Відео

Not sure who you are buying from.. I bought a kWeld a year or so ago from a US supplier for 235 USD with free shipping. That's just $60 more than directly from Germany, not including whatever VAT or shipping would've been ontop.

@@nickldominator I wonder if things changed in the year. Because the manufacturer indicates for USA and Canada the one and only distributor is the Canadian one. At least as of late 2022. There would be zero VAT on top for exports from the EU to the USA.

@@nonionbeezness Not sure if my reply ever went through, but I can send you the email to the supplier I bought mine from. I checked their live spreadsheet of availability, and they have 13 units in stock ready to ship and still sell for $235 free shipping.

@@nickldominator sure I’d be interested in taking a look at a reasonably priced US source.

Will be looking for one of these in the foreseeable future. If I must wear the label of “Stupid” or worse to get the real deal - I will, and pay the surcharge as well. I need all the help I can get.

Yes, good point. I only use when switching off to drain the capacitors.

I do not know were you got the information that these are supercaps. They are large Electrolytic capacitors Max voltage 25V so they can contain lots of energy.

Hello! Could you please give more information to how you managed to purchase the boards only?

The reason for this is 2 fold: -The tracking part is hardly determining performance -Over 90% of the MPPT tracking algorithm used is perturb and observe so this is hardly a discriminating feature The loss in the converter however is a major contributor however. If it is not very low loss al the benefits of an MPPT are basically nullified. An example: A simple PWM charge controller with a MOSFET with a low Rdson can achieve an efficiency of 99.5% @ 10A a poor MPPT will achieve 90%@10A. If your panels are hot the benefit of an MPPT will be around 8% because the MP shifts to the battery voltage. So now your regular charger will perform better. So you have an expensive MPPT, charging your battery slower, having to dissipate 13-14W often in hot boat, tent or camper van. So efficiency is key!

yes or something with CAN-bus

in the interrupt routine you also sample the state of the input that is not stable, I do not understand why.

@@jansunman7393 Thanks for answering my comment. As far as i see , you are doing the same thing in your code. In the routine part for A your last( inner) if-statement is : if(A==GPIO_PIN_RESET) count--; and therefore i read the state of A

@@BerndSchmitt-Martinique the statement you are referring to is not an assignment! in your ISR for A, you assign (sample ) input A at the moment it is not stable: B_new = digitalReadFast(PIN_B); // read stable B A_new = digitalReadFast(PIN_A); // read current A <- A must be sampled in the ISR for B not here

@@jansunman7393 I recognize my bug - thank you very much for your quick help . Your wonderful routine is working like a charm ( i think, this is the adequate english expression ). It is by far the BEST and SHORTEST and most EFFECTIVE code for ROTARY ENCODERS. ( and i have tried a whole bunch ). I am a retired Dipl.-Engineer from germany. My hobbies are FLIGHT-SIMULATION and all kinds of microcontrollers and peripherals for PANELBUILDING. These ( desktop - ) panels are equipped with encoders , buttons , switches and displays to control the planes and get useful informations out of the simulator. Using a panel gives more realistic feeling than using the mouse on the sscreen. If you are interested in technical discussions here is my email adress : bs.fortdefrance@gmail.com in any case - thanks a lot - you have done a fine job . Bernd Schmitt

@@BerndSchmitt-Martinique thnx Bernd, you are welcome

it is not the kweld software, the mosfets are just switched with the time that is set

@@jansunman7393 so how i can do the same as you?

@@jansunman7393 Uh yeah, it is, because the software on mine does NOT give me the opportunity to set the pulselength, only energy in Joules. The pulse-length is automatically determined by the unit, based on the current it has seen flowing through the weld.

@@weeardguy yes, I re-programmed the default software because with the default k-weld software it is not possible. I also have a scope attached to monitor the discharge to stay in safe boundaries wrt MOSFET failure.

@@jansunman7393 In what programming language is it written? How did you modify this program?

Is this based on my example?

@@stefanvandepas4249 no

But I am asking what is the reason for timer being stuck in interrupt routine

Be sure you are clearing the timer's interrupt pending flag inside the interrupt routine. I believe you clear it using __HAL_TIM_CLEAR_IT().

Hi Brian. I m interested to modify my kwelder. Did you manage to find any information on the modification? What capacitors? Thanks

Judging by the size they are some 300000 to 500000 uF caps. My 35 or 40 Volt caps are about twice as short for 100000 uF, so paired with the lower maximum voltage of the ones in the video, I consider this a good estimate. The kWeld most likely performs nothing different: it doesn't know what it's power supply is and will just drive the FETs like normal. It's the ultralow impedance caused by the large copper busbars and caps that it can (apparently, it's not shown that it's actually able to push that amount of current through the wires, which I highly doubt) that increases the maximum current. But to be honest, I doubt one manages to get it to 3800 amps. At a fully charged 12,6 Volt from the LiPo I use, it manages to draw nearly 1800 Amps in calibration mode. But the wires and electrodes feel this stress and warm up from this calibration pulse and even though the wires are very short on mine (about 120 mm in total from kWeld to battery and about 20 centimeter or so on every electrode cable) my setup gets a mad jolt from the electromagnetic forces exerted on the whole unit while mounted in my Proxxon drillstand (as I have the electrodes hanging down). I seriously do not want to think about what happens to the cables and unit at the given 3800 amps. I also think one has to wait LONG for the power supply to get the caps back up to a decent voltage. As I now work at a different place, where I can use a sophisticated milli-ohm meter, I'm now thinking of taking my kWeld there and see what the optimum torque on the bolts and nuts does to the overall resistance of the unit.

I am afraid not dometic seemed not to be interested. To bad it is working great using it for 2 years now and there is much interest on the beach :-)

@@jansunman7393 what a shame... Any idea what I must do to get my hands on one? I still haven't found a suitable charger... I even built an arduino "mppt" charger but it turned out to be a performance let down but at least it was a nice learning experience.

@@jansunman7393 any news or github repository?

STM32CubeIDE 1.42 / STM32CubeMX 6.01-RC3

adding another thumb up.. YT lost the 1st one :)

Well I would not use a 18V input and reduce to around 5V efficiency will never be very good with that kind of voltage drop. Also note with a constant output voltage like 5V you cannot do MPPT.

@@jansunman7393 Oh I didn't know that, appreciate the information! I read in the description of one dc to dc charger module that the higher input voltage the higher the efficiency. See in description s.lazada.co.th/s.XMN5Q I also ordered that one, looked modern and neat! :) So for mppt the output voltage must be variable... I see, good to know.

@@jansunman7393 They are so dirt cheap I ordered also this one s.lazada.co.th/s.XMDr2 and this one s.lazada.co.th/s.XMSHV They say an efficiency up to 96% which sounds pretty good! I figured that I can't use mppt for my daily driver which is the 6v 12W panel so (6v input to 5.4v to 4.9v output is a very small range and makes not much difference to use mppt I suppose) I got these dc to dc converters since they claim 90 to 96% which is pretty all right. Previously some of these charges destroyed my powerbanks (guess poor circuit and too high output ripples) or they performed lousy. They also get super hot. Hope one of these new ones will perform better than my present one. Will have to wait a few weeks till they are here since shipped from China.

@@jansunman7393 Also thank you so much for replying!! I really have a lot of respect of smart people who can make things better than others! It's so hard to find an affordable, efficient and handy charger. As well not many people review those available online. Your videos are so easy, straightforward and informative, I like them! In case you have a few of those chargers and sell them I would be happy to buy one or even 2 or 3 of these if this is more attractive for you. If you can make an efficient version for a 6v panel with 5.4 to 4.2v output (or whatever the usual USB range is) then I would also be very grateful to be able to purchase one of those. I use a 6v panel daily whenever I am living in the forest without electricity which is usually several months a year. Looking forward for your reply! 😊

Hi Mr Max I saw a review: ua-cam.com/video/liYZ5pYOZDE/v-deo.html does not seem to be very good....

The expensive one has a large cooler for only 5A mppt charger that must be total crap.

@@jansunman7393 Thanks!

@@jansunman7393 Thank you, good to know!

@@jansunman7393 I actually ordered this cheap one (3euro) s.lazada.co.th/s.XMJaVa (I am in Thailand and lazada is like ebay here) few hours ago to start with something. Since I wanted to use lifepo4 I had not much choices, most of these chargers are for li-ion or lead acid. I also hope to get two of these cheap meters (5euro) to run some tests s.lazada.co.th/s.XMEWb.

its to limit the input voltage to not overdischarge them. I dont know whats the minimum input voltage tho

Because 1C is an appropriate charge current for this battery.

The battery is discharged at 1C so 2A. The input voltage determined by the load. It is not explicitely set.

Hi Tore, I would be surprised if there would be that kind of demand. Apart from a solar charger it also supports connection to an adapter an it can be used as a battery charger. What kind of batteries should be supported in your opinion?

@@jansunman7393 Just a CC CV scheme with a potentiometer for end voltage and a solder bridge selectable minimum current "end of charge" or continuous float option, and just leave the rest to whatever BMS is connected to different battery chemistries. I think you are too modest, to my knowledge, the only other option for a true MPPT charger in the <5A range is the Genasun, then there are the "simulated" MPPT chargers, that work by a panel specific manually set PP voltage, but they are unable to cope with dirt and season changes i panel characteristics, Like the BQ24650, that even has panel temperature and battery temperature compensation, and do a fairly good job if tuned for the exact conditions 94%, but it is not for everyone to calculate temp sensor resistor divider through a half page differential equation in the data sheet (I skipped that part). So an efficient MPPT charger that doesn't require reading datasheets that could be a building block for whatever university project that needs charging, or the DIY'er like me or a cheaper alternative to the Genasun bare board version that costs EUR 75 for OEMs. You would really fill a niche! So you can order populated boards in small quantities, so do a pre order for 20 PCBs and have people sign up and pay beforehand, and you won't bare much of the risk. Then people will try them out, make youtube videos about them and before you know it, you'll get requests for more. You might want to do the controller programming and soldering yourself including the manual soldered connectors just leaving the SMD parts to China, not to be beaten by Chinese knockoffs appearing on Aliexpress, and to save costs on the first order?? I'm convinced you got something unique here.

Hi Tore, well it could be done. Is there a way to reach you with a PM to further discuss?

@@jansunman7393 Cheers, my email is svp@km.dk and my YT name is my real name (I'm not afraid of scammers spamming me, I'm a guy and my profile picture is of my daughter). I just made a suggestion to how this could work, but I will sign up to be on the list, so you got one already! I'm in Copenhagen Denmark.

This is a very good remark. The reason I take 17V is that mostly in "12V" panels have 34 or 36 cells meaning that the maximum power point will be around 17V. This controller does not support series connected panels. So that is the reason why. But in general you are right the larger the voltage drop over the converter the less efficient. To emulate 2 in series I would take 34V. " Anyway, only the newbies worry about efficiency" -> I have to completely disagree here. Tracking is really simple and it is easy with a simple micro to do it 15-20 times per second. Making a highly efficient converter however is very complex, with many trade-offs. "Efficiency numbers are highly over rated." -> I disagree also here efficiency is one of the most important spec. points for an MPPT because otherwise your regular charge controller will out perform it. A simple PWM charge controller with a MOSFET with a low Rdson can achieve an efficiency of 99.5% @ 10A a poor MPPT will achieve 90%@10A. If your panels are hot the benefit of an MPPT will be around 8% because the MP shifts to the battery voltage. So now your regular charger will perform better. So you have an expensive MPPT, charging your battery slower, having to dissipate 13-14W often in hot boat, tent or camper van. So efficiency is key! Do not trust MPPT chargers with large cooling fins.

@@jansunman7393 You really have a clue of these things, can you please recommend me an efficient mppt charger you approve of?

On what event? If it gets too hot?

UnderVoltage

it does not cut off but it wil give a bied signal cuts off voltage rises and starts again. So you have to watch it!!!!!! This is my oppinion.

I have allready burned up 2 mosfets and i mean the divice caught fire !!so be carefull !!!

Hi outthereandaway, yes I do real tracking, meaning that the panel voltage is adjusted 15 times per second to maximize the output current to the battery. The maximum PV voltage is 28V the minimum battery voltage is 3V. Initially the current is determined the power that the PV panel is able to generate but limited to 3A because that is the thermal limit of the converter. If the current is limited than of course you choose a too large panel for this converter. When the battery is at its maximum voltage the current is reduced to maintain that voltage (CV charging), it is all a matter of software. So that is suitable for lipo charging (without balancing). Wrt schematics I do not want to share them at this moment because there is a novelty in the circuit that I have never seen and that makes it small, cheap and efficient. Btw you also would need the software if you want to reproduce it. What kind of power do you need for your iot stuff?

jan sunman Hi, thanks a lot for your reply! Very interesting. Oh yeah, I totally understand, that you do not want to share too much details. I was just curious, what components you used since I might try building a mppt for very small modules on my own in the near future. In order for this to make sense, it would have to be small, efficient and also cheap... Just like yours ;-) I just ordered some cheap 6V modules and I will do some experimenting with a couple of mosfets... Just to see, if I could use them for charging 18650s. Just for fun ;-) By the way, you showed efficiency in full sunlight (probably around noon). Here, your device performs really well. Congrats! Have you also done some tests in rapidly changing or cloudy weather, changing module temperatures and lowlight situations? Have you had the chance to collect any (efficiency-) data under such conditions? I think, the whole point of using a full-blown mppt is to be able to get the most out of the pv-module, especially in those situations...

Hi outthereandaway, For clarity I think you have to distiguish between "tracking efficiency", how fast do you itterate to the MPP and "converter efficiency" how much energy gets lost in the conversion. The main converter loss is in the MOSFET's at high currents. With respect to the tracking efficiency I can tell that simulation shows the the control system that controls the panel voltage stabilizes within 10ms after a step response in the irradiation. So the perterb and observe loop could run at 100 Hz. The perterb and observe algorithm needs about 10 itterations so it will keep up with irradiation variations of say 2Hz without being off the MPP.

@@jansunman7393 Jan, truly a shame that there is no commercial version available even after 2 years have already passed. If I would be a businessman I would finance it. Do you still plan to do one or will you take it to the grave so to say? I know at least two DIY developers who would be very interested in your knowledge and understanding and who would share their expertise for free (well for joy but not for money) guaranteed in case you ever decide you want to gift your findings to the DIY community. They are really nice guys who enjoy helping others with their knowledge, even absolute beginners like me. I eventually managed to complete one SoftPower version but unfortunately it draws too much power as well mppt didn't really work. It turned out to be not what I was looking for, a cheap reliable small mppt charge controller with little self consumption and basic security functions. Mayve have a look discord.gg/3QjNuyZm And the "open solar project" as well. Maybe you like to do something with them, they always happy to meet capable people who also share their love for tinkering. Both are engineers, one still working the other one retired. Very sweet guys, they had a lot of patients with me since it was all new land to me. Its fun for sure but not easy... 😏

Hi Adrian, I am using a hill climbing method so perturb and observe at 15Hz. Simulations indicate an efficiency above 95% but I am also very curious how good it is in real life. Have to buy a second Portapow ;-)