The Otherpower discussion board

Homebrewed Electricity => Solar => Topic started by: george65 on October 16, 2017, 03:25:55 PM

Title: Solar Panels to Direct electric water heating.
Post by: george65 on October 16, 2017, 03:25:55 PM

Been researching this and while there is a bit on the subject, can't find anything relating to what I CAN or want to do so I'll see what the brain bank can advise.

I have 8 spare panels for the time being I have no use for. 190W/ 36V Pmax.  Also getting an old but serviceable water heater i'd like to experiment with.
I would like to couple the panels in Series to make wiring easier and hook them to the heater.

YES, I know it's not efficient and I DO NOT care. That is not the purpose of the exercise, low cost and simplicity using what I have is.  For the time being I'm looking at this just  to heat up an outdoor shower that won't be used every day.... although with summer coming up and if it works, it just might be and if that's the case I'd look at putting it on the house heater.

I have also seen different controllers and read that one can make a PWM very cheap to make things much more efficient however I cannot find a schematic or code to do that. If anyone has links, I'd be very interested to see it and have a crack at building one but for the moment.....

Rather than the usual thing I have read of people wanting to run the heater element at lower voltage, I have enough panels to drive the thing at over voltage at least on specs.  Not sure what the panels will pull down to dead loaded but I'm thinking it will still come it at or above the elements rated 240V.
Running the element a bit higher than rated shouldn't hurt either.

The main problem I see is the thermostat and here's where the questions start.....

I would like to use the built in thermostat to switch a DC SSR .  Hard to find them in the voltage I _think_ I need at a reasonable price. Found a 400V one for $77. Not going to happen.  Can find them at 220V for a reasonable price but I'm wondering how sensitive they may be to voltage? There would be heaps of headroom on the amperage but the voltage could go well over. Would this be a problem?

Next thought was to use AC SSR's that would handle the Volts and amps easily  but I am aware of the DC arcing.  On my solar panel setup the disconnect is across both the Pos and neg lines so they cut at the same time to stop the welding flash.  Would using an AC rated SSR on each side of the inputs and controlled together stop the arcing ( or whatever will kill an AC SSR with DC) and let the things work reliably with decent longevity?

Even though the current will be will down with this setup, The time factor should be good and coming up to summer the daylight hours will be long.
Again, efficiency is not the object here and I understand the losses may be significant but for the time being I have no use for the panels which are mis matched tot he rest of my system so till I work out if I can add 190W panels to a 250W array without the need for another inverter, this is a better use than them sitting round.

I have filled my other goal with an array to back feed my analogue meters. I took all the loads I could off the smart meter and put them on the spinny meters and am backfeeding the solar. Putting over 150KwH back in a month so far. Don't know our consumption having just moved in but I am getting 20 KWH+ per day on clear days and seem to be hitting 16 Kwh with amazing consistency on partially cloudy ones which is much more than I would have expected.  The only thing I can't backfeed is the smart meter that has one leg of the 3 phase AC and the off peak hot water. If I could solar feed  the water heater even 3 months of the year, that would be a saving.  For the house I could look at automating the DC/Ac heating switchover for when there is insufficient sun for the solar once I have the basic setup working.

I would be very interested in building a controller if I could get the info and it's simple enough for ME to build or if someone will build me one for a reasonable price. :0)

Title: Re: Solar Panels to Direct electric water heating.
Post by: frackers on October 16, 2017, 04:47:05 PM
A bunch of MOSFETs is the way I would go, using the existing thermostat to put the correct volts onto the gates of the MOSFETs. Something like these I got a while ago for controlling solar input to a GTI to control its output on an OZInverter. (
Assuming a switch on at 100volts, and based on 10volts gate-source voltage to turn on the FETs then a 10:1 ratio of resistors (R1/R2) from drain-gate-source via the thermostat contacts with a protection 15v zener from gate to source.

This is straight off the top of my head but a couple of those high voltage FETs on a heat sink should do the job thus:

----------------- solar +
   |             |
   |            R1
heater           |
element          |
   |             |
   |          thermostat
   --|           |
FET  |---------------------
   --|           |          |
   |            R2       zener
   |             |          |
-------------------------------- solar -

Title: Re: Solar Panels to Direct electric water heating.
Post by: Mary B on October 16, 2017, 04:51:27 PM
40 amp 220 volt DC
Title: Re: Solar Panels to Direct electric water heating.
Post by: OperaHouse on October 16, 2017, 06:48:29 PM
I use a similar circuit in my dishwasher.  Just make sure you have at least 10V on the gate to get the stated on resistance.  FET are cheap and I always put at least two in parallel for very low on resistance and no heating.

Figure out what series power point voltage you want and compare what the amps would be with your heating element.  Your panels should be capable of providing at least 2.5 times that current.  While not efficient, it will be effective.

Not a big fan of those FET modules.  People report they need a lot of heat sink.
Title: Re: Solar Panels to Direct electric water heating.
Post by: george65 on October 17, 2017, 01:48:40 AM

Tried a test this afternoon with the panels and an old Urn.  Open circuit on the panels was 336V. Dropping the urn on them direct, the voltage came down to 80.  Admittedly the panels weren't at full output but it was a learning point in how much the panels pull down like this. I measured the ampreage at 2.4A so the urn was getting about 200W.  Going to take a long time to heat that up at that rate, a HWS is going to be useless I think.

I expected some voltage drop but not that much. I'll try again tomorrow but clearly the fall off makes this even more inefficient than I realised.

With this lesson learned I revise my initial thoughts and look for a cheap way to raise the efficiency of connecting the panels.
I keep reading of a PWM setup but not sure how this works. Is it to supply limited power but allow the panels to get up on their curve for decent output?
I also read about putting caps in the supply. I take it this is to let them charge up in between pulses then drop that into the element in the on time?

I can find PWM boards that do 220V but wondering if the 300V+ open circuit will cook them? Given they are not cheap like the little AC ones, I would prefer to ask the question and potentially look more stupid than get one and blow it because of obvious ignorance.

Also pondered the idea of using a cheap low voltage PWM controller to trigger mosfets as mentioned above given their higher voltage and amperage ratings.

I keep reading of these cheap and easy controllers for this purpose but buggered if I can find the schematics to build one.
Amazing how damn difficult these simple things can be when you are not an electronics engineer.

Title: Re: Solar Panels to Direct electric water heating.
Post by: OperaHouse on October 17, 2017, 05:10:37 AM
I just came back from camp for the season and thought I was coming back to the good house with no problems.  Be more than a week before I can deal with other peoples issues.

This is fairly simple.  You need:

10-18V electronic wall wart
a couple 50A or more 400V FET
uno or nano
1K ohm resistor
three 270K-330K resistors
10K pot
opto isolator
half dozen silicon diodes
Dozen 220-470uF 400V electrolytic caps
some other miscellaneous parts from an old VCR or TV

The capacitor bank is the key to storing energy.  You may not want to build it but anything smaller will not last.  Gather up your stuff and tell me what you got.
Title: Re: Solar Panels to Direct electric water heating.
Post by: george65 on October 19, 2017, 04:29:24 PM

The caps are going to be very exy. Anyone have suggestions where they can be had from for the right price?
I have NO electronic scrap having got rid of every bit of junk I could when I recently moved.  Is there any particular type of electronic products that I may be able to liberate suitable caps from?  I did have a HEAP of removable server power supply's, might have been a candidate but all gone now.
Anything I didn't need was disposed of. Brother in law knew a very nerdy but probably brilliant guy who was interested in all the electronic scrap I had. His eyes glazed over when I loaded him up with brand new UPS units, servers, desktops, routers and all manner of stuff. All of it was good, Just had no use for it or too much already.

I'm having a re think about wiring the panels for 24V to a couple of small batteries through a UPS and a PWM like I did before. Might be the easier way but then again I know that and there is no learning in it. I'll see what I can figure out with a DIY solution for the moment.

Anyone know if I can use a DC PWM board off fleabay with the caps to make a suitable controller?  I'm thinking the caps are placed in front of the board to store the energy on the off cycle and discharge it in the millisecond the thing switches on.

I did try the direct coupled urn again yesterday. Switched it off at 9:30 am before I left the house as it was already boiling. With a controller there may be an effective if not highly efficient amount of heat to be gained here.

I'm seeing things on YT and other sites about having a pre heater for the hot water which is solar powered. This is an idea I had some time back but only get see others are now thinking of it/ setting it up as well.  If the water got up to temp witht he solar, I'd add a 2 way valve to the plumbing so I could switch the main heater off altogether during the sunny weather and run on solar alone.
Title: Re: Solar Panels to Direct electric water heating.
Post by: joestue on October 19, 2017, 06:21:18 PM
Anyone know if I can use a DC PWM board off fleabay with the caps to make a suitable controller?  I'm thinking the caps are placed in front of the board to store the energy on the off cycle and discharge it in the millisecond the thing switches on.

capacitors in front of the pwm switch, an inductor after, and you have a buck converter.

if you do not have an inductor, then the moment the switch turns on the current rises very quickly to as high as it needs to go, in order to satisfy the laws of physics. often the ESR of the capacitor, the ESR of the battery, and the inductance of the cabling is the only thing limiting that current.

so because there's an inductor you also need a diode.. the inductance between the capacitor, diode, and switch needs to be as low as possible.
Title: Re: Solar Panels to Direct electric water heating.
Post by: tanner0441 on October 20, 2017, 06:52:40 AM

A source of high voltage high value caps, is dead switch mode power supplies. they normally fail on diodes or chips. If the caps go they usually do it in a most spectacular way that leaves you in no doubt.

Title: Re: Solar Panels to Direct electric water heating.
Post by: OperaHouse on October 20, 2017, 02:39:56 PM
That is the trouble with living in paradise, no parts.

Saw this interesting video on youtube on MPPT water heating.

And you think I can't explain things!
Title: Re: Solar Panels to Direct electric water heating.
Post by: george65 on October 21, 2017, 02:46:37 AM

Yeah, that's one packed full of info Vid!
Love your comment on it! 
When you don't know, pretty sure the rest of us are well and truly stuffed!   ;D
Title: Re: Solar Panels to Direct electric water heating.
Post by: george65 on October 23, 2017, 06:59:12 AM

I have been playing with the panels some more and learning a lot.

I found my old urn I have been testing with is 1800W which is half the element in the HWS and a size I can get for it.
In bright but slightly hazy sunlight I got up to 212V with the urn connected.  This was near as dammit full output and had the urn boiling as it would on mains power.  While observing this, some cloud came over and the voltage dropped to 30.  The boiling stopped being a roll and the element was probably only replacing heat loss.
Clearly the direct coupling would be fine with the 8 panels in good sunlight but the ramp up in the morning and fade off in the afternoon would be pretty poor.

I started wondering how much power I needed for my water heater. If I needed to heat the 250L of water my tank holds from 50 to 70 as I guesstimated on average, I'm going to need 6KW. If I need to heat from 40, it's going to be 9Kw. As I have around 1500W in the array, I clearly wasn't going to get 6 kilo out of it directly coupled but I could get that and maybe a fraction more when I had the panels on an inverter.
As I am on grid and my main array is to wind the meter back I thought the cheapest, easiest and simplest thing is to hook the array to an inverter and run the heater off the normal mains which I am backfeeding into rather than the offpeak I can't rewind on the digital meter.

Then another thought occurred..... the normal power costs me .30C Kwh. Off peak is 11c.  As I am probably going to be using a bit more than I am generating once the warm weather comes and the AC kicks in, I'm financially better off, nearly 3 times better off,  to put all my generated energy into the normal circuits and just pay for the offpeak to heat the water.

That still left me with 5Kw of  250W panels and 1.5 Kw of 190W panels.  Very different open and loaded voltages, amps etc.
I looked up mismatched arrays but could only find info on mis matched panels within and array, not different arrays or strings.  I worked out that the loaded power point for the 190 array of 8 panels was 293V and the 10X 250w panels was 306V. Close enough I thought.
Figured I had nothing to loose trying  to parallel them up so I grabbed some cable and hooked them together.

The 250W panel array was doing about 1500W output. The 190 panel array was doing about 880. I hooked them together in parallel and maxed out my 2kw inverter which was really pleasing to see.  I tried swapping the 2Kw inverter with teh 3kilo one I have and it showed about 1880W output.
Lot of possible reasons for this comes to mind but I'll wait and see what I get tomorrow in hopefully better light.  Even a 500W gain is probably going to average out better than directly hooking the panels to the HWS.

It will be interesting to see tomorrow what the 3Kw unit maxes out at but I think there is a gotcha in the plan.  The 3Kw inverter has dual inputs. I am pretty sure that neither input will take the max load. I seems to remember they overlap, as in you can put 2Kw on each side but not 3 Kw.
I'll see what sort of output I get tomorrow.

I'm also thinking I can probably offload these 190w panels for the same price I can wheel and deal 250's so I'm probably better off getting rid of them and getting better matched panels and being done with it.  That way I can get max output from them in any config and it will be much easier to incorporate them in with the other panels.

As I'm on grid,  the cheapest way for me to do anything with solar is to just backfeed my meters with what I generate and run whatever I want from there. I have inverters and panels are always going to be cheaper and easier than more controllers and setups. Not as much fun or as interesting but efficient from both a power and cost POV.
For the moment I'll backfeed all I can and just pay for the offpeak power to heat the water. If I start getting into credit on the meter's, I'll put the hot water onto the normal power and use that.  I can see where possibly spring and autum when I don't use the AC much I might be able to run the water heater off the mains and then in winter and summer when I am using more amps I can run it off the offpeak. Or.......
I can put another few KW of panels up and loose the off peak all together.

See how I go when it's properly set up and running but I am having fun and learning a lot with this along the way.   ;D
Title: Re: Solar Panels to Direct electric water heating.
Post by: frackers on October 24, 2017, 04:30:31 AM
I'm surprised nobody has mentioned OpenEnergyMonitor yet. (
This can divert power to your water heater while the solar is exporting and balances the in/out to try and make the net amount sum to zero so you use the power as you generate it without it being registered by the meter.
 Couple of the guys at the local robotics group have built an Arduino based design from this site and they just work out of the box!!
Title: Re: Solar Panels to Direct electric water heating.
Post by: DamonHD on October 24, 2017, 06:47:49 AM
The OEM guys (and their stuff) are good.

Plenty of consumer-grade diverters out there too.


Title: Re: Solar Panels to Direct electric water heating.
Post by: george65 on October 24, 2017, 06:51:28 AM
Thanks Again Frackers!

I saw that before and have been trying to find it since but every search I did came up with something else, mainly expensive commercial units.
it's all in the terms you use and frequently, I haven't got a clue what to call things or they are nothing like what I would call them.
I'll look that over some more when I'm less tired so I can take it in.

That is what I need for the Digitally metered phase.  Just put in enough to balance consumption and nothing more. A friend tells me there are plans afoot to remove all analouge meters within 4 years here and replace them with smartarse meters.  That's to help people save money on their bills of course, nothing to do with making the power companies yet more profit still. People really have nothing better to do with their time than get out their phones and see how much power they have used at any time of the day or night.  Yeah right!  ::)

I tried the mismatched arrays today and they seemed to work fine.  My inverter seemed maxed out at 1880W all day. I take it was maxed because it was still doing the same output fairly late into the afternoon where the light levels and sun direction were way off optimum.
This is a 2 input inverter, an aurora 3KW. I have taken out one side with over voltage from my IMAG I was playing with a while back.l Id really like to have this thing working properly. Wondering if it might just be a fuse I could change or something I could have repaired that wouldn't cost more than the unit was worth.

I have a new Bosch inverter but remembered today when I dug it out I didn't get the AC connector with it. I have emailed Bosch to see if they or an agent can supply one although I bet the cost will be " inspiring". Might see if I can get it apart and just remove the connector all together and attach some mains cable. Decent mains cable.

Despite the clipping inverter, I managed a new high of 28KwH today which did surprise me. Was fairly hazy and overcast most of the day and I was cursing because there was neither rain nor clear sunshine.  Don't know what it is with hazy days but it seems to give real good outputs on the solar. Either the haze improves some part of the spectrum the panels like or they gain from running cooler or something.
I think I'm still missing 3-500W with the inverter limitations so it may be possible to get 30Kwh out of what I have now which would be great.  Still tempted to get some more panels though.  Once the heat comes and the AC goes on, I can easy see the consumption going up a LOT. 
Winter will be much the same and the solar input less.

I can get another 3Kw on the west facing Verandah roof which gets sun by 9 in the morning and full sun from midday on. Even another 10KWH out of that array would be real worth while.  I hope to achieve the goal of having to turn an array off so I don't go backwards on the meter at some point from the last reading.  :0)

Wonder if something like the monitor could be set to measure your total in and out and allow back feeding to catch up with night usage and  keep the ratio neutral?
THAT would be ideal!