 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Solar : Simplified Solar Hot Water
  Page 5 of 8   |
|||||
| Author | Message | ||||
| Davo99 Guru  Joined: 03/06/2019 Location: AustraliaPosts: 1577 |
Hi Tony, Thanks for the List. I have ordered/ already got everything I need bar the mosfet gate Driver and the Mosfets. Could you recommend a source for these? I do happen to have a bunch of CLA50E1200HB if these would work at all? |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
That CLA50E1200HB is a thyristor not a mosfet. Not going to work in this unfortunately. https://pdf1.alldatasheet.com/datasheet-pdf/view/499741/IXYS/CLA50E1200HB.html I ordered ten FOD3182 gate drivers from Xon electronics. https://www.x-on.com.au/?gclid=EAIaIQobChMIyPG3o_2E6AIVwhaPCh3xnwL6EAAYASAAEgKfffD_BwE More expensive than China, but much faster as I do not have any here at all. Likewise the mosfets are available from e-bay as well. These are going to be the most expensive parts, so its well worth shopping around. Edited 2020-03-06 14:15 by Warpspeed Cheers, Tony. |
||||
| Davo99 Guru Joined: 03/06/2019 Location: AustraliaPosts: 1577 |
Frustrated none of the boards I have been waiting on arrived and being an idiot and blowing the other board I had accidently shorting it while still putting it together, I decided to hook up what I had and test a theory. Got an arduino board and with great time, effort and with rat cunning and inginuety, put the blink sketch into it and modified it to run 1/3rd of a second pulse. Wired it up to a DC SSR (Whats the difference between DC and AC on the output side of an SSR anyway?) and hooked that to the cap bank. Took it out and put it on an array of 4 X 250W panels in series I propped up flat on milk crates (so as not to burn the Kaka out of the lawn) and hooked that to the little water heater. Being a nice cloudy day direct connection to the 3600W element gave me about 1.5 A @ around 50V. At the 1/3rd of a sec on/off, I was seeing the volts go up to about 120 and fall to about 60 On the multimeter and the analogue amp meter was swinging up to about 6 A and falling to about 2. I changed the timing to 100/th of a sec on and off and the amperage seems to be going to about 4 and the voltage is reading about 90. Pretty certain the pulsing is just too quick to get an even reading. I'll have to go back and re read how to level it out with a cap and resistor. I have a DC power meter but that only goes up to 100V so don't want to risk that. The main thing is the water IS getting warmer far quicker than what I would expect. After only 90 Min the temp has Climbed about 9 oC in the 50L tank. MUCH more than I would have expected. Looking at one of my GTI's, it's doing about 1000W atm when on a sunny day it would be pushing 4KW so the solar generation is definitely down. I'm somewhat surprised this very crude setup has worked as well as it seems to be. So far been running it in the middle of the day although I don't think the light has changed that much since about 9:30 am. Maybe I should take some light readings to confirm. Minimal testing but it does give some early vindication to the idea that even a basic flat Pulsing would be an advantage. Unfortunately the panels are well matched to the element but I don't think that works so well on days like today. When I short out the cap bank, Both the amp and the voltage readings fall significantly. I don't think the pulsed side is accurate but I do think it is indicative. I might look at adding 2 more panels tomorrow and splitting them so I can stay under the 100V and add my power meter to try and get an idea of what it is producing straight and pulsed. I did find a battery charger board on Fleabay that runs Mosfets instead of relays. This could be really simple to hook to a voltage Divider and drive an SSR with a cap bank for those not fortunate enough to have the help and resources here. I think that will work well as a poormans PWM/ MPPT and be miles ahead of Ohm matching. This would track the different conditions and adjust the pulse speed as needed for best output. I need to read up more on voltage dividers. I put one on the cap bank I have set up and still seeing over 100V on the meter when I was aiming for about 12 and should be less at this voltage. :0( I'll give the tank a couple more hours and go out and do another check with the heat gun to see what I get. Scored some points with the Mrs as well. She was hanging out the washing lamenting the humid conditions from the rain last night and the lack of sun and wind. Took the car radiator fans I have under the back verandah circulating the air that gets trapped which are hooked to a couple of panels and put them on a ladder near the clothes line. Washing is spinning quite well now and Mrs is happy and confident washing will be well dried. I can see a nice baked Dinner happening tonight! :0) Was thinking to put a panel on top of the clothes line and a fan on one end of an arm and have it propel itself round and round like that. Might just have to patent the idea first! :0) |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
If its anything like the Turnigy power meter I have, all you will see is scrambled numbers. My Turnigy definitely needs steady dc to take a useful power reading. You could try measuring the power from just two of the four panels, and assume the other two are doing an equal job. That way you can measure the current, but only half the final voltage. Don't expect to hear anything from JCL until later on in the week, then it will be a few days after that until the boards finally show up here. But I have everything here now ready to complete one board. Cheers, Tony. |
||||
| Davo99 Guru Joined: 03/06/2019 Location: AustraliaPosts: 1577 |
Yes I think it would be difficult for most non highly expensive meters to get a stable reading. I buy lots of Multimeters, have a fetish for them with the variety you can get these days but only buy the ebay cheapies mainly. Have a couple of half decent ones I am very careful with. You could try measuring the power from just two of the four panels, and assume the other two are doing an equal job. That way you can measure the current, but only half the final voltage. Brilliant Idea Tony. Again I am in awe and embarrassed by your simple and effective solutions. I'll have to work out how to do that when they are all in series and I'm using an amp meter. DC is such a pain compared to AC. The weather turned erratic this afternoon. For the most part overcast to gloomy with a patch of sunshine. Noticeably pushed the volts up when the sun came out the meter was reading. Not sure how to interpret that. The amps stayed the same pretty much given they were up there from the start but I don't know if the Voltage being under Vmp ( Which Climbed in the sun) meant the On time was too Long. I'll load the arduino programmer on my laptop tonight and play with different times tomorrow and see what effect they have. Might give me an idea of the charge up and discharge times for the caps. Seemed my water temp measurement was out. I used a heat gun on a couple of the brass plugs on the tank thinking they would be representative of the tank temp. Seems not so. After a few more hours I only got another 4oC temp rise. At first I thought there wasn't a lot of power to be had then I did some mental arithmetic and it didn't seem right. I opened the valve on the tank and could feel the water was way warmer. I let it run in my hand and shot it and got 42 oC which seemed spot on. I can't be sure now what the initial temp was but no doubt the tank Climbed a good 20oC or more in about 4 hours. No way it would have been above 20 when I started with the weather we have had and the sub 20O Nights, around 15o. I'll dig out my Thermometer with the flexible thermo couple and stick it through the top port in the tank and see what I come up with tomorrow. Late this afternoon it came over very dark and stormy looking and the readings were very telling. The voltage was hovering around 60-80V and the amperage appeared to drop slightly, maybe 1A. Direct connecting told a story though. Less than 1A @ 22V. Even pulsing, I'm sure the output is far Higher over all with the caps. I learned much more than I though I would with this little half baked exercise and raised a few questions to find answers to as well. Very convinced pulsing the output through caps can make a big difference though. Last winter I had a 100L water tank in my little greenhouse as a thermal ballast to stop everything freezing at night. I had 3 Kw of direct panels on that and it struggled to get above 40o. I'm sure with a simple pulsing setup I can get it much warmer than that with 1Kw of panels. With a voltage tuned setup.... Be that much better again. |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
It will make an even bigger difference once its properly cycling between two well chosen voltage points. It will then self adjust over a very wide range of solar input. Cheers, Tony. |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Dave, DHL delivered a nice little package for us about an hour ago. If you can pm me with your address, it could be in the mail to you today along with the four extra blank boards. Just bench tested it, no surprises, everything works exactly as anticipated. Interestingly, the +12v supply is sightly low at only 11.74v which causes the switching points to coincide almost exactly with the resistor values fitted. 180K=180v 220K=220v 270K=270v etc.. These little supplies always tend to be a bit short of voltage, typically 11.9v. I have never actually seen one at the full rated 12.0v.  Edited 2020-03-12 15:04 by Warpspeed Cheers, Tony. |
||||
| Davo99 Guru Joined: 03/06/2019 Location: AustraliaPosts: 1577 |
Made my week Tony, Literally! My own efforts have been very depressing. Blew a few boards with different accidents and stupidity but I guess I learned something. Then the one DC SSR I had gave up which I was really confused about till I did some research and found this is not uncommon. Haven't been able to play with this at all and it's very frustrating but I am trying to learn more in able to do things properly. I have been reading up on how to use the mosfets and studying this circuit quite a bit and going back and forth learning how it works and what and why you have done things on it. Got most of it now and can appreciate it's simplicity and how clever you have done it. Just need to read up on the Mosfet driver to understand that better as what I have read so far seems a bit conflicting. The thing that is going to be hard to learn for me is the physical construction of something not on a PCB. Breadboarding confuses me a lot! These boards look great! Very different to how I pictured them and a very neat layout! Seeing things laid out like this makes me confident even a Dilettante like me could put them together. Tis a thing of beauty! My idea was to put the Mosfets on a CPU heat sink with a fan. I take it as long as I keep it under 100o ( Mosfet is Limited to 150 right?) that should be ok or do they have to be cooler as I would try for anyway? I have sent you a PM. Thanks again Tony, VERY much appreciated! |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Pm received, boards will be in the mail first thing tomorrow morning. The mosfet driver chip does three things. It inverts the signal from the LM555, as the output of the LM555 is "upside down". Pin three is logic low when we want our mosfets to switch on, and goes logic high when we want the mosfets to switch off. The gate drive chip also provides a nice clean very fast high power drive waveform to the mosfet gates, and can easily drive a couple of giant mosfets. The gate driver chip also has an internal undervoltage shutdown. At sunrise and sunset the small 12v dc supply really suffers and struggles, and the undervoltage shutdown ensures that the mosfets cannot "half" turn on and create even more problems. The mosfets will either be held off when the +12v supply is very low, or start working when the +12v is high enough for the system to function. It will probably still have fits and starts when the sun is just below the horizon, but its not going to be destructive. Its an interesting thing to watch any solar control system struggle pitifully into life at first light, and usually die in convulsions at dusk. Anyhow, I wished to keep the whole thing very simple, with the fewest possible parts, and an opto isolated gate driver has the exact three functions I wanted all in one chip, even though the opto isolation feature is not being used. That part of the circuit is a little "odd" but its the simplest fewest parts solution I could come up with. Cheers, Tony. |
||||
mason Regular Member  Joined: 07/11/2015 Location: CanadaPosts: 86 |
Nice circuit Tony, we'll probably see them on AliExpress in the coming weeks lol |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
I hope so Mason, that would make it all even easier. Just a reminder to everyone, I can supply Gerbers for this board to anyone that wants them, but I cannot attach the Gerber files to the thread or a PM directly. You will have to e-mail me first so I can then attach the Gerbers to a reply. To do that, go to the top of this page and click on "members". Click on "W" then scroll down to "Warpspeed" and click on that. My personal details should come up, along with my e-mail address. That way, you can order your own boards from China direct. Five boards cost about three dollars, but the postage will be more than that. If I order in a big bunch of boards myself, I have to re post them and mess around with small ammonts of money and make numerous trips back and forth to the post office every time, and I really cannot be bothered with all of that. Much easier for both of us if you order your own boards direct from China. Feel free to pass the Gerbers on to third parties, its all completely public domain. This is all so wonderfully simple, it really has no commercial value except perhaps for the Chinese, and I would like to see the whole concept become as widespread and known about as possible. Cheers, Tony. |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Dave, in the mail, tracking number R332013001563524 Slightly less likely to end up in The Twilight Zone. Cheers, Tony. |
||||
| Davo99 Guru Joined: 03/06/2019 Location: AustraliaPosts: 1577 |
Thanks again Tony, So much appreciated. The timing is really bad though, I have to go up to Dads for about a fortnight and I'm going to be dying to try this out the whole damn time I'm away! :0) His neighbour that wants to go off grid has been asking Dad when I'll be back up swo I surmise I'm going to be setting up his little system. Funny guy, Tight as a fishes Bum but seems to have no problem spending significant money on the idea of saving money that he will never recover or has next to no chance of. Often one can't tell some people that sometimes just paying the bill is the cheapest option of all. If I get the boards before I leave, I might just have to load up some panels and the water heater and take them with me. :0) I have different thoughts on this to you as far as commercial value. I can see a great range of uses for it. I think as well as off grid, it could be a great supplement to on grid users as well. I have seen what you are talking about with the sunrise/ sunset situation. I have hooked up a little USB converter to a panel on the roof Over my office. I put a little digital volt meter on it and have been charging Phones, tablets, batterys etc with it. Given it's on a 190W panel, it works well even on the wet days. The sunup/ sundown is interesting. Even with just the volt meter, there is a period where it flicks around and flashes on and off and carrys on. I also notice that the voltage drop is VERY rapid. The panel is either working or it's not. People still maintain that the panel Voltage is light dependent but one only needs to see the rise and fall off to see what is completely false... unless you are talking about MA of current at the time. Thanks again Tony, Can't wait to try this out! Been something that's been on my mind for a couple of years at least. |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Not to worry Dave, its all still going to be there when you get back in a couple of weeks  Cheers, Tony. |
||||
| davef Guru Joined: 14/05/2006 Location: New ZealandPosts: 499 |
Read through the whole thread to see where the discussion on the "sunrise/sunset situation" was. I assume this has to do with how the system fires up/down when powered solely from the PV input. Just some observations ... if I power a LM358 off, say 0-30 volt ramp through a 1K resistor with a 12V Zener to ground, a signal voltage on the +ve input AND it is configured as a non-inverting op-amp it fires correctly. When configured as an inverting op-amp with the signal on the -ve input it does NOT fire up correctly. Does the LM555 always fire up in the proper operating condition as the PV ramps up? Dave Edited 2020-03-15 17:33 by davef |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Yes it works hazard free. The whole system will power up at about +45 volts, but that is highly variable. Those little supplies come from multiple different manufacturers in China and are not all exactly the same. But at some voltage around that figure, the +12v supply will suddenly spring into life. It will probably keep on going, because its receiving its power from the large electrolytic across the solar panels, but if the solar conditions are truly marginal, it might conceivably pull down the electrolytic to perhaps +30v, where the small supply fails and turns off. It may wink on and off a few times before it has sufficient power to keep going at dawn. The voltage sensed by the LM555 will still be way below what is required to turn on the mosfet, which will only turn on when the large electrolytic hits the high voltage upper trigger threshold. It will then do a full high power dump into the heating element, until the voltage has quickly ramped down to the low voltage threshold. If conditions are truly marginal, the power dump into the heating element will all come from only the stored energy in the electrolytic. In a clear blue sky, the constant current from the solar panels adds directly to the stored energy, so each energy dump will be correspondingly more powerful with a higher current pulse into the load. The LM555 has two internal voltage comparators plus a flip flop which ensures very positive switching between two voltage thresholds. Exactly what we need. Just a simple single op amp like an LM358 is a much more tricky beast, especially if the dc supply to it is variable. A much better way to build a window detector would be to use each half of the LM358 as a separate voltage comparator, and then use the outputs to toggle the set/reset inputs of a flip flop. https://en.wikipedia.org/wiki/Window_detector Its much simpler to just use an LM555 which has all of that circuitry already there. Cheers, Tony. |
||||
| Davo99 Guru Joined: 03/06/2019 Location: AustraliaPosts: 1577 |
Go pro is set up in front window trained on front verandah with letterbox in background. Wifi Booster tuned in and set up to relay signal. Tablet in office getting good reception and hooked to charger for constant use. There is no missing that mailman or courier bringing those boards now! They cannot escape Detection!! :0) Trip to fathers has been put back. Doesen't want me up there. Well, said no point coming while it's raining at least. Might go up there later in the week or next week. Overcast weather here, perfect for testing. Interesting to see how long all my china orders take. I noticed on Fleabay delivery times are being all pushed out to months as in some being 3 or more. The pessimistic side of me says this could be a great rort for sellers. Push delivery times right out and if anyone doesn't get their stuff, they say we specified longer delivery times due to world wide shut downs. 3 Months later when it still hasn't arrived because it was never sent, time has elapsed to leave feedback so seller takes the money and is home free. I'll guarantee PLENTY have thought of that already and ARE probably taking advantage of it. |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Stuff is still arriving from China in drips, I received two small deliveries of components last week. Cheers, Tony. |
||||
| Davo99 Guru Joined: 03/06/2019 Location: AustraliaPosts: 1577 |
Ah well, no Christmas box today. Something to look forward to for tomorrow. All I got was a Voltage converter I can't remember why I ordered it now..... Think it was for a battery charger for Dad. Ordered on 19th of feb so not slower than it would be any other time. Ordered many things from this supplier and a month is about standard delivery time. |
||||
plover Guru Joined: 18/04/2013 Location: AustraliaPosts: 302 |
I wonder if anybody has the same experience with https://jlcpcb.com/ when ordering pcb's I am trying to order 10 boards (warpspeeds, of course) and all goes well until shipping cost, which initially is about $US 7 about I think it was so my total order would be about $12 US delivered. For some reason which I do not understand shipping ends up being US$45 Had 3 emails from support just telling me it is the shipping company that has decided I am in remote area, surcharge $ 26 US and add the transport around $ 18 US The suggestion is that I try another shipping address. Has anybody had that experience. Something new blamed on Corona Virus? |
||||
| Page 5 of 8 |
|||||
 Print this page |
