@Sarel.Wagner I didnt like the idea of switching the sonoff on and off too rapidly, Im still concerned it will fail oneday.
what % 0f the Sonoff capacity are you switching? Normally, if they lasted a month or 2, they will keep on keeping on once past the first few months. Infant mortality is a thing. As long as you do not overload the Sonoff, it should last well…
Groetnis
Ita a sonoff pow R2, they rated at 16A, its been running for 2 years switching the 3KW element, thats quite close to its rated capacity, So after changing to the 2KW element its significantly lower, and even more so if it switches on with the power draw reduced to 50%.
Yes, for a big load like a geyser, you do not really want to do phase control - no inverter likes it (I have seen in a manual where they state that it is explicitly forbidden, but cannot recall in which one). And it creates a lot of noise on the power lines. For a load like a geyser the ideal control is referred to as “PWM” but not in the sense that most people know - here it refers to switching a complete cycle on at a time. Optimal control is dithering with something similar to Bressenham’s line algorithm. This leads to the highest frequency and most accurate result - at 50% it is on one cycle and off one cycle.
Unfortunately this method is also not ideal for an inverter as a source…
With that in mind, what you should actually do is high frequency PWM control with either two opposing MOSFETS, or a MOSFET and a bridge rectifier - or better yet an IGBT and a bridge rectifier. (but an IGBT is more comlicated to control…)
As long as you still have the thermostat as fallback in the geyser, relays with big loads like to weld themselves into the on state…
My geyser are controlled by SCR zero cross switching, kind PWM but not. It will switch half cycles and for 50% it will do like 5 on and 5 off or 10 cycles on and 10 off, always zero crossing switched.
Next I will try phase angle switching with an industrial controller, same as the zero crossing controller. @_a_a_a What is your take on this.
These are the controllers: CCI Control
Phase angle switching is phase control…
If you are doing it from scratch, consider high frequency switching (the software will be much simpler, hardware about the same). Just remember to also add appropriate filtering, otherwise you will create more noise than the other methods.
Something like this:
from:
(notice the filtering is not included)
1 Like
You mean it will switch full cycles in bursts?
Very neat! What did this cost you and where did you get it?
Thank you @_a_a_a
Maybe V2 will be based on some PWM control
Trolling fleabay and the CCIPower website to get the model numbers squared away. The model # tell ya everything…
Fleabay general search, normally between $40 and 150…
CCI Power Phase angle controller, the new one I want to testa different brand
Groetnis
A bit pricey.
One could use a regular SSR as I see it…
And then there are those who sommer use a contactor (like me!)
But then you would not get the proportional control, which is what is wanted here (cause this is an Axpert, no grid mixing).
That too. A stupid edge-leading dimmer (the most common kind) causes a very poor power factor, and usually inverters don’t like that too much. But I think if the inverter is actually capable of the whole load, and your reason for “dimming” the geyser is to not discharge the batteries, it is probably fine.
(Besides, it’s an Axpert… if it blows up it’s not too big of a tragedy 
).
You do get dimmer versions as well…
See the first offering here:
ASH-XX-XX_AOLE.pdf (749.7 KB)
One thing I learned with females and them needing boiling water to shower … if they don’t have it, the shiite storm that hits you is indescribable.
The Marital Laws state: Thou shall ensure thee wife has boiling water at all times.
As such, I have never and will never bother with geyser automation bar GeyserWise timers.
If the timer says geyser must go on, sunny day or not, the geyser shall go on.
At LEAST I got the females to agree to some sort of schedule … which may change at any point depending on how “they feel” at that point in time. So yeah, even that agreement is tenuous at best.
Just an update on this project.
SCR worked flawlessly, I noticed a significant change in solar generation over the initial winter months.
So it worked a charm to cope with the slightly lower generation.
I had an “issue” with running the 2KW element in winter, specifically how long it took to heat the geyser in the morning if anyone took a bath at night.
Now who just adjusts the schedule to just run it a little earlier? Not me, thats who.
So I figured I could put the 3KW element back, and still limit the power to a max of 2KW on solar and full tilt on Grid. Worked a charm for 3 days and the SCR died.
They were cheap enough that I replaced it, guess what, new unit died within an hour of fitting it.
So I was pretty certain that they were rated incorrectly, as they state they are rated for 4KW.
Ordered a 5KW version and its been running for about 2 weeks now.
1 Like
Thanks for the update! Looks like SCRs are still with us…
If you were to start this project over from scratch what would you change?
Yes, the ones from china are almost always way under specced. A 40A aliexpress solid state relay usually has a BTA16 triac in it. (They are rated for 16A).
Rob a bank and order a fancy SCR from the states, but If I had to do everything over, id have a different inverter and solar setup, and wouldn’t bother with this project.
1 Like
Oh man, they are excellent (if ancient) little devices. Much lower drop than a Bipolar transistor (therefore smaller heatsink), cheaper and higher voltages available than MOSFETs. Of course there is that thing that they latch in the “on” position, but for AC-control that is not a problem.