Author Topic: Wanted, MPPT inverter that will keep running and just lower AC output voltage.  (Read 3965 times)

0 Members and 1 Guest are viewing this topic.

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 9837
  • Country: nz
Yeah, probably need to run panels in parallel to get ~48V @ 22A and then DIY a boost converter to get that to like 60-130V to drive the current 3kW element at 1kW at 130V down to whatever min power 60V is.

Or, might be simpler to just get a cheap generic 48V solar inverter to 110 or 230V and then rectify the output and PWM that into the element with some caps to reduce avg current loading the inverter sees to what the solar system can supply. Might be a bit tricky to get tuned right though. Has to respond faster than the inverter can so it doesn't shut down from min input volts error.
And pwm freq would have to be fast to avoid needing huge caps.
« Last Edit: June 05, 2023, 01:20:45 pm by Psi »
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline fourtytwo42

  • Super Contributor
  • ***
  • Posts: 1179
  • Country: gb
  • Interested in all things green/ECO NOT political
You might find this forum https://diysolarforum.com/forums/solar-water-heating.60/ helpful except most of them live in the desert and have never heard of clouds and short winter days  ::)
Lots of ideas though and mostly off-grid isolated systems.

Have you ever heard of or used LTspice ? very useful tool for modelling solar panels and trying out circuits without actually building them!
Also try this site https://re.jrc.ec.europa.eu/pvg_tools/en/tools.html#DR for insolation information in your backyard (seems to cover NZ).
« Last Edit: June 05, 2023, 02:22:04 pm by fourtytwo42 »
 

Offline marekm

  • Contributor
  • Posts: 17
  • Country: pl
Another solution to consider - switch the PV panels in series or in parallel to better match the resistive load, depending on the amount of sunlight.  Less efficient but no EMI issues, use MOSFETs as DC switches without arcing.  If you can have more than one heater element - these could be switched too.
 

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 9837
  • Country: nz
Another solution to consider - switch the PV panels in series or in parallel to better match the resistive load, depending on the amount of sunlight.  Less efficient but no EMI issues, use MOSFETs as DC switches without arcing.  If you can have more than one heater element - these could be switched too.

Yeah, at the cost of a few contactors its probably a good idea,  just need to be DC rated.
Or could use semiconductors I guess.   Maybe cheap AC contactors and single N-Mosfet in series so contactor changeover is done while current is zero.
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline Ed.Kloonk

  • Super Contributor
  • ***
  • Posts: 4000
  • Country: au
  • Cat video aficionado
Yeah, probably need to run panels in parallel to get ~48V @ 22A and then DIY a boost converter to get that to like 60-130V to drive the current 3kW element at 1kW at 130V down to whatever min power 60V is.


I don't know if I would run a water tank at >50vdc. The themostats just isn't rated for that. Apply a leakage breaker at least.
iratus parum formica
 

Offline Ed.Kloonk

  • Super Contributor
  • ***
  • Posts: 4000
  • Country: au
  • Cat video aficionado
How about this:


from:
https://www.instructables.com/safe-and-simple-AC-PWM-Dimmer-for-arduino-Raspberr/

I've yet to get around to seeing what it looks like in spice with a beefier fet, say 25A or more.
iratus parum formica
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
I've been doing that for years with water heaters.  I use a fixed power point voltage so I can use it in parallel with MPPT charge controllers for diverting excess power. It can temperature correct to follow the natural power point or just adjust seasonably. Simple and efficient. It pulses the heater element at no faster than a 50Hz rate with an arc interrupt off period should the ON pulse last longer than 10ms.  This allows standard mechanical thermostats to be used without burning contacts. I heat three separate tanks and a priority for each tank can be set. Graph is a half hour of diversion in passing clouds.

 
The following users thanked this post: Ed.Kloonk

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 9837
  • Country: nz
Yeah, probably need to run panels in parallel to get ~48V @ 22A and then DIY a boost converter to get that to like 60-130V to drive the current 3kW element at 1kW at 130V down to whatever min power 60V is.


I don't know if I would run a water tank at >50vdc. The themostats just isn't rated for that. Apply a leakage breaker at least.

Other than the contact current ratings, why would the thermostat care if the water is being heated by 50V or by 220V?  I assume its just a mechanical thermal switch triggered by water temp?
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 9837
  • Country: nz
I've been doing that for years with water heaters.  I use a fixed power point voltage so I can use it in parallel with MPPT charge controllers for diverting excess power. It can temperature correct to follow the natural power point or just adjust seasonably.

Not too sure I understand that correctly.
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline marekm

  • Contributor
  • Posts: 17
  • Country: pl
Other than the contact current ratings, why would the thermostat care if the water is being heated by 50V or by 220V?  I assume its just a mechanical thermal switch triggered by water temp?

The switch is rated for AC which extinguishes the arc at zero crossing, when interrupting high power DC the arc will burn for much longer, could catch fire or weld the contacts closed, overheat and "rapid unscheduled disassembly".
 
The following users thanked this post: Ed.Kloonk

Offline Jeroen3

  • Super Contributor
  • ***
  • Posts: 4045
  • Country: nl
  • Embedded Engineer
    • jeroen3.nl
I'd find a suitable microinverter, have the panels operate in grid mode. This way they do as you ask.
Then get one of those small controllers, or make one yourself, that just chops the boiler heater the same amount as the exported power.
Eg: https://www.gpceurope.com/product/gpcontrol-smart-boiler-controller/

Just so I understand you correctly,  you're suggesting a micro-inverter that's setup for grid-tie but in a grid offline state 24/7.
Then using that controller to command the micro-inverter to drive the heating element instead (since its in grid offline state) in a way where it switches to driving a resistive load (water heater) and will maintain max solar power into the resistive load with a very wide voltage range?
No, what I mean is that you use the micro grid tie inverter as designed. Exporting to grid.
Then you wire a switch or controller with a CT or power transducer into your export point and have it turn on the boiler heater when the power is negative. The same way Dave's myenergi charger works on eco mode.
This way the panel will operate at mpp regardless, that's what the micro is for. And even if your boiler is full you can use the excess yourself or export for the negligable amount of money it gives you.

Other question: Is using (pwm) DC in electric heaters a problem with galvanic corrosion?
 

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 9837
  • Country: nz
Other question: Is using (pwm) DC in electric heaters a problem with galvanic corrosion?

Fair question. 

but I think the element is electrically isolated from the water
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
There are a number of commercial products available, epically in Europe for stand alone systems. I think a using PV only to heat domestic water is a little goofy since the idea is for it turn off. There are a lot of other places that energy can be used.  But panels are cheap.  Search ACTii and a number of other companies in Poland. My laptop died and I lost the list of products.  Here is one using my design in Lithuania.   It looks like for adults only, not for people who sit at the kiddie table of solar.



     AC7391 video

https://create.arduino.cc/projecthub/stevetearle/loadmaster-xp-a-smart-pv-mppt-solar-hot-water-controller-4a813f

« Last Edit: June 06, 2023, 02:21:19 pm by Seekonk »
 

Offline f4eru

  • Super Contributor
  • ***
  • Posts: 1079
  • Country: 00
    • Chargehanger
Wouldn't it be better to use the Sun to heat the water directly by sending it through a solar collector?
The solar panels + conversion + losses in the wiring, etc. can't be more efficient than that.
From the energy efficiency, yes.
From economic efficiency, it does not make sense to invest in thermal solar.

I have thermal solar on my roof, at the time of install it made economic sense at my location, but today, not any more.
PV is widely in front, because it costs the same, but you get not only hot water, but also excess electrical energy for the same install cost.
This is especially true if using a thermodynamic water heater, but even with resistive, it makes plenty of economic sense.

Offline janoc

  • Super Contributor
  • ***
  • Posts: 3775
  • Country: de
From the energy efficiency, yes.
From economic efficiency, it does not make sense to invest in thermal solar.

I have thermal solar on my roof, at the time of install it made economic sense at my location, but today, not any more.
PV is widely in front, because it costs the same, but you get not only hot water, but also excess electrical energy for the same install cost.
This is especially true if using a thermodynamic water heater, but even with resistive, it makes plenty of economic sense.

Not sure what kind of thermal solar are you talking about but the small water heating setups are literally just a lattice of water pipes painted black and connected to a reservoir. Plus you want a pressure release valve.
 
Good luck finding a PV setup + inverter + installation costs for the price of a bunch of scrap piping and some valves.  :-//

Heck, you could even get a small portable system - they are commonly sold to heat pools ...

The bigger sets that heat enough water for 4-5 people household (~300 liters or so) cost about 3000€ (see e.g.: https://www.solar-eshop.sk/c/solarni-termika/solarni-sety/ ) if you want to buy an off-the-shelf system. A PV system with comparable power output is not going to be cheaper for sure.

« Last Edit: June 06, 2023, 04:52:47 pm by janoc »
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
I'll put my PV water heater against your scrap pipes any day. $20 for electronics, #300 for a KW of panels, and a free water heater someone gave me because the plastic cold water down tube broke off. And mine only operates on excess solar normally wasted with 70% of my panels shaded at any one time. It would be absolutely impossible to heat with solar thermal at my location.

This is a 1KW control made from a $10 ebay inverter board.
« Last Edit: June 06, 2023, 11:21:01 pm by Seekonk »
 

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 9837
  • Country: nz
Yeah, I've been thinking about the best approach for my application.

Best plan for me is to buy a generic 1.5kW 48V to 230V inverter.
That way i can actually power 230V stuff in an emergency during the day.

Then for water heater thing, just rectify it and PWM into the current 3kW element at like 1khz and smooth out draw with some caps.  Max duty is going to be around 50% to max out panels but that's fine.

Saves me having to do much work other than making something to PWM the element with feedback to keep panel voltage at max power voltage. (and low current loop through thermostat).

No need to change water cylinder element or anything.
and inverter is off the shelf
and pwm + filter for 1kw shouldnt be that hard.
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 8922
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
You don't actually need AC, just a non continuous current for the thermostats to be happy. How about a boost stage and then a fixed 100Hz or so PWM to interrupt the current? The boost stage operates at whatever duty cycle it needs to in order to stay at MPP, unless the output voltage is already at the 300V or so limit.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 9837
  • Country: nz
You don't actually need AC, just a non continuous current for the thermostats to be happy. How about a boost stage and then a fixed 100Hz or so PWM to interrupt the current? The boost stage operates at whatever duty cycle it needs to in order to stay at MPP, unless the output voltage is already at the 300V or so limit.

DIY DCDC boost stage would work well, but would then need to spend time actually designing, building and testing one.   I kinda want to get a proper 230V inverter anyway, just so i have the option to run things from the panels

So it's easier to just use a off-the-shelf inverter, rectify then filtered PWM so it can drive the 3kW element with 1kW and not freak out the inverter.  But that creates the need to separate out the connection to the thermostat so it doesn't have to switch full DC load. But it needs a MCU anyway to do the PWM so pretty easy to add an extra input. Probably do more than one temp cutout just for safety. 

I guess it could be inverter and the two stage pwm, first pwm does mpp and runs at like 1khz,  2nd pwm runs at 50hz with no filtering. Then the thermostat would be happy and wouldnt have to do any rewiring of the cylinder.
« Last Edit: June 07, 2023, 04:16:49 am by Psi »
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline Ed.Kloonk

  • Super Contributor
  • ***
  • Posts: 4000
  • Country: au
  • Cat video aficionado
Then just use the triac method. I cheated and used that common arduino ebay light dimmer module with zero cross detection and substituted the 8A triac with the BTA41-600B Triac 600V 40A remoted on a giant bit of aluminium. Snubber builtin is iffy but circuit worked with 4.8kw element, 2000w inverter (that is really max 1500w   >:( ).

Just looked the 1.8kw element was fed about 4kwh @1kw today. The inverter can and does power computers and whatnot up to 400w sometimes as well so the waveform isn't -too- bad. Apparently. :-\

 :-+
« Last Edit: June 07, 2023, 07:20:21 am by Ed.Kloonk »
iratus parum formica
 

Offline PsiTopic starter

  • Super Contributor
  • ***
  • Posts: 9837
  • Country: nz
You're saying to just use a MCU controlled Triac dimmer on the 230V AC inverter output to power the 3kW heating element at 1khw and under for doing MPPT?

Hm.. yeah, might need an inverter with a higher surge current, but that's ok. They are usually a lot more than 1kW anyway.

Definitely quite a bit easier to wire up
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline Ed.Kloonk

  • Super Contributor
  • ***
  • Posts: 4000
  • Country: au
  • Cat video aficionado
The software for the dimmer is already in the arduino lib. You call set_power(x) percent and there is a 100 int array that maps the power percent to timer wait-off time after the zero cross. So yeah, at 50% it turns on right at nearly the peak of the wave. After so long now, I still haven't done the sums on how well the supplied snubber matches the inverter impedance versus the grid it was designed for. Also to account for some inches of wire the triac is now away from the snubber. Does that even matter?

That's why I said a PWM would be a bit kinder but more complicated. I just set it up to see if it could be done not to be elegant.

The early iteration with the 4.8kw element and a rather small AGM battery, yes it often crashed when feeding 1.5kw and the shade blew over the panels. Until you get the algo tightened up with the proper power setting, your battery should be able to withstand the load until your controller decides how to respond.

Polishing the home-made algo that decides on the power setting is the tricky bit. Depending on how you detect the instantaneous output of the solar panels, the mood of the mppt and the charge level of the battery, hold your tongue right.

You'll get a good power result if you rely on the battery to pick up the slack and then recharge, but you prolly shorten the life of AGM batteries by doing severely that over many days and months.

I should point out that depending on the MPPT controller, ramping power up and down can drive it nuts. You have to decide whether it prefers a slow gradual ramp (up or down) or a stepped ramp (up or down). That's why, I think, the project suggested above included their own MPPT as well, as a closed loop. Bells and buckles.

Now, I'd probably use a Pi instead of the Arduino simply because of the ability to surf on in from anywhere to tweak the algorithm performance across four weather seasons and if you add more panels. Though my Arduino has Ethernet, you can't really change the code so easily, obviously.
« Last Edit: June 07, 2023, 11:02:24 am by Ed.Kloonk »
iratus parum formica
 

Offline Ed.Kloonk

  • Super Contributor
  • ***
  • Posts: 4000
  • Country: au
  • Cat video aficionado
I forgot to mention my algorithm listens to the serial output of the MPPT and specifically the MPPT state flag. Ramps up slowly with a j-curve and drops about 25% power when the MPPT flag raises. But of course, YMMV.
iratus parum formica
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf