Maker Pro
Maker Pro

Switching 240v source automatically

flippineck

Sep 8, 2013
358
Joined
Sep 8, 2013
Messages
358
Source #1 is a solar power setup as follows:

3 x 80 watt solar panels "nominal 12V" in parallel
One allegedly MPPT but probably PWM, 10A solar charge controller
2 x 12V 110Ah lead acid deep cycle batteries in parallel
one 1000W cont / 2000W peak pure sinewave inverter (which can realistically handle 600W)

The controller has a pair of input terminals for the panels, a pair for the battery, and a pair marked 'load'.

The 'load' terminals supply 12VDC and are switched off automatically when the battery voltage falls below a programmed level. When battery voltage rises above a seperate programmed level, the 'load' terminals begin to supply 12VDC again.

The controller will only accept certain programmed values such that, the 'load off' voltage is always a bit below the 'load on' voltage.

Source #2 is, a standard UK 13amp domestic mains ring outlet.

You can feed a 12V load directly off the controller 'load' terminals. The controller is only rated at 10A though. I never get much above 9 or so amps from the panels, but I often want to take much more than 10A to feed the inverter. So, I have the system rigged up such that the controller only controls the battery charging; the load terminals are left unconnected and I draw power directly from the battery for the inverter.

The controller seems happy with this and charges ok.

What I was thinking about doing was to use the 'load' terminals on the controller, to power a ganged 3 way, 2 pole relay coil.

Connect the 3 commons to the final outlet socket, one for L, one for N, one for CPC.

Then wire each fixed contact of the relay with one pole going to the inverter (source #1) and the other to a 13Amp plug to go in source #2

That way, when the controller thinks there is enough battery power to run the inverter, it will switch the inverter through to the final outlet. but when the controller decides the battery is too flat, it will changeover the relay and disconnect the inverter before connecting the grid mains through to the final outlet.


Problems I've thought of are..

(1) There'll be a 'brown-out' on the final output when the relay switches. I'm powering kitchen appliances like a steamer, can opener, waffle toaster etc though rather than sensitive electronic devices so hopefully that won't matter? They don't seem to mind when I replug them manually.

(2) What should I do with the CPCs / grounds / earths?

One idea would be to use a relay with three ways, so I'd switch live, neutral and CPC fully and completely.

This would mean that momentarily the load would be left without the CPC connected. Only for a brief time whilst the contacts flick over, and during that time, the live conductors would also be disconnected. My worry here would be behaviour under fault conditions?

Also 2-way relays are common and cheap whereas ones with 3 seperate gangs are less so, so I thought it might actually be better to just connect all the CPC's together permanently and only switch the lives and neutrals.

But would this be OK? I'm guessing (dirty word) the inverter's output's earth / ground / CPC pin wouldn't mind being tied down to the grid's CPC but I vaguely recall something about computer power supplies not being straightforward in the earthing department, wondered if power inverters tended to be similar in that regard? Should I keep the grid's earth and the inverter's output socket earth away from each other?

I was wondering.. if I used a standard relay, it would take 75mA to keep energised all the time the battery was deemed to be OK. Could I convert the load output from the controller to a pulse output with a suitably sized capacitor / resistor, and then use some kind of latching relay rather than one which requires to be permanently energised? Like some kind of flip/flop latch on / latch off type.. I dunno if these are commonly available. but, 75mA at 12V is something of the order of one watt, that should only cause a negligible battery drain shouldn't it.. is it worth bothering making it operate off a momentary pulse?

Thanks for reading to the end of the ramble, thoughts appreciated
 

flippineck

Sep 8, 2013
358
Joined
Sep 8, 2013
Messages
358
This is the kind of relay I was thinking of.

http://www.deltrol-controls.com/products/relays/magnetic-latching-relays-5-10-amps/105ml/20039-81

http://www.deltrol-controls.com/sites/default/files/brochures/105ML Series 102011.pdf

If I inserted a relatively large capacitor inline with it's coil and the load terminals of the controller, I think it might latch when the power comes on due to the capacitor allowing a short pulse of current? but then draw no current as long as the status of the load terminals didn't change

When the power from the load terminals ceases, I'm thinking there will then be another transient pulse of current through the relay coil but of the opposite polarity, that should cause the relay to de-latch?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
If I get this right, you want to power your 12V appliances from the battery (while the battery voltage is OK) and from a mains operated 12V source when it's not.

A relay operated from the regulator output would seem to be an acceptable solution. The controller already probably allows for short dips below the trigger voltage (to allow for surge loads) so it might not need anything special.

I'd probably go for a relay rated at a significantly higher current than your peak load because it's going to be switching under load. 20A to 40A would be my preference. Given that it's a 12V rail, an automotive relay might work.
 

flippineck

Sep 8, 2013
358
Joined
Sep 8, 2013
Messages
358
Sorry, seems I've confused you.

The controller provides a 12V DC output which goes live when the main storage battery voltage rises above let's say 11V, and goes dead if the battery voltage falls below let's say 10V. These voltages are programmable within limits that always ensure a bit of hysteresis.

This output is intended by the manufacturer of the controller, to be the main 12V DC power output to the load, for the system.

The controller is only rated at 10A though, for both solar panel input and load output, and I need to take a lot more than 10A from the output sometimes.

So, I connect my load (which is a 12VDC --> 240VAC power inverter) directly to the storage battery with very thick cables, and just let the controller handle the charging current from the solar panels through to the battery.

The system works fine like this, but the kitchen appliances sometimes run the inverter down into undervoltage shutoff etc.

The solar controller does a fine job of monitoring the state of the storage battery so I was hoping to use it's load output, just as a control signal for a meaty control relay, whose 3 sets of contacts could be used to select between 240VAC from the inverter or 240VAC from grid mains (using one set of changeover contacts for live, another for neutral, and the last for earth)
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
OK, I understand. Yeah, that sounds OK. It would be best to have something with a small time delay so it doesn't switch back and forth rapidly as there will be nothing keeping the phase relationship between them.
 

flippineck

Sep 8, 2013
358
Joined
Sep 8, 2013
Messages
358
Do you think, tying the inverter's 240VAC output earth pin permanently to the grid's earth, would be okay?

(To do away with one set of relay contacts, and make the load have complete permanent unbroken connection to an earth)

Can't think of a reason why it wouldn't but I just remember, PC power supplies make some kind of active use of the earth which means their earths aren't totally and completely isolated from L & N - are PC PSU's pretty unique in this or is it something I should watch out for with inverters too

ISTR I was working with an electrician one time & we had problems tripping earth leakage circuit breakers in the distribution panel, it turned out the problem was caused by the normal, non-faulty operation of a large number of PC PSUs connected to one particular ring.. they were all contributing a small trickle to earth by design, and the sheer number of them took the ELCB over it's trip milliamps.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I can't think of any reason why I'd ever want to disconnect the earth lead in this sort of application.

Leakage to earth can cause problems, but it is exactly that, it's leakage, and that's what earth leakage breakers are supposed to detect.
 

flippineck

Sep 8, 2013
358
Joined
Sep 8, 2013
Messages
358
In the case of the actual country-wide grid, earth really does mean earth as in 'ground' as in 'soil'.. earth rods etc..

But the earth pin on an inverter can't be the same because.. well there's no earth rod?.. I'm assuming here, but I suspect an inverter's mains output earth pin is actually a second neutral pin? so relative to another discrete electrical system such as the planet, the inverter's earth would be independant & 'floating'?

The relay itself is going to momentarily disconnect earth if I use 3 full sets of changeover contacts.. so it's just a case of trying to ascertain whether connecting grid earth to inverter output earth (and load earth) as a permanent fixture is safe... and get by with just switching L & N through the relay. Can't for the life of me think why it wouldn't but many times before been in the situation of being 100% sure only to be proved 100% wrong!
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Yeah, the earth on the inverter should be connected to an actual protective ground.

But you're working with high energy (mains supply) so it's always worth being very careful.
 

flippineck

Sep 8, 2013
358
Joined
Sep 8, 2013
Messages
358
Guess I have to consider the fault condition of one or more contacts welding themselves together, so that the load is stuck seeing:

Inverter Live
Grid Neutral
Grid Earth

Inverter Neutral
Grid Live
Grid Earth
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Neutral is probably bonded to Earth at your switchboard for your mains connection. You need to check if neutral is bonded to the earth at the inverter or floating. If it's also bonded to earth then the relay contacts for neutral may seem superfluous. However if the neutral relay jams, you may generate earth currents which in turn may trip ELCBs

The live relay sticking is electrically similar to the neutral relay sticking, so you may also get ELCBs tripping.

If neutral is floating on the inverter, a sticking relay in either direction will result in an open circuit, albeit with mains voltage possibly present.

First determine if ground is bonded to neutral on your inverter.

Then determine if earth is bonded to neutral it your switchboard (AFAIK this is required for ELCB's to operate reliably)

Then draw up the two incoming supplies and the relays and see if you can find a path for current to flow in any fault situation.

Note that your relays should be connected so that the load goes to the common connection with the power sources on the NO and NC contacts. It should be almost impossible to short the NO and NC together, a situation that could be disastrous for your inverter.
 

flippineck

Sep 8, 2013
358
Joined
Sep 8, 2013
Messages
358
Are inductive loads liable to cause scenarios that need attention?

Let's say, the system is powering a food mixer from the grid on a sunny day when by & by, the battery voltage becomes high enough to turn the controller 'on'.

Relay coil is energised and

(1) grid is disconnected
(2) load floats momentarily (milliseconds)
(3) inverter is connected, whilst food mixer's motor is still running at full pelt

Could there be a back EMF that would cause a problem.. or would any effect likely be insignificant?

I've been turning all this over in my head and it seems what I'm actually building, to use a mechanical analogue, is a 'crash gearbox'. What could I incorporate in terms of a synchro?
 

shumifan50

Jan 16, 2014
579
Joined
Jan 16, 2014
Messages
579
If you want to be super safe, then you could disconnect from the one supply and delay before connecting to the other supply. This could best be achieved with a micro controller that monitors the 12V supply and drives 2 sets of relays, one for each source. That way the switching delay can easily be managed and as fault conditions will in most cases cause a drop in the supply voltage, this could also be controlled. By adding a current sensing circuit, the source can also be switched when the load exceeds a preset current. By monitoring the current it will also be possible to avoid switching 'on' both supplies at the same time as you can wait for the current to drop to zero on the source being switched off before switching over.
Delaying the switchover will cause momentary loss of power, so would not be suitable for powering electronic devices like computers - but in your first post you said you will be powering appliances.
 

flippineck

Sep 8, 2013
358
Joined
Sep 8, 2013
Messages
358
I've never played with a microcontroller before. Could be an interesting project :)
 

shumifan50

Jan 16, 2014
579
Joined
Jan 16, 2014
Messages
579
For what you want to do the hardware will be really simple, but it would need some programming effort. It will also open up a new world in electronics for you.
 

flippineck

Sep 8, 2013
358
Joined
Sep 8, 2013
Messages
358
If I was going to go microcontroller, I wouldn't mind trying to go the whole hog & try to develop something that could accomplish not only the above, but also management of the power flow from solar panels to storage battery. The controller I'm using calls itself 'MPPT' but a lot of folks are saying it isn't due to the price & lack of any kind of inductor on the board. They say it's actually a PWM controller.

Decent controllers that I can verify are true MPPT seem to start at a very high price so maybe I could do something via my own programming.

I'm better with BASIC than C & assembly throws me a bit. Pointers flummox me. Expect a bunch of beginner postings in the microcontroller section ;-)
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
MPPT isn't that hard, and you can do it yourself.

Basically you have a buck (typically) regulator charging the batteries. This means that the output voltage is fixed, so changing the mark/space ratio of the switching device changes the output current.

So, in essence, you monitor the output current and always seek the mar/space ratio which yields the highest current.

The simple way to do this is to store a direction that you're going (either increasing or decreasing). You measure the current, then make a step in whatever direction you're going. If the current increases you do nothing (i.e. the next time you'll step in the same direction. If the current falls, you reverse direction for the next step. This will track the mark/space ratio required to transfer maximum power.

On top of that you need to determine if the battery is charged, and stop when it reaches that point -- typically by reducing the mark/space ratio to track a particular voltage on the battery. At this point you're wasting energy because the solar panels can provide more energy than you're using.
 

Gryd3

Jun 25, 2014
4,098
Joined
Jun 25, 2014
Messages
4,098
MPPT isn't that hard, and you can do it yourself.

Basically you have a buck (typically) regulator charging the batteries. This means that the output voltage is fixed, so changing the mark/space ratio of the switching device changes the output current.

So, in essence, you monitor the output current and always seek the mar/space ratio which yields the highest current.

The simple way to do this is to store a direction that you're going (either increasing or decreasing). You measure the current, then make a step in whatever direction you're going. If the current increases you do nothing (i.e. the next time you'll step in the same direction. If the current falls, you reverse direction for the next step. This will track the mark/space ratio required to transfer maximum power.

On top of that you need to determine if the battery is charged, and stop when it reaches that point -- typically by reducing the mark/space ratio to track a particular voltage on the battery. At this point you're wasting energy because the solar panels can provide more energy than you're using.
Sorry to jump in like this. I've been toying with a Solar idea for a while. This would require the Panels to be wired in series, or would any configuration higher than the battery voltage work? (MPPT just sounds like an automatically adjusted PWM...)
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
In general the topology is such that the panel voltage is higher than the battery voltage, simply because it's easier to do that and because that results in lower cable losses (and buck converters tend to be more efficient than boost converters). However there's no reason why it has to be done this way.
 
Top