing!
You are welcome. And I am only a mere mortal ;-)
Great informative video, I was hoping this would cover the circuits needed to charge and regulate the voltage to 5v and 3.3v but still a great video
Thank you. I have been looking at videos with similar content for the past few days and I yours has been the most helpful. I hope you continue to provide more. Wish I could give you two thumbs up
Glad it was helpful!
Hi Andreas, the charging unit is very important part of this project I guess. I can't wait to watch your video on that. Thank you
You are welcome!
Excellent. But this deserves a new 2021 version. I'm sure a lot has changed in 4 years.
Good idea! Recently, I did a video about the topic. Maybe you watch it (if not done already).
Great tutorial. This open up so many applications. Im more I treated in the charging units. If there is bigger, smaller and also if you can order them online
So, you have to be patience till I learned enough about that topic...
I would like to suggest that you spend considerable time on the solar panel and the sun's path in Basel. On June 21 at solar noon here in Phoenix it is ~82 above the horizon; yours is ~50'. I noticed that your test panels all seem to have a reflective surfaces. At 50' the sun at solar noon is quite low compared to Phoenix. The most efficient solar angle to the plane is 90' so with your cell lying flat in your picture you are off by 40'. With the reflective surface most of the light is going to bounce right off just like a mirror. So now the question is does tilting the plane make any sense or a tracking panel even better. Here in Phoenix the panels around town all are pointed north and at an angle off perpendicular determined by some formula for year long maximum efficiency. I am old enough to remember seeing the first solar cell back in '56 made by Bell Labs; tiny little thing. Hope this will help you in your decision making. Out here we have been experimenting with the sun for a very long time. APS has a solar farm experimental station at Mcclintock and University in Tempe just down the street from ASU. Try a Google Earth look see. Heat can be a problem for the panel so let it breath. 106' and the sun has set.
I tried to tilt it to the maximum, but the difference was not big. So, I decided to do it flat to be easier comparable between all panels. After your post, I looked into solarelectricityhandbook.com/solar-irradiance.html and compared the two values for Basel/June. Flat, the factor is 5.7 and at 57 degrees (best for summer) it was 5.43. So, also here, the difference does not seem very big.
In this application, It would be advisable to chose the angle for the worst month of the year as we will have too much energy in summer.
I hope, that the surface of solar panels is penetrable for sunlight and does not act as a mirror, but of course, I cannot test it.
BTW: I loved Phoenix and its surroundings when I was there. I flew also to the Sedona airport (with a Cessna) and will never forget that...
mini solar produce constant power you can use( buck converter) or( boost converter) first then you storage power to battery .
the milliamps and voltage change all the time when you produce power and you can change it at any ratio you like , to make it more efficient.
Buck and boost can replace the solar panel controlller?
Thanks a Lot Andreas! If I ever get a chance to come to Basel, I'll definitely drop by!
Cool!
That's an interesting project! One thing though, since these small solar powered devices will stay outside for some years ideally: Did you account for the degradation of the 18650 when fully charged in summer? The cell will degrade much faster then it has to if the temperatures are up and it's fully charged, even though it doesn't need the additional capacity in these conditions. You could even reduce the charge limit dynamically based on the mean temperature to get maximum lifespan. Thanks for your informative videos!
You mention another complication to the project. It will depend on the way I will implement MPPT. If I use a micro controller, I can control all these parameters. With normal MPPT controllers, this is not possible (I assume).
Andreas Spiess That makes sense, it wouldn't bring huge enough benefits for the time spent implementing it.
Sir, since discover your channel, I like it very much and with your kind guidance, I learn a lot!!!! Thank you very much.
Welcome aboard the channel!
Thanks for the calculations! You saved me months of learning! Greetings from Turkey!
You are welcome!
I agree with the positive comments. You are an excellent tutor. Thanks for your efforts.
+Alan Collins :-)
I like the pace and thoroughness of your video. I'm currently on 11:29 and trying to figure out with no results where you obtained the "Solar irradiance figures" presented on the left-hand side of the screen. Thank you.
As usual, the links are in the description (solarelectricityhandbook)
Thank you! Very interesting. I've been thinking of doing some solar energy experiments myself, sadly northern Sweden is lacking in sun hours most of the year. :)
Maybe you still do the calculations... The difference between Dubai and Switzerland also was smaller than I thought...
Certainly your channel deserves more subs.
Thank you!
I have ordered some TP4056 from AliExpress. These connects great to the 18650 battery and deliver the right start and stop for charging. Additionally these are incredible cheap. Looking forward to get them delivered. If they support the MPP I do not know. I assume not.
They are great, but do not support MPPT They are more for normal charging.
You can find something like this on the Avasva page. Full step-by-step instructions right on your desk.
I did not find the page. Only marketing stuff. Do you have a link, please?
Marketing spam. Social marketing bots are spamming TH-cam videos from people like you to make money from subscriptions. Middle-women/men nobody needs. I prefer your Swiss attitude, because it is ethical.
Very good. Very very good. I live in Brasil, Mato Grosso state, very warm and have very sun, and was thinking about and esp8266 for solar weather station, and this video will help me very much, thank you!
So, the results of your calculation will be completely different to mine ;-)
yes, but rules are same, and tips are usefull (don't too low voltage battery, thing about bad weather, etc).
Looking forward very much to the next video especially cutting off charging the battery at 4.2v and at the low voltage point. This should solve issues I've been having using the piZero and the sparkfun sunny buddy MPPT controller.
We will see how I will solve these issues...
Very infomative and entertaining at the same time.
I see a good use for a INA219 I2C module to measure voltage/current...
I think too, that it could be an interesting part. But I also think (or fear), that we have to keep the number of parts as low as possible because they will use energy all the time, also when there is no sun...
Maybe one could switch it on only when needed.
I think, this is a good idea. Maybe even for the whole charger...
Nice vid and a fresh view on the sizing question. But I think you for got one big problem. How to survive freezing conditions with lithium type batteries. For me the fun part of solar is to run something standalone where there is no power like a shed or maybe a bird house or something. Been looking at many solar projects but almost all ignore freezing conditions. The only solution so far (for small devices) is to protect the battery by no charging it. But what is it freezes for 3 or 4 weeks. than the light or sensors will stop working for sure or the wifi on the eps8266 has drained the single 18650. Heating pads require loads of power so back to lead acid than? It all feels very disappointing. How was the beer? :)
I have no long experience. But most documents I read about Li-Ion batteries in the cold are if you charge them with high currents. That is not what we do. And I had viewers reporting that their batteries survived winter. So maybe it is time for a test in your situation? The loss would not be too big, I assume. And you have a fair chance to succeed.
Andreas very useful video. I have been trying to do a similar thing for my esp8266 could not decide on a charge controller. Looking forward to the next episode.
Which controllers were in your selection?
Sunny Buddy - MPPT Solar Charger from sparkfun and Adafruit USB / DC / Solar Lithium Ion/Polymer charger - v2 both quite expensive and then there was also this one thepihut.com/products/solar-lipo-charger-3-7v?ref=isp_rel_prd&isp_ref_pos=4
ofcourse you made no error in your calculation. You are Swiss. Also, a very clear explanation of the MPP
If I remember school, this was not always the case ;-)
Andreas Spiess yes, but that's were your Swissness was honed and polished :-)
Very good video for the right reasons. I have a question: which would be the algorythm to calculate needed panel for chargins a 18650 battery?? Thanks in advance
You can use the formulas in this video.
Hi Andreas,
good information, thank you.
I am running an arduino based weather station with solar power in Zurich. I made good experience with a charger module. This is the one: 03962a / TP4056. but there are two versions, one with dedicated battery connectors and one with shared ones. I prefer the newer version with dedicated connectors.
Cheers
SJ
I should have one of these and will try it out. Thanks for your feedback.
Nice explanation. But an additional factor that has to be concidered depending on location ist the dust! E.g. in Dubai the solar panel would get dirty/dusty after some time and thus convert less energy. This can happen without a dusty desert ... some air polution + dirty rain or snow can destroy the calculation also.
You are right.
Awesome video. I would only start from the other end. You know how many days/hours you need to run the device from a battery, then calculate the size of solar panel. If you need a week of battery life and power consumption of your device, you have a size of a battery. Then you can calculate size of solar panel to charge the battery within a day or two
I think you can do that with the information given.
@@AndreasSpiess thanks for the reply. Would you consider using solar to power up motors? I would love to do rover-like robot. Or powering such device from solar is not feasible?
Nice work. However, i was wondering, wouldnt it be possible to instead of relying on solar panel power in the worst month (The way you do calculations now), incorporate a larger battery to compensate for the winter months? i.e. Store surplus power generated in the summer and use it during the winter? Much like you would for nights and bad weather periods? Or would be battery become too large? Well, if you need 14 batteries to surive 14 days, and you could reduce your power consumption to 10%, then the 14 battery setup should survive 140 days which should be plenty to carry you through the winter. Of course this will increase weight, but the sun-exposed-footprint of the device will be smaller which might make it more convenient. There is probably some room for cost-optimalisation though.
I think, there are many scenarios possible. The formulas will be the same. Maybe it is even possible to "simulate" different scenarios, compare the results, and chose the right one for your power need and location.
I bought a small solar panel from AliExpress sold as "5V 400mA". I measured the open circuit voltage at 6.5V and short circuit current at close to 150mA. I determined the MPPT of the solar panel to be about 300mW at 5.8V and 50mA (measured while adjusting a 10k rheostat in series with the panel). Does this mean I should replaced the resistor on the TP4056 to only draw 50mA?
Optimal power point tracking may also be relevant to mechanical generators like wind generators and bicycle dynamo. That being said, is wind a viable way to support a wireless sensor? If it's not viable on its own, it may provide just the right supplementary power just for the bad weather days, if your climate happens to make rainy days also windy.
I think, wind is also viable, if available in quantity. In my region, we do not have stable winds, so, the effort would not outweigh the reward
Excellent video. Perhaps we need to consider other ways of generating power, such as small wind generators? I haven't seen anything suitable yet, but it must now be possible to make anemometer-style generators that can provide a small but usable charging current to supplement solar panels, or even replace them in areas where wind is more common than sun?
For sure there are other possibilities, but the sun is very handy and everywhere available. Wind, for example, is not available where I live (at least not in quantities).
Andreas I think we have more wind than sun! I will get busy on a wind generator...
The only wind generator I had was the anemometer on my bicycle when I tried to do a power meter ;-) There, I had to produce the wind.
Thank you Andreas, I was smiling a lot watching this video.
Thanks! The world is unsmiling enough. So, at least, our hobby should be different...
Very useful and interesting! You deserve them beers on a hot day sir!
Thank you kindly!
Well, your calculations may have left enough margin for error there. But I feel like it's worth mentioning, that if your battery is large to get you through the night (not the winter), you should design things to work in the worst case scenario and not on average across the whole year. In the summer you will be getting more power than the average and in the winter less. Therefore if your solar panel produces enough power on average over the year, that doesn't mean it will produce enough power in the winter. Especially not on a unusually cloudy day after your solar panel has collected some dust etc. Personally I would oversize the solar panel just a little bit more to be safe.
I think the formulas are here and you can select whatever scenario fits for you. I just made examples.
As always a great and instructive video, I don't know how you find the time but am really glad that you do
Thanks!
+Andreas, thanks for this great video! One question: What is the 10% you mention at 11.25? Is that solar panel efficiency? Because earlier you were using 15% as the efficiency multiplier.
Also, at ~3 minutes in the video you're making calculations based on the year-around solar radiation, but you take into account 15% efficiency for the solar panel, and then another 33% for the charging circuit. Then at end you make the same calculations again using the data for December instead of year-round, however here you only take 10% efficiency into account, so why 10 instead of 15, and what happened to the 33% of the charger circuit, I think we should still take it into account, right?
The 10% take also the losses of battery and charging device into account. The solar cell itself has around 15% and you lose some energy if you charge and discharge a battery
Thanks, I realize now I misunderstood the formula (possibly related to the fact that I was watching it at 2AM :)) I thought you only deducted 10% for the efficiency, but in fact you're deducting 90%
:-)
Thanks Andreas you saves my times and moneys . Very helpful and experimental video Great job.I hope God gives you a more healthy and long life.Best Regards
Thanks!
Thanks for sharing I learned a lot from this and your videos. I am building a weather station using ESP8266 and this is spot for my power requirements. I live in Malta so solar irradiance should be similar to that of Dubai. Many thanks.
You are welcome. It should be easy to generate enough power where you live ;-)
I'm just starting to look into this, so I don't have your hands-on experience, but I was thinking about how one can monitor solar intensity with respect to solar panels. A Lux meter is not too bad, because most panels use the 380-750nm range, but it's not a full correlation with Lumens. The cheaper Lumens meters use two photo-diodes, and calculate Lumens based on their outputs. But their output is not linear, although it does roughly correspond the the human eye. A solar panel has a much more linear response to wavelength, in the visible region.
Maybe you could just use a standard solar panel of the same type as you use in your tests, and record the power generated from it during each test. Then that would automatically take into account the effects of atmospheric moisture, dust etc on solar radiation intensity.
You could even use that to normalize your test data by using the standard power output for a given series of tests: just record it as you do each test. Then if something happens during a specific test you will both know it and be able to compensate for it.
I think, I will only try to monitor the energy which really came from the panel, because I am not interested in the phases with energy in ambundance, only for the critical phases. This is different to most applications where you want to get must out of your panels.
That's not what I mean't. You tried to use a Lux reading to compare readings taken at various times for your system, and my point is that Lux is not an ideal measurement for solar power: the wavelengths and weightings over these wavelength ranges are different for Lux and solar radiation.
I understand that you want to concentrate on the low energy times: Winter basically. So you need some way of making a quantitative measurement of how well a specific system behaves for a given amount of solar energy. So you need to be able to measure solar energy.
Complimenti!Spiegazione Chiara ed esaustiva
Grazie!
Good technical information. Quite thorough and incisive. I look forward to the next video.
Thanks!
Very good analysis. I tried using 3W Poly crystalline panel in India to power the NODEMCU without a battery (using a voltage regulator) but wasn't successful. Can't wait for your next video to see your Battery size recommendation..thank you
Battery recommendations are already in this video. The next will focus on charging the batteries.
Andreas Spiess yes, got it thanks..will watch out for next video on how to charge..will it be a charge controller or charging module for LiPo..will wait and see
Love it, very useful. I'm working on IoT device that will send data only once per 20mins, then in sleep mode.
Thank you!
Interesting video and very nice explanation. For the next episode for the mppt, pwm, etc chargers please take into account that mppt chargers usually draw much more current than the other ones and that it might not be as good as a pwm one especially for low power situations.
+Luc Peeters This is very important! Do you have examples of boards or tests?
Most of these chargers are not for this scenario They use quite high voltages and are rated for many amperes. I did not find many of them suited for low power applications so far.
So so true. Seems we need to make it ourselves or modify . I have a couple of these but with the display, which can be turned off.. www.ebay.co.uk/itm/1PCS-5A-MPPT-Solar-Panel-Controller-Voltage-Step-down-Module-Constant-Current-/292153107129
You need to make sure to use a Schottky diode (place for that on the PCB) and most of these mppt boards since when you have no solar the power from the batt goes back into the board. But this makes it a bit more difficult since the tolerance of all components might make the batt charge over 4.2 Volt. So if I would use these kind of things to make a L-Ion charger I would prefer 2 instead of 1 cells but to make it charge only to 4 Volts which is better for Li-Ion cells anyway. Already looking fwd to next Sunday.
Wonderful narrative style, Andreas!
Thanks!
I love Your videos.
You are encountering and describing with easy words (and tests!) many of problems with building IoT devices. Not only simple "Hello World" projects. :)
Thanks! There are many others around for the "Hello world" stuff. And sometimes, I am also using these examples if I start something new...
Yes, me too. But sometimes I wonder if all TH-cam Tutorials stopped on the same "Hello World" examples and that is all you can get from IoT.
I drank that beer when I was in Switzerland! It's really good!
Thank you!
Great video. Result ( with deep sleep for 10 seconds ) is panel 10x10cm and one 18650 battery. The information about requested capacity of battery is missing. 3000mAh?
2600mAh for 18650 battery.
I think one thing not taken into consideration is low temperatures affecting battery power and life. That can be a difficult obstacle to overcome in higher latitudes. A housing that can absorb some heat from the sun will help, but obviously not when there is no sun. Discharging can also be an issue at lower temperatures, even if it is less of an issue. However, keeping a device running for long periods of time when batteries are cold will be an issue. It looks like Basel doesn't experience temperatures much lower than zero in December and January, but where temperatures are near to or below zero charging batteries can be a problem. Some Lithium batteries will simply not charge below freezing, and will have issues charging below 5 degrees, only allowing charging at very low currents.
I wonder if you have been forced to take this into consideration for your projects?
You are absolutely right. I think here lead-acid batteries are still a good solution, especially if space/weight is no issue.
Hi Andreas. I wonder if Amorphous silicon cells could offer any advantages during cloudy winter months. They have a much lower overall efficiency, but better low-light efficiency.
Maybe. But these days you can afford to take the next bigger panel in many cases.
Sir your explanation is awesome...!! Thankyou..and we need more videos from you.
Every week one (and sometimes a "bonus" one.)..
Andreas Spiess sir we need gsm, sensors (proximity) video. Where GSM sends msg only twice a day but sensor senses actively.
...and i do have another hint. solarcells r working way more efficiently if they are directly pointed to the sun. this means 254,5° movement horizontal, and vertical can be fixed somewhere at the middle between 0 and 65,9° azimuth (also based on basel). 22° az or little more will be optimized for winterseason. it wouldnt be that hard to make the cell moveable, and combined with a solartracker it would be a very nice and interesting topic for your channel :-) i was up to build such, but... yah. didnt.
I did a solar tracker with 4 photo resistors and a dc motor for a school project. it is really simple to make. I bet it would be even easier with a servo motor
I could upload it if you don't mind that is all written in spanish.
Patchoulino Don't think so. Servos usually rotate 180deg i just read. As calculatet more as 250deg is needet. And exact positioning isn't needed. Follow up, and endswitches should be good enough.
I am sure that this is great for high power applications. But it would be interesting to fin out for small panels, if the energy used for the servos can be made up with the additional energy you get...
Andreas Spiess i wont glue on servos. Thinking more on low power gear motor. Needn't to be fast or strong...
Great thinking. The thrill of building variable vector control would justify doing this. However, the practical consideration of the added equipment (and load) to extract nominal gains would perhaps be defeated by what would more easily be achieved by adding one more cell. In a well balanced system, much of the power is not used anyway to protect overcharge, right? But, yes, someday I'd like to try the servo control anyway.
If a 180 servo is centered on the 254 horizontal sweep, only 37 degrees would be missing on each side. Better yet, orient the sweep to the most critical conditions knowing that maximum voltages will be clipped by the battery regulator anyway.
These videos are wonderful treasures Andreas - Thank You!
You are welcome!
Another very useful video, Andreas! I assume it gets below freezing many days in the Swiss winter, true? I think that Lithium batteries should not be charged below freezing, true? What type of battery would be good to use for a solar project that will be below freezing (say down to -10C) some days?
I newer tested batteries below freezing so far, but I read about it. Maybe stuff for another video?
Hi thanks for the video.
Yes designing a 24/24/365 solar powered system is quite different than to design a device to maximize solar harvesting during sunny days.
Once you design it to gather enough power to pass bad winter days, you will discover that the panels will deliver > 500 times too much energy during hot sunny days.
So one thing to consider is what to do with that energy?
My current approach is to use that energy to power fans, that will cool down the whole system.
The fans are a very good idea ! I hope, the factor will not be >500...
Ups! Of course >500 was exaggerated. I meant factor ~50. I have tested my 5W panel to deliver 15-20mA on rainy days compared to 800mA maximum in bright sun. Anyhow these little 20mA are reducing the energy deficit by ~25%.
I will activate the fans by the signal "fully charged" of my charger modules, essentially to avoid a source voltage increase above the specs of the regulator when it stops dragging current. Additionally the Arduino will be able to activate them temporarily as well, when its own temperature exceeds a threshold, that will improve the panel efficiency.
Sounds like a good idea! Thanks for the tip
Wow! Thanks for sharing your efforts! I may have to make a solar charged battery pack now!
:-)
Great detail, appreciate the info. Yes, I agree with other comments, a well organised video. Thanks again Andreas.
:-)
Love your videos! Can't wait until I can begin using solar panels myself
Go for it!
I didn't close my eyes one single time. Great video.
About the battery size for 14 days of bad weather. Wouldn't this assume that there is absolutely no sun for 14 days? Even if the weather is bad and it is pretty cloudy, the solar panels should convert at least some energy. Or do panels not work at all when it's really cloudy?
The power drops to nearly zero if it is cloudy.
Hi, Andreas, very interesting tutorial (as usual). I've got a question: To calculate the watt per m2 available, is it realy correct to divide the annual solar radiation by the number hours in one year (8760) ? Since there is no radiation at night, wouldn't it be exact to divide this annual solar radiation by the number of hours of sunshine each year ?
Maybe you are right and I made a mistake. I do no more remember the details :-(
Thanks for sharing your experiences! Curiosity: Are you doing similar work for a living?
No, not at all. I am building ERP systems for global companies...
One important variable that should be considered also is the environment temperature. I have a 500W solar panel that reduces ~20% when we have hot weather. So, with clear direct face of the sun (and perfect inclination) i have almost have full power in the winter (498W was my best measure) and with around 30ºC or more, i have around 400W
+Nuno Justo Thanks for your info. This is more important for the usual setup. For this setup it is even an advantage, because we need every Watt in winter and we have to much in summer.
yeah, definately. Soon i will do my first project with a 100x60 cm panel. This will power up my new toy omega 2 plus. you should take a look at it. onion.io/omega2/
Awesome! @8:17 , any good? Haven't seen Quöllfrisch in the Philippines, but we do get plenty sun.
Yea, you for sure can live with smaller solar panels than we can. But you probably need more beer ;-)
A Review of a full setup would be nice. E.g. OpenHAB2, MQTT, ESP32, Mongoose OS, powermanagement as well as solar and lipo setup. This would be interresting for building wireless sensor devices and integrate them into a free home automation system.
I think, the systems setup would be your job ;-). I concentrate on the parts. Like that, my content is more widely useful.
Nice videos by te way 😉
:-)
I was just wondering about this....and my favorite Swiss vlogger comes through
Coincidence or luck? Or maybe both...
Very very educational. Thanks for your knowledge, Andreas.
My pleasure!
Vielen Dank für das sehr interessante und lehrreiche Video.
bin in diesem Thema ganz neu unterwegs und habe folgende Frage:
Muss das Solarpanel zwingend 5V haben wenn es an eine Batterie angeschlossen wird oder können die Volt auch höher sein?
Die Spannungen müssen immer inn erhalbden Spezifikationen sein. Die findest du normalerweise dort, we do das Bord kaufst. Oft sind sie für leicht höhere Spannungen ausgelegt.
@@AndreasSpiess Vielen Dank für die rasche Antwort! Folge dem interessanten Kanal sehr gerne :)
Thank you so much Andreas. You are doing an amazing job. Your videos are simply perfect.
Thank you for your nice words!
Hi Andreas, thank you for this video.
Do you think the WeMos D1 Mini can resolve some of these issues? Eg. Charge Battery Through Solar Panel, Detect if power goes above or below 4.2v or 3.3v?
+Anton Chan The Wemos are similar to the other NodeMCU boards and have many components on board. For low power applications, I prefer plain ESP-12 modules
Tollen Video, ich freue mich schon auf die Fortsetzungen....
:-)
Exactly what I was looking for. Thank you!
You are welcome!