Inverter for 48v off grid - PIP xxxxx or something else?

[paulvk]

Could I put a filter on the 240vac supply

If I was building a system now I would put MOV voltage units on the system in a number of locations
probably where I have circuit breakers, I have a pack of 275v units and will incorporate them in the future
The other problem people have had is the 63v capacitors (which have been upgraded in later units)
I have not had problems I think because I try to keep the temps below 40c by adding external fans
and I do hardly any AC charging this in particular would stress them.

[Revlac]

As yet I have never had an issue running brush motors or induction motors, within respect to its limits.
2400w angle grinders with soft start have run for years, the metal drop saw I use an SCR speed control just to start it off, vacuum cleaners 1000w direct start and 2400w soft start, water pumps 1100w, the 900w bore pump runs all day through 120m of 15amp cable, had sand get into the pump and resulted in a locked rotor for 15 minuets, the inverter just kept going, this happened several times until I fixed the impeller Clarence problem and a better circuit breaker.
Lots of different appliances and configurations so results will likely be different to others.
Have had one blowup and repaired it had nothing to do with running anything mostly due to a faulty switch/connection on the battery side.
I have Low Frequency inverter and it likes the direct start motors more than the soft start type.


As for the PCM60X MPPT, 2 out of 3 of them are running perfectly, the last one I bought at the time when MPPsolar were changing there address (new location) They said they where updating the firmware on it and it would be about 3 weeks.
It now has a voltage reading up to 2 volts different than the inverter, put up with this for maybe 2 years until I decided to look inside, found the inductor terminal where it connects to the board was a little loose and had been very hot, melted the shrink wrap, fixed that.

Connected the unit to a small 48 volt battery without the sensor connected (I think) then reset the controller to the default auto voltage detect (voltage still reading wrong), disconnected then connected it up to the main battery bank with the sense wire connected left it to power up the next morning.
It was running perfect for a week until I changed something and now its voltage is drifting out again, still not sure what the problem is.

Cheers Aaron

[Honu]

I strongly don't recommend connecting 7.2 kW of panels to an inverter (any brand or model) that is rated for 4 kW of PV. I know it will only use what it needs in theory, but the Solar Charge Controllers assume rated panel power (or a little higher), and can overshoot and undershoot quite badly with severly "overclocked" PV. When the Pylontechs disconnect due to over-voltage, unusual and sometimes bad things can happen.

HOOo i did not know that, noted ... i will then use a separate MPPT to deal with the 3.2kW remaining panels.

Also, the bifacial panels are quite new; I would call them an untested technology. Do you intend to mount them vertically to get east-west sun, or conventionally on a roof? I've not heard of the former before, and would not know how to mount them safely. The 6 mm thickness modules seem fragile to me (perhaps you are considering the ones with the frames, but those are only 10 mm wide at the top and bottom, and 30 mm wide at the sides). I note that I could not see any 300 W models; the lowest models I saw were 305 W (unframed) and 310 W (framed).

I said 300W, but they are in fact 310-330W : https://www.jasolar.com/uploadfile/2020 ... 344322.pdf
I will mount them on the ground, vertical 55°, from south 0°, i plan on perhaps creating a kind of "box" around them to preheat the air that come into my house. Box from which the top will be opened in summer.

- 10kwh of Pylontech (3xUS3000) (those already got their BMS : http://www.pylontech.com.cn/pro_detail. ... 121&cid=23)

Those will be able to be charged at 37 x 3 = 111 A, which should be able to take the full power of your PV array (nearly 12 A per string, or some 96 A).

Yea that what i've calculated too, i've got little room there, which is good. 10-12A per string, 8 strings.

I will perhaps add a "SCC MPPT 3kW" to cover what the PIP can't

I would recommends this (or any other external MPPT). Or a second PIP in parallel to take half the PV strings (see below).

Not sure about the second PIP .. but .. i keep the idea since they do not cost that much more then ther MPPT alone and it could replace my first inverter in case of problem.

By "transformer", I assume you mean a DC-DC converter, since your 48 V is presumably DC. This is an attractive idea in theory, but consider the problems:

• A completely separate wiring system, which should be physically separate from the AC system (not sharing the same conduits or channels, for example).
• The difficulty of switching 48 VDC at significant currents. Ordinary AC switches will not be suitable, as they will arc. I get away with AC switches in my house for a few lighting loads, because I use 24 V (makes a huge difference), and the loads are very low power (< 1 A). 95% of my lights, and 100% of my power, are 230 (or 240) VAC. The 24 VDC system is a legacy that I keep as an ultimate backup in case two other systems fail, or something trips the house Residual Current Device.
• The difficulty of fusing DC (fuses are larger and way more expensive)
• 48 V cables need to be some 5 times thicker for the same power level.

Yea sorry, i meant DC-DC converters, i do not plan on using more then 1000W on my 48V circuit, something like 20A.
My water pump will be 1m from the battery pack/inverter (in an underground concrète shed, just under the panels, some tubes are running under ground and are cooling the system, convection act like a natural pump), seems 2.5mm2 is enough. 4 meters from this inverter/battery you draw a circle of 5 meters in diameter and all my electric devices (appart from lighting) will be in it => My circuits are pretty short, i can use 2.5mm2 or even 4mm2 if needed for the 48V. In fact i want to be able to use my important devices even if my inverter fail.

I'm using this table for cable sizes :

The goal is to switch the inverter off each night automaticaly, cause those 30W are way too much in my opinion and loss in transforming also add up while my 48V-12V transformers kind lose near nothing.

I note that if you have a charging source (PV all day, or you have utility AC-in at night), then PIP inverters don't switch completely off, even if you turn off the switch. If you don't have AC-in connected, then leaving the switch off will cause the inverter to switch off at sunset, but that's presumably not what you want. Turning off 230 VAC power at night means every clock in the house will lose power (apart from battery powered ones, of course). Personally, I find the clock in the microwave to be very useful.

Yea you are right, i do not want it to switch off at sunset, i want to be able to watch a film or something like that then switch it off when going to bed.
But .. for what i've read, the PIP 5048 got 3 modes : ON under load, ON without load (45W) and idle mode what i call sleep mode which is 15W .. m i right on the values here ? Cause 15W is perfectly acceptable. And if i remember correctly, the peep wake up from idle mode when the "demand" rise over some watts... isn't it ?

This is another idea that is attractive in theory, but difficult in practice. You'll need several DC rated contactors, and if you connect to the panel side, it's ~100 VDC, so those are different elements than ones suited for 50 V. Resistive elements draw 4x the power at 2x the voltage; you very likely can't overpower the elements by 300%. You can't ue the ordinary thermostat; it's only rated for AC, so you'll need a special thermostat, or some other sensor, and then there are safety concerns. It likely can be done, but it's very tricky.

What about connecting the resistive elements on the charge port of the pip ? WOuld be something like 35A with all 3 elements on.

My questions : What is the best voltronic / Pip to buy today, i need 5kW, i think i will not need parrallel capability cause 5kw is already plenty for me, and this would double the sleeping energy cost.

I thought that 4 kW would be enough for me (one PIP-4048MS). But now I'm running two in parallel. The main thing I didn't cater for is charging an electric car. You may not have one now, but in 5-10 years, that could well change. So I'd at least get a parallelable model, even if you go with only one inverter for now.

The 450 V or 500 V max SCC models make it easier to wire the solar panels, that's all. PIPs don't have insulation monitoring, so the higher voltage models aren't legal in Australia; I don't know about other countries. There is no patched firmware for the higher voltage SCC models, although there may be one or two lightly patched versions (just fixing the premature float bugs) in future.

Ok you bought me on this one, we already got electric cars, since 1997, i'll go with the 145dc parralel compatible device then..
If i need to add more panels in the futur.. i'll be ready.

One last question, for people that already tried both, would a victron system be really more efficient in terms of sleep energy cost, in term of inverter efficiency, in term of Mppt reaction time.. voltage start up .. ?

Total : 2900€ ... vs a PIP 5048MG that would cost me 650€... that a quarter of the price .. !

Wow. I haven't checked the price differential for years; it was something like that ratio back then. I thought the newish Multiplus II was supposed to be much more competitive.

Perhaps there a way to get it cheaper, like 1400€ .. but i'll still need a 4kW MPPT to match the PIP posibilities.

[Honu]

I would not be running 48v for lighting even with the loss in the inverters 230v led lighting is much more efficient.
I have compared 12v led lighting to 230v and I can light a whole room brightly with two 10 watt 230v down lights
I could use 40 watts of 12v and not get it that bright
I have 48v DC gear running off my system modem , ether net switch , wifi access point, 48v soldering irons and a number of other things
but these were there long before solar.
For hot water you can get 48v elements but I have a 3600watt 230v element that is two 1800watt in parallel
so it can be 3600 , 1800 , 600 watts with relay switching
Others have used SCR solid state relays to pulse width modulate the element.
Some others turn off one PIP during the night to save power.
Victron is a good brand but as you see 4 times as expensive you can put the extra money into more batteries and panels.
I run two homes with parallel PIPs and purchased 6 units to have two spares nearly 6 years and they are still spares.
There is one type of motor I do not run with my PIPs that is brushed AC motors, vacuum cleaners and power tools
these type of motors put voltage spikes on their power source and as the PIPs are transformer-less
the out put semiconductors are directly exposed but with battery power tools and vacuum cleaners so common
I do not use the others much now.

Hello, most led (recent) deliver 100 lumens per watt, in fact AC got certainly more loss due to the switching power converter in their "220V" lamps. Led only work on DC ... well they can work on AC but only on half the cycle.. they are LED (aka diode).
I got a flashlight that work on 3.7V .. and it deliver 1000 lumens, i can say it's pretty bright for a so small thingy.
Certainly in your test you do not have the same kind of led, not the same density, but surely something is different.
Psychologically i can't convert my PV DC .. TO AC .. THEN DC again to power my lights .. i mean, it's non sense to me.

Yes i can put on more batteries, you are right .. a 3.5kwh US3000 cost like 1400€. It's still cheaper then a victron solution, and those 50W*10 hours.. (let's say 10) = 500Wh .. are pretty easily taken.
Good to know PIP last more then 6 years when properly used.. !
But yea .. i got some power tools for wood work mainly, they are for most of them under 2000W.

[coulomb]

But .. for what i've read, the PIP 5048 got 3 modes : ON under load, ON without load (45W) and idle mode what i call sleep mode which is 15W .. m i right on the values here ? Cause 15W is perfectly acceptable. And if i remember correctly, the peep wake up from idle mode when the "demand" rise over some watts... isn't it ?

Sorry, I have zero experience with power saving mode.

What about connecting the resistive elements on the charge port of the pip ? WOuld be something like 35A with all 3 elements on.

I don't know what you mean by the "charge port of the PIP". If you mean the solar panels, the problem is that the voltage is quite variable, and you still can't exceed the element's maximum power rating (at least, not for very long) without endangering its service life. And it's still DC; switching 35 ADC is not trivial.

[Honu]

But .. for what i've read, the PIP 5048 got 3 modes : ON under load, ON without load (45W) and idle mode what i call sleep mode which is 15W .. m i right on the values here ? Cause 15W is perfectly acceptable. And if i remember correctly, the peep wake up from idle mode when the "demand" rise over some watts... isn't it ?

If someone know about it .. i'll take all informations you got ..

I don't know what you mean by the "charge port of the PIP". If you mean the solar panels, the problem is that the voltage is quite variable, and you still can't exceed the element's maximum power rating (at least, not for very long) without endangering its service life. And it's still DC; switching 35?ADC is not trivial.

I meant the charging port, the one that is connected to the batteries. I would connect my resistive element in parrallel to the batteries, with an SSR to be able to disable those elements as soon as i detect a consumption on the shunt of the batteries, i will make a sketch to explain it more precisely.

I was thinking of something like this :


I've read somewhere that PIP got a contact being closed when the battery is fully charged, that contact would switch on my SSRs... when battery lose some volts.. it should open this contact and then subsequently open all SSRs.

Question is .. how does this work .. PIP send current to those batteries, than when correct voltage is met it stop sending this current .. right ..? Does this mean i can't draw current on this charge line without sucking the batteries ? At least when they are 100%¨state of charge ? Do i have to set the battery BMS to stop charging let's say at 49V while i tweak the charger to stop at 50V ..? This way it would still send current to this charge line even when batteries would be full ..? I'm kind of lost here on how it work.

In fact i would like to avoid that my batteries do the YoYo .. charge /discharge between 95% and 100% state of charge.... is it possible ?

Why i got 3 SSR ? Cause the idea is to close them depending on what is available.. while not using energy stored into the battery, i was thinking on piloting them with an arduino, i got Shunt info (charging/discharging battery) and the battery voltage (fully charged ?)

It certainly looks pretty messy ..

[T1 Terry]

I would more likely use the direct solar feed to the heater element loads. By diverting some of the solar once the end of bulk charging is reached the solar current to the battery is reduced stopping the "bounce" excessive solar can cause because the battery voltage jumps up suddenly rather than a slower creep up to the 100% SOC point. You could select battery voltages to turn on each solar diversion load, 54v on for the first and 52v off, 56v on and 54v off for the second and 58v on and 56v off for the 3rd. You could shift all of these voltage down by 1v if the end load of 3.6v per cell balanced is too high for you.

T1 Terry

[Honu]

I would more likely use the direct solar feed to the heater element loads. By diverting some of the solar once the end of bulk charging is reached the solar current to the battery is reduced stopping the "bounce" excessive solar can cause because the battery voltage jumps up suddenly rather than a slower creep up to the 100% SOC point. You could select battery voltages to turn on each solar diversion load, 54v on for the first and 52v off, 56v on and 54v off for the second and 58v on and 56v off for the 3rd. You could shift all of these voltage down by 1v if the end load of 3.6v per cell balanced is too high for you.

T1 Terry

Hello,
Yes but it's not easy, you say "some of the solar"... yea .. how much..? And in case you suddenly ask for power on the AC side of the inverter, there need to connect those PV back to the inverter. I've turned this many ways and it seems to me that the charger line is the best way, cause we can react pretty fast, cause we take advantage of the MPPT, cause the cables here are not very long, between inverter, batteries and the resistive element.. (well .. depen on your house for the element)while to be able to connect disconnect PV we would need some long cables, at least to be able to do something that "adapt" to the insulation. What do you think ?

The MPPT and converter are pretty efficient (something like 97%) and while it put some wear on the condensators here ... i feel that replacing them in 10 years or so is not that big of a deal. Plus you keep the benefit of the MPPT .. ! Which give you the most of your panels while a direct connect to the element will lose a lot of energy, in fact in most situation it will certainly be bad. But perhaps someone that really tested those things can give us some advice...


The problem in my schematics is that... i think, but i'm pretty bad with electricity, that the battery pack internal resistance is far lower then the resistive element (by design it is pretty high on this one) so the battery would always be "under charge" if current was to be sent on this parrallel circuit.. which is pretty bad no ?
Would it be... i know .. just be gentle with me, this is not my scope of expertise.. , could i put diodes in there to prevent the current to flow TO the battery when i want to use my elements ?

Does this solve the problem ? Everytime i want to heat water, i open SSR0, then close SSR1, 2 and/or 3 ? Then when i want to charge again, i open SSR1, 2 and 3 then close SSR0..
The diode is here in case the inverter suddenly need power to feed the ac outlet. Cause i do not think my arduino and SSR0 would be fast enough to react.



PS : i know that what i've written is certainly dumb... so .. be free to criticise it. (and apologize for my english which is far from perfect too)

[solamahn]

My existing hws 415L has 2 4800w elements. Previously the top element had off peak rate grid power connected via the top thermostat continuously and the bottom element had off peak rate timed grid power connected via the bottom thermostat. I connected both elements in series to give a total load of 2,400w via the top thermostat. Manual control can be made using a circuit breaker mounted inside the house only or automatically with the addition of a contactor controlled via the dry contact and prog 12 and 13.

[Honu]

My existing hws 415L has 2 4800w elements. Previously the top element had off peak rate grid power connected via the top thermostat continuously and the bottom element had off peak rate timed grid power connected via the bottom thermostat. I connected both elements in series to give a total load of 2,400w via the top thermostat. Manual control can be made using a circuit breaker mounted inside the house only or automatically with the addition of a contactor controlled via the dry contact and prog 12 and 13.

I do not know if i understand all, you connected 2x2400W elements to your inverter is that right ? On the AC side ? And you have a manual breaker to use it .. right ?

[solamahn]

I have connected the 2 existing 4800w elements in series. This gives a total of 2400w load. The reason I did this is because 4800w for 1 appliance is a bit much for 2 x 4048V inverters. There would only be 3200w left out of the 8000w total available. Yes on the ac side. The modification to the HWS is easy. Just move a wire from one terminal to another terminal.

[T1 Terry]

I would more likely use the direct solar feed to the heater element loads. By diverting some of the solar once the end of bulk charging is reached the solar current to the battery is reduced stopping the "bounce" excessive solar can cause because the battery voltage jumps up suddenly rather than a slower creep up to the 100% SOC point. You could select battery voltages to turn on each solar diversion load, 54v on for the first and 52v off, 56v on and 54v off for the second and 58v on and 56v off for the 3rd. You could shift all of these voltage down by 1v if the end load of 3.6v per cell balanced is too high for you.

T1 Terry

Hello,
Yes but it's not easy, you say "some of the solar"... yea .. how much..? And in case you suddenly ask for power on the AC side of the inverter, there need to connect those PV back to the inverter. I've turned this many ways and it seems to me that the charger line is the best way, cause we can react pretty fast, cause we take advantage of the MPPT, cause the cables here are not very long, between inverter, batteries and the resistive element.. (well .. depen on your house for the element)while to be able to connect disconnect PV we would need some long cables, at least to be able to do something that "adapt" to the insulation. What do you think ?

The MPPT and converter are pretty efficient (something like 97%) and while it put some wear on the condensators here ... i feel that replacing them in 10 years or so is not that big of a deal. Plus you keep the benefit of the MPPT .. ! Which give you the most of your panels while a direct connect to the element will lose a lot of energy, in fact in most situation it will certainly be bad. But perhaps someone that really tested those things can give us some advice...


The problem in my schematics is that... i think, but i'm pretty bad with electricity, that the battery pack internal resistance is far lower then the resistive element (by design it is pretty high on this one) so the battery would always be "under charge" if current was to be sent on this parrallel circuit.. which is pretty bad no ?
Would it be... i know .. just be gentle with me, this is not my scope of expertise.. , could i put diodes in there to prevent the current to flow TO the battery when i want to use my elements ?

Does this solve the problem ? Everytime i want to heat water, i open SSR0, then close SSR1, 2 and/or 3 ? Then when i want to charge again, i open SSR1, 2 and 3 then close SSR0..
The diode is here in case the inverter suddenly need power to feed the ac outlet. Cause i do not think my arduino and SSR0 would be fast enough to react.



PS : i know that what i've written is certainly dumb... so .. be free to criticise it. (and apologize for my english which is far from perfect too)

It depends what voltage you bring down from the solar to the MPPT controller. If it is up around the 140vdc mark you can switch that into an existing 240vac or what ever voltage element you have, as long as it's above the open circuit voltage of the solar raw feed to the MPPT controller.
If you bring the 140vdc open circuit down to the MPPT controller is multiple strings and connect each in parallel, you can use the battery voltage sensing system to switch one group at a time to the element. If a sudden load is applied, the battery voltage will drop below the diversion switch off point for each group so the solar now goes to the MPPT controller to supply the battery charging required to bring the battery voltage back up again.

T1 Terry

[Honu]

I have connected the 2 existing 4800w elements in series. This gives a total of 2400w load. The reason I did this is because 4800w for 1 appliance is a bit much for 2 x 4048V inverters. There would only be 3200w left out of the 8000w total available. Yes on the ac side. The modification to the HWS is easy. Just move a wire from one terminal to another terminal.

Ok that what i understood, tthose 2 elements in serie will draw half current so i get the idea, but i would really like to not connect on the AC side, cause... the pip is something like 90% effective (not amazing) and i would like to not wear it too fast, and heating water consuming so much power in the long run, i think it's better to connect it on the DC side. In my house water will only be heaten by PV, only some days in winter i will fire my wood water boiler. In fact i have a 800l water tank and it will be used for shower and to heat the house to, so .. lot of power will go in it. In fact i would like to harvest near all electrons from my PV and put them in my 800l water tank (23-90°C)



That's approximatively 60 kwh of energy stored in this 800l tank.

[solamahn]

You could just connect panels direct to the hws heating element. Say 8 x 300w 36v panels in series.

[coulomb]

You could just connect panels direct to the hws heating element.

I would think that you'd want some sort of over-temperature protection, even if in your part of the world it might rarely be needed. And the stock thermostat would only be AC rated.

[Honu]

It depends what voltage you bring down from the solar to the MPPT controller. If it is up around the 140vdc mark you can switch that into an existing 240vac or what ever voltage element you have, as long as it's above the open circuit voltage of the solar raw feed to the MPPT controller.
If you bring the 140vdc open circuit down to the MPPT controller is multiple strings and connect each in parallel, you can use the battery voltage sensing system to switch one group at a time to the element. If a sudden load is applied, the battery voltage will drop below the diversion switch off point for each group so the solar now goes to the MPPT controller to supply the battery charging required to bring the battery voltage back up again.

T1 Terry

Hi Terry,

If i understand what you said : My pvs are 41V open circuit, 32V max power => strings of 3 in series => 123V. I will have 8 strings (not all on the pip), so i could take 2 strings (6 panels).. connect those 2 strings in serie then connect to my resistive elements... would then be "Open voltage : 246V", "Power voltage : 6x32=192V". Then if a load occur, i would sense it, disconnect the serie SSR and reconnect the 2 strings to the MPPT.

A fast calculus give me that a 5.6 Ohm resistor would be good, 6 PV in 2 strings of 3 would give me 99V and max 18Amp

This small table give power meet by the 5.6 Ohm resistor depending on the Amp given by the panels.

Amp / Power
1 5.6
5 140
10 560
15 1260
18 1800

See how 18 Amp give 1800W while half that intensity, 10 Amp give only 560W ? => direct connect can't work under low irradiance, for half intensity it give a quarter of the power (the square responsible for that), i would need a 10 Ohm resistor to get 900W at 10 Amp ... but .. the resistor for a fixed internal resistance. Or i could serial them 2 of 5.6 would be ok... but it's impossible toi make it "linear".. => lot of losses.
An other option would be to have ... something like 10 elements with different resistance, and combine them to show the most efficient resistance to the PV array, would be like an MPPT anyway.

Direct connect is a bad idea.. cause Max power changes a lot with irradiance => can't work with a fixed resistive element.

As info, here are the delta for different cities, with MPPT .. without MPPT for a 2kW array, direct to element : (it's between 20 and 35% less effective)
PS : sorry it's an European document...



On this one you can see why a fixed resistive element is bad :
- Point 1 is the max power of the panels => resistive element should be 28 Ohm, now if irradiance go down to 200W/m2 .. you are on point 1 without MPPT ...(56W) and on point 3 with an MPPT (400W), those 400W could be achieved with a 169 Ohm resistor. AND a 169 Ohm resistor in case of 1000W/m2 of insulation will give something like 450W instead of 2000W (point 4)

[solamahn]

The way my hws is setup is good. Cheap, easy and it works.

[T1 Terry]

The method I described is not the most efficient, but it is waste solar anyway so what does it matter? It does work because I've used it on a number houseboats and off grid systems. For the parallel connected 12v PWM controlled set ups, we use a 24v element, not efficient but a high level of over voltage protection so the life isn't reduced even if the solar voltage was to reach open circuit output for some reason. The 24v systems we use a double 24v element and connect it in series. 3 of these would leave some head room for high voltage panel output and work effectively on 900w, yet the full 1800w could also be connected and the only effect would be the voltage increasing.
Easy enough to get a DC thermal control.

If you are looking for a simple bolt in set up, check out these units https://www.kickstarter.com/projects/el ... icty-and-h

T1 Terry

[Honu]

The method I described is not the most efficient, but it is waste solar anyway so what does it matter? It does work because I've used it on a number houseboats and off grid systems. For the parallel connected 12v PWM controlled set ups, we use a 24v element, not efficient but a high level of over voltage protection so the life isn't reduced even if the solar voltage was to reach open circuit output for some reason. The 24v systems we use a double 24v element and connect it in series. 3 of these would leave some head room for high voltage panel output and work effectively on 900w, yet the full 1800w could also be connected and the only effect would be the voltage increasing.
Easy enough to get a DC thermal control.

If you are looking for a simple bolt in set up, check out these units https://www.kickstarter.com/projects/el ... icty-and-h

T1 Terry

Hello,
Ho yea it work by directly connecting, but i need more then that cause my house only rely on PV as energy source, it need to be efficient. On a boat you even more need efficiency. The resistive element need a constant current and variable voltage From PV that generate constant voltage and variable current.

I know Dacian from Electrodacus, but his system got flaws that are not comaptible with what i need. There is a way to connect an inverter to his card but it do not accept the full power of all panels connected to this same board.. making it useless to me. Plus the think is pretty expensive if not used for what it is... it's basically a heater, Dacian use it so heat his own house through resistive wires.

[Honu]

The way my hws is setup is good. Cheap, easy and it works.

It works but it's not efficient, like my grand pa car work .. but it burn 10l per 100km while my BMW only use 6.2l. Efficiency is ... in my opinion, one of the most important thing in our world, people buying a sandwhich and throwing half of it .... not efficient, being efficient is being against wasting things. I know that in your case, it's not really wasting... but if your system was more efficient you could have saved some panels (20%)... same idea as the sandwich, of course the system to make this more efficient got to be less "expensive" in terms of money/material/labour then the panels saved. It's because humans are improving for milleniums that we are where we are.. (it's bad for some things though.. )
Making this more efficient is the way, more efficient without making it overcomplex or unreliable.

Ps : This is in no way a critic on your system which is already far more efficient then what most people do by not using they overgeneration of electricity, but i would like to go a little further.

[Honu]

Is there a device able to transform whatever Voltage x Amp (max 140V, max 40 Amp) to a linear Voltage x Amp .. ?
In short, a DC to DC convertor with variable input and variable output, output just have to be linear, this line will depend on the resistance of the element.

In this example, we got some MPPT for different insolation : the thing to do is to convert those different points 60W, 125 .. 315W and "move" them on the line .. this mean converting the 60W, 33V/1.8A point to the 13V/4.5A.
The "line" slope depend on the resistive element and got to be "tweakable" in the circuit.
As you can see on the graph, all cases are decreasing voltage to create some Amp, those arrows seems parralels .. perhaps dut to the panel itself, can't be sure it will be on every panels.
Panel is a Ja Solar 315W. (https://www.jasolar.com/uploadfile/2020 ... 344322.pdf)

[coulomb]

Is there a device able to transform whatever Voltage x Amp (max 140V, max 40 Amp) to a linear Voltage x Amp .. ?
In short, a DC to DC convertor with variable input and variable output, output just have to be linear, this line will depend on the resistance of the element.

It seems to me that an off-the-shelf MPPT solar charge controller would work. It expects to be charging a battery, but instead you connect the resistive element. It seems to me that as long as the resistance of the element is low enough to take maximum power at maximum insolation ("sun power"), this should work.

In all your cases, the voltage to the element has to be less than (or nearly equal to) the input (PV) voltage. So a standard buck mode MPPT charge controller should work. It would automatically find the correct output voltage, since it is adjusting the output current to maximise power transfer. No need to "tweak" for the resistance of the element, provided that it's low enough to take full-sun output from the panels without boosting the voltage.

I hope I'm not overlooking something obvious.

[Honu]

It seems to me that an off-the-shelf MPPT solar charge controller would work. It expects to be charging a battery, but instead you connect the resistive element. It seems to me that as long as the resistance of the element is low enough to take maximum power at maximum insolation ("sun power"), this should work.

That's what i'm starting to feel too .. but my electricity logic is wrong somewhere.

In all your cases, the voltage to the element has to be less than (or nearly equal to) the input (PV) voltage. So a standard buck mode MPPT charge controller should work. It would automatically find the correct output voltage, since it is adjusting the output current to maximise power transfer. No need to "tweak" for the resistance of the element, provided that it's low enough to take full-sun output from the panels without boosting the voltage.

I hope I'm not overlooking something obvious.

My feeling after doing some more calculus is that you are right .. BUT i need to understand where i m wrong :
I think i misunderstand some concept.. , let's say i got a 7 Ohm resistive element rated for 2000W Max.
I'm applying 33V and 2A, what is the power dissipated by the element ..? And if now i apply 33V and 4A ?

Ohm's law give me 3 ways of finding the dissipated power .. but their results are different, yea i know Ohm is not wrong, it's how i use it that is the problem .. (V = R x I)
A : P = V x I
B : P = R x I x I
C : P = V x V/R

Which formulae to apply ?

For 33V/2A
A : 33 x 2 = 66 Watts
B : 7 x 2 x 2 = 28 Watt
C : 33 x 33 / 7 = 155.6 Watt

I'm doing it wrong .. but .. why .. where ?

[coulomb]

I think i misunderstand some concept.. , let's say i got a 7 Ohm resistive element rated for 2000W Max.

Ok, so that's rated for V = √(P·R) = √(2000 · 7) = 118 V. Perhaps it's designed for 120 VAC.

I'm applying 33V and 2A, what is the power dissipated by the element ..? And if now i apply 33V and 4A ?

Your problem is that you're choosing both the voltage and the current; you can only choose one and the other will follow from Ohm's law.

If you apply 33 V to a 7 Ω element, the current will be I = V/R = 33/7 = 4.7 A. If you want to cause 2 A to flow, you'll need V = I·R = 2 · 7 = 14  V. For 4 A (double the current), you'll need double the voltage, or 28 V. You can figure out the power using any of the formulae. For example, applying 33 V: P = V²/R = 33²/7 = 156 W. Or P = V · I = 33 · 4.7 = 155 W (the 1 W difference is a rounding error). Finally, P = I² · R = (4.7)² · 7 = 155 W.

Which formulae to apply ?

Usually, the one that has no unknowns in it. Otherwise, you have to use another formula to find the unknown.

[T1 Terry]

Is there a device able to transform whatever Voltage x Amp (max 140V, max 40 Amp) to a linear Voltage x Amp .. ?
In short, a DC to DC convertor with variable input and variable output, output just have to be linear, this line will depend on the resistance of the element.

It seems to me that an off-the-shelf MPPT solar charge controller would work. It expects to be charging a battery, but instead you connect the resistive element. It seems to me that as long as the resistance of the element is low enough to take maximum power at maximum insolation ("sun power"), this should work.

In all your cases, the voltage to the element has to be less than (or nearly equal to) the input (PV) voltage. So a standard buck mode MPPT charge controller should work. It would automatically find the correct output voltage, since it is adjusting the output current to maximise power transfer. No need to "tweak" for the resistance of the element, provided that it's low enough to take full-sun output from the panels without boosting the voltage.

I hope I'm not overlooking something obvious.

I think you are missing the fact an MPPT controller needs to see a battery voltage so it can determine the voltage/current that is the optimum output from the input it sees.
This is where the Electrodacus MPPT unit comes in, it will change what ever is coming in to a useable output designed for heating via a resistive element. Have you looked at the SBMS0, it is basically a control device that monitors cell voltage and drives remote devices including an MPPT charge control as long as you design the on/off to work from a low current open/closed contact supplied by the SBM0.
If you are looking for a constant voltage then why not just add a simple square wave inverter powered by the battery bank, a resistive load is not fussed about wave shape, it is just a controlled short circuit designed to convert electrical energy into heat energy.
If you want efficiency then you need to look at a heat pump to heat the water rather than a resistive element, then you do need clean pure sine wave 240vac.

T1 Terry