Switching solar panel strings between inverters/SCCs

[fafrd]

I’ve got a 1S3P 1kW DC-coupled array that I am thinking about using for grid export until a power threshold is reached and then switching one panel at a time away from the Microinverter over to an SCC.

One of my 380W panels has an Isc of 11.5A, so a bit beyond what can easily and cheaply be switched with a 10A relay, but Imp is 10.96A and I’ve never gotten past 90% of the 380W rating, so I’m ready to take the risk of losing a low-cost relay or two.

The idea would be to have all 3 panels tied to the Microinverter in the morning to generate and export power until the inverter nears saturation at 580W of output.

Each 380W panel will be supplying over 190W of power when the Microinverter is maxed out at 580W, meaning current of ~1/2 Imp or ~5.5A.

So my thought is se a current switch set to 5 or 5.25A to drive a pair if 10A relays that will switch that panel from the Microinverter to the SCC.

At that point, the two remaining panels will saturate the Microinverter at a power of ~290W each or ~0.76Imp (8.4A), so a second current switch set to 8A on the second panel can switch it from the Microinverter to the Isc.

The Microinverter can always remain connected to the third panel since it will never output enough power to saturate the Microinverter.

The two relays feeding the Microinverter will never see over 10A, so should be OK.

The two relays feeding the SCC might see current of over 10A if conditions ever allow the panels to exceed 91.26% of rating (which I’ve never seen over a full year of use).

Since this 3-panel 1kW DC-coupled array is a SE-facing array with earlier output than my SW-facing 4kW grid-tied array, I’m also thinking about controlling the second panel with a current-sensor tied to output current of the larger AC-coupled array or also using a timer to only allow switchover at 12:00 when the AC-coupled array gets up past 80% of peak output.

I’m interested in anyone else who is successfully switching DC solar panel output between different energy converters.

Any advice on recommended relays or transfer switches as well as techniques that work or do not work appreciated…

 

[McKravitts]

I am reposting my last comments from the other topic in hopes we can get a few more, perhaps more eknowledgible people involved.

I have installed two Functional Devices Inc. RIB24P30 dpdt relays.
One controls the outputs of two solar strings either to charge my battery bank via mppt controllers or connect to a grid tie inverter based on the soc of my batteries. If the state of charge drops below 90 % (as measured with a smart shunt) the power is directed to charge the batteries. When soc rises to 95 %. it switches to grid tie. As a backup my off grid inverter/charger will go into charge mode when the soc drops below 20 %.

The batteries drive an off grid inverter/charger which outputs to a 6 circuit transfer switch.

The other relay can bypass the two of the spdt switches in the transfer switch to switch these two circuits back to line input whenever the solar power drops below a set reference, presently controlled by a small solar panel connected to a resistive load.

The following is still under construction:

My grid tie inverter will shut down when it detects it's output power is higher than one leg of the split phase grid power being imported.
A current pick up coil on the connection line going low/near zero power can be used a the logic for directing relay controls

if your grid tie inverter has no such disconnect features one caan easily be built as follows.

By clamping identical power pick up coils to both the incoming grid power and to the output of the grid tie inverter, rectifying both and feeding the outputs to a simple comparator circuit(s) should provide the necessary logic to interface to the relays. Since I am on a split phase grid two such comparators will be necessary with the outputs logically OR’ed.

I’d like to believe that I am trying to reinvent the wheel. Perhaps someone on the forum is aware of some device(s) that may be available to simplify the task.

The relays are rated 25 amps dc and my strings max scc is about 20 amps. The max I've logged to date was 17 amps per string.
My strings are 3s2p 180 watt panels.
I am using a thornwave powmon 5s and setting the relay to generator control using soc.
Since I use my 4.8 Kwh battery bank as backup for power outages, I set the switch points to 90 % soc and 95 % soc.

 

[fafrd]

I am reposting my last comments from the other topic in hopes we can get a few more, perhaps more eknowledgible people involved.

I have installed two Functional Devices Inc. RIB24P30 dpdt relays.
One controls the outputs of two solar strings either to charge my battery bank via mppt controllers or connect to a grid tie inverter based on the soc of my batteries. If the state of charge drops below 90 % (as measured with a smart shunt) the power is directed to charge the batteries. When soc rises to 95 %. it switches to grid tie. As a backup my off grid inverter/charger will go into charge mode when the soc drops below 20 %.

The batteries drive an off grid inverter/charger which outputs to a 6 circuit transfer switch.

The other relay can bypass the two of the spdt switches in the transfer switch to switch these two circuits back to line input whenever the solar power drops below a set reference, presently controlled by a small solar panel connected to a resistive load.

The following is still under construction:

My grid tie inverter will shut down when it detects it's output power is higher than one leg of the split phase grid power being imported.
A current pick up coil on the connection line going low/near zero power can be used a the logic for directing relay controls

if your grid tie inverter has no such disconnect features one caan easily be built as follows.

By clamping identical power pick up coils to both the incoming grid power and to the output of the grid tie inverter, rectifying both and feeding the outputs to a simple comparator circuit(s) should provide the necessary logic to interface to the relays. Since I am on a split phase grid two such comparators will be necessary with the outputs logically OR’ed.

I’d like to believe that I am trying to reinvent the wheel. Perhaps someone on the forum is aware of some device(s) that may be available to simplify the task.

The relays are rated 25 amps dc and my strings max scc is about 20 amps. The max I've logged to date was 17 amps per string.
My strings are 3s2p 180 watt panels.
I am using a thornwave powmon 5s and setting the relay to generator control using soc.
Since I use my 4.8 Kwh battery bank as backup for power outages, I set the switch points to 90 % soc and 95 % soc.

The only DC rating I see on those relays states:

‘25 Amp Resistive @ 28 Vdc’

Does that mean your Voc or at least your Vmp is under 28Vdc?

I’m using 380W panels with voltage levels higher than that and the relays/contractors that can handle higher voltages all seem to be even more expensive.

So I’m starting to think that rather than trying to reroute DC inputs from solar panels, it might be easier to reroute AC output from Microinverters to an AC battery charger or other dump load.

600W @ 240VAC is only 2.5A, easily switchable with a pair of inexpensive relays.

Another idea I’m considering would be to connect panels to an SCC and switch the DC SCC output between a Microinverter or charging the battery. SCC output can be kept under 28V and the MPPT of the inverter should settle on whatever boost voltage the SCC has been set to.

 

[McKravitts]

The only DC rating I see on those relays states:

‘25 Amp Resistive @ 28 Vdc’

Does that mean your Voc or at least your Vmp is under 28Vdc?

My vmp is about 54 volts.
The general rule of thumb for dc relays is to multiply 28 times the rated current to get max dc watts and then try to not exceed that figure in watts.
I'm going to parralel both poles of the relay to get a maximum safe operating current of 50 amps and max wattage of 1400.
The absoute maximum any one string of my system will producs is 1080 watts.
Even if the relay contacts aren't perfectly balanced it should still be safe.
I'm only using one pole in my protype but the maximum power I measured so far is 900 watts on one string, about 30 % above capacity.
Relays aren’t like fuses. They don’t blow when you exceed their rated current. You just shorten the useable life.
I got them on ebay pretty cheap and figured they'd be good prototyping relays and if it works I can parallel them.

I’m using 380W panels with voltage levels higher than that and the relays/contractors that can handle higher voltages all seem to be even more expensive.

So I’m starting to think that rather than trying to reroute DC inputs from solar panels, it might be easier to reroute AC output from Microinverters to an AC battery charger or other dump load.

600W @ 240VAC is only 2.5A, easily switchable with a pair of inexpensive relays.

That's not a bad idea.
Do you have microinverters that can operate both in grid tie mode and in off grid mode or are you contemplating additional circuitry?

Another idea I’m considering would be to connect panels to an SCC and switch the DC SCC output between a Microinverter or charging the battery. SCC output can be kept under 28V and the MPPT of the inverter should settle on whatever boost voltage the SCC has been set to.

I considered that as well.
Most of the inexpensive grid tie inverters I looked at all have built in mppt charge controllers.
I'm guessing two charge controllers in series will probably reduce the efficiency overall.
Have you considered a high efficiency bucking converter on the solar output rather than a scc.

 

[fafrd]

My vmp is about 54 volts.
The general rule of thumb for dc relays is to multiply 28 times the rated current to get max dc watts and then try to not exceed that figure in watts.

Didn’t know that, thanks. So ta 25A 28VDC-rated relay can handle ~700W.

I'm going to parralel both poles of the relay to get a maximum safe operating current of 50 amps and max wattage of 1400.

So 2 25A relays on positive and 2 25A relays on negative. It has also not occurred to me to double-up relays for higher current capacity.

The absoute maximum any one string of my system will producs is 1080 watts.
Even if the relay contacts aren't perfectly balanced it should still be safe.
I'm only using one pole in my protype but the maximum power I measured so far is 900 watts on one string, about 30 % above capacity.

Confused by what you mean here. 900W is 64% of your 1400W relay capacity - do you mean ‘more than 30% below max capacity’?

Relays aren’t like fuses. They don’t blow when you exceed their rated current. You just shorten the useable life.
I got them on ebay pretty cheap and figured they'd be good prototyping relays and if it works I can parallel them.

Yeah, $60 is more than 10 times the price on Amazon for the DIN-mounted 10A / 280W relays so hopefully you picked them up close to $20…

Now that you’ve turned me on to the possibility to parallelize relays, a one pole-relay rated to 1400W would need 5 of those 10A relays at a cost of ~$30.

My 380W panels max out at 340W so I could switch one of them using 2 10A relays in parallel.

I’m going to have to think about this some more…

That's not a bad idea.
Do you have microinverters that can operate both in grid tie mode and in off grid mode or are you contemplating additional circuitry?

The Microinverters are grid tie (as are my GTIL inverters. Off-grid and backup power is not a priority for me. I built the small 1kW DC-coupled system to provide backup power during an extended power outage but don’t need that to be automatic. That is a separate 3kW battery inverter I only turn on during rare extended outages.

Now that I’ve got that additional solar power on the roof, I’m looking for ways to add some output to my AC-coupled solar production without surpassing my 3.5kW export cap.

So the Microinverters will only be operating when grid-tied and my issue is how to throttle them back as the larger SW-facing Microinverter powers up near peak production levels on clear summer days…

I considered that as well.
Most of the inexpensive grid tie inverters I looked at all have built in mppt charge controllers.

Yes, pretty much any Microinverter will have an MPPT.

I'm guessing two charge controllers in series will probably reduce the efficiency overall.

SCCs have very high efficiency (93-96%) but you are correct, a Microinverter driven from an SCC would deliver lower efficiency than driving that Microinverter directly off of the solar.

Since I’ve got both an available Microinverter and an SCC, that was the appeal of using relays:

Overdrive the Microinverter with full DC-coupled array until it nears saturation and then switch one panel at a time over to the SCC in the morning and then back over in the afternoon…

Need to cogitate on this a bit more (though very interested to learn from your experience as you get your system working).

 

[McKravitts]

Didn’t know that, thanks. So ta 25A 28VDC-rated relay can handle ~700W.

That's just a rule of thumb. In reality there are many complicating factors involved in relay design. When contacts close for a very tiny fraction of a second there is an arc between the contacts and they act much like an arc welder, transferring material from one contact to the other. With AC current the transfer will go back and forth, but DC current transfers in one direction only which causes pitting on one of the contacts and eventually leads to failure. See https://en.wikipedia.org/wiki/Contact_protection.

So 2 25A relays on positive and 2 25A relays on negative. It has also not occurred to me to double-up relays for higher current capacity.

Not exactly. The two legs will never make contact at the exact same time, so one will wear down faster than the other. You can increase the overall life by paralleling, but probably not double it.

I'm using a common negative, so I only switch the positives. All my equipment is suited for a negative common ground.
The negatives from all my strings are connected together on a common bus bar to which the scc's and inverters etc. negatives are also connected.
Most manufacturers specify their equipment compatibility with common grounding schemes.

Confused by what you mean here. 900W is 64% of your 1400W relay capacity - do you mean ‘more than 30% below max capacity’?

The maximum rated output power of solar panels is rarely achieved, at least in my area. Humidity, haze, air polution, solar angle etc. reduce the power. I have 6 180 watt panels per string. 6 times 180 is 1080. That is the maximum output per string that I could ever achieve. To date I haven't measured any more than 900 wats, so I'm using that figure for relay sizing.

So the Microinverters will only be operating when grid-tied and my issue is how to throttle them back as the larger SW-facing Microinverter powers up near peak production levels on clear summer days…

I was thinking about that approach you suggested in your previous post regarding transferring the ac power into a battery charger.
Since the micro inverters must be grid tied, it would be very difficult and likely impossible to control where the power is going, however you could monitor it and use that measurement to modulated the power taken from the grid for battery charging. It would take some circuitry design but it is feasible.
A device like that would assure that you always take more from the grid than you put into it.
It would greatly simplify what we’re trying to achieve and be applicable to a variety of configurations.

 

[fafrd]

That's just a rule of thumb. In reality there are many complicating factors involved in relay design. When contacts close for a very tiny fraction of a second there is an arc between the contacts and they act much like an arc welder, transferring material from one contact to the other. With AC current the transfer will go back and forth, but DC current transfers in one direction only which causes pitting on one of the contacts and eventually leads to failure. See https://en.wikipedia.org/wiki/Contact_protection.

Not exactly. The two legs will never make contact at the exact same time, so one will wear down faster than the other. You can increase the overall life by paralleling, but probably not double it.

Got it. So there is some benefit to going with a single larger relay. Or you could just go with more smaller/cheaper ones in parallel and check for the first failure on some maintainance schedule - is there an easy way to know when a relay has failed?

I'm using a common negative, so I only switch the positives. All my equipment is suited for a negative common ground.
The negatives from all my strings are connected together on a common bus bar to which the scc's and inverters etc. negatives are also connected.
Most manufacturers specify their equipment compatibility with common grounding schemes.

If you can use a common ground, the other idea I’ve considered is to use only a single relay with a diode connection to the DC String / SCC. This assumes you have one DC string to an SCC, a second DC string to an MPPT such as a Microinverter, and a third DC string you want to switch from the Microinverter to the SCC once the power to the Microinverter gets high enough (approaches saturation).

The diode means that as long as the Microinverter MPPT is at the same voltage as the SCC MPPT, it will get all of the power from the switched string.

A shunt on the string permanently feeding the Microinverter can decide when the switched panel is no longer needed and used to open the relay.

Once the relay is open, power from the switched panel will flow to the SCC (minus 0.7V/Vmp%).

The beauty of this is that the closed relay can be controlled to only pass a limited morning level if current. Peak generation currents only flow through the diode.

Two MPPTs on the same string is always a big question but if the switched string is smaller that the permanent strings it gets added to, there will be a limit on any ‘race to the bottom’ and the MPPT getting power through the rely will always win out.

I’ve got exactly that situation (3 DC strings) so I may need to revisit that concept with the new information you’ve provided regarding relays. (As always, the question is whether there is enough payoff in terms of increased power generation to justify the added cost and complexity…).

The maximum rated output power of solar panels is rarely achieved, at least in my area. Humidity, haze, air polution, solar angle etc. reduce the power. I have 6 180 watt panels per string. 6 times 180 is 1080. That is the maximum output per string that I could ever achieve. To date I haven't measured any more than 900 wats, so I'm using that figure for relay sizing.

Yeah, my 1.14kW array has never put out more than 1.0kW (88%).

I was thinking about that approach you suggested in your previous post regarding transferring the ac power into a battery charger.
Since the micro inverters must be grid tied, it would be very difficult and likely impossible to control where the power is going, however you could monitor it and use that measurement to modulated the power taken from the grid for battery charging. It would take some circuitry design but it is feasible.

A simple current switch plus a 120VAC relay is all that is needed. Since I know my maximum export limit and I know to the saturated output of my additional Microinverter, I can monitor generation off of my large array and activate the relay / battery charger once it get’s closer than Export Limit minus Saturated Microinverter.

So triggering the ‘dump load’ will be quite simple and relatively inexpensive (<$20 all-in). [more on the battery charger as dump load below]

A device like that would assure that you always take more from the grid than you put into it.
It would greatly simplify what we’re trying to achieve and be applicable to a variety of configurations.

Well firstly, at a high level, this is precisely what the Smart EV chargers do. They monitor export and adjust EV charging power to maintain export below a small threshold (or maintain consumption below a small threshold).

I’ve been looking for / waiting for an AC battery charger that can do the same thing (controllable charging power) but none exist (at least not yet).

The Schneider Conext XW has a controllable battery charger and a protocol that can be used to change the charging power, but that’s the only non-EV charger I’m aware of with that capability today (many DC chargers can do this, but I’m talking about AC chargers).

The Enphase battery may do it as well, but only in an all-Enphase system which does not interest me.

So hence my idea to turn in an AC battery charger as a dumpload. It’s very crude, but if the amount of excess generation you are concerned about is roughly equal to the battery chargers power, it’s a nice way to capture most of that excess energy in a battery rather than throttle it back or truly dump it.

Turning on and off the ‘battery charger dump load’ is straightforward, as I wrote, but there is an additional problem to solve: it does not act as a dump load once the battery is fully charged.

So my plan is to have a second current switch in the dump load itself monitoring the current going to the battery charger. Once the battery charger enters Float and charge power begins to decrease, that second current switch will activate a second relay to a true dump load such as a resistor.

I’m trying to handle 300W of excess peak generation, so once the main AC-coupled array puts out over Export Limit - 300W, it’ll activate the battery charger dump load and the battery will start charging at a consumption of 320W.

If the battery charges enough to drive the battery charger out of Boost into Float, consumption will drop below 300W at which point the second current switch will close a second relay to turn in a 300W load.

Once the large AC-coupled array drops below Export Limit -300W in the afternoon, the first current switch opens the relay and the battery charger dump load is turned off (meaning any remaining power from the additional generation source (Microinverter in my case) gets added to power being exported.

So I’m in the market for cheap and reliable ways to dump 300W of 120VAC power. I’ve found 100 Ohm power resistors supposedly rated for 100W, so a 2S4P network of those should consume 288W with no individual resistor seeing more than 36W.

But it seems llike U’ll need to mount those power resistors in a heatsink, etc, so I’m very interested in easier ways to dump 300W…

 

[fafrd]

Looks like $25 all-in is my budget for a 300-350W 120VAC dump load: https://www.amazon.com/heater-Progr...cphy=9032080&hvtargid=pla-1598031172706&psc=1

The dump load will be needed primarily in summer, so I’d rather dump the energy by cooling rather than heating, but it’s a start.

 

[McKravitts]

Got it. So there is some benefit to going with a single larger relay. Or you could just go with more smaller/cheaper ones in parallel and check for the first failure on some maintainance schedule - is there an easy way to know when a relay has failed?

The contacts fuse together and it stays in a slosed state.

Well firstly, at a high level, this is precisely what the Smart EV chargers do. They monitor export and adjust EV charging power to maintain export below a small threshold (or maintain consumption below a small threshold).

That's news to me. I never had a solar ev charger, but I'm definitely gogint o look into it.

I’ve been looking for / waiting for an AC battery charger that can do the same thing (controllable charging power) but none exist (at least not yet).

Every off grid inverter/charger I ever owned had a small potentiometer that adjusted the charge rate.

I just checked the specs on one unit and they claim the charge rate can be adjusted from between 0 and 100 %.

There are several grid tie inverters on the market that have what they call limiters. They limit the amount of power they’ll output to the grid so as to not export.

If I could get the schematics of the units I could design a simple interface circuit to control the charge rate of the off grid inverter/charger to accomplish the task.

You could leave a few circuits on the off grid so as to maintain available battery capacity.

I’d rather dump the energy by cooling rather than heating, but it’s a start.

I made my own whole house air conditioner that is super-efficient. It only takes about 700 watts and outputs about 30,000 BTU of cooling. You have to be in an area of the country where you have access to ground water 60 degrees F or colder.

 

[fafrd]

The contacts fuse together and it stays in a slosed state.

So once a relay has failed you will be shorting the string to both MPPTs which is probably not catastrophic and should be easy to detect. But I suppose that if you’ve got 2 or 3 relays in parallel, it means two times or three times the current through single failed relay (probably not good).

That's news to me. I never had a solar ev charger, but I'm definitely gogint o look into it.

These are Smart EV chargers (no solar required): https://uploads-ssl.webflow.com/5ff...395847d4b83206_EV Charger Technical Specs.pdf

Every off grid inverter/charger I ever owned had a small potentiometer that adjusted the charge rate.

I just checked the specs on one unit and they claim the charge rate can be adjusted from between 0 and 100 %.

Would appreciate a link - are you talking about an AC battery charger or an inverter?

There are several grid tie inverters on the market that have what they call limiters. They limit the amount of power they’ll output to the grid so as to not export.

Yes, I know. I’ve got two GTIL inverters offsetting my loads as we speak (thhe y work great and are fantastic value for money).

What I want is something similar on the consumption side: use CT sensors to monitor export and charge battery with exactly the power needed to offset export.

If I could get the schematics of the units I could design a simple interface circuit to control the charge rate of the off grid inverter/charger to accomplish the task.

If you know enough to do that, you just may be my hero!

The key is a charger that can have it’s charge power and power consumption controlled by an electronic signal (potentiometer or whatever). I’ve searched but have not found anything.

If you have a unit like that, developing a circuit that controls that signal through CT sensors should not be too difficult.

Please share a link to whatever charger you are talking about.

You could leave a few circuits on the off grid so as to maintain available battery capacity.

I don’t actually have any ‘off-grid’.

I have AC-coupled solar and then I have a small DC-coupled solar array that charges a 24V battery all day long.

When peak hours start, I turn on my 2 GTIL inverters (one per 120VAC Leg) and they begin offsetting consumption (they are both grid-tied to their respective legs).

The GTILs will keep offsetting loads until the battery is depleted, generally past midnight.

So I have no ‘off grid’ power to switch loads to, but I’m now starting to focus on adding an air conditioner that can be turned on automatically anytime too much power is being exported (which will generally only occur in the summertime).

The AC-coupled array is 4kW and works great.

I just added the 1kW SC-coupled array last year and it’s also working great (generated 1400kWh of load offset over the past year(.

I’m now looking for ways to add another 1kW which will mean squeezing out some more AC-coupled power within the remaining gap below my export limit and filing up the remaining headroom to charge my battery up more fully….

 

[fafrd]

I just discovered this battery charger from another thread: https://www.thunderstruck-ev.com/tsm2500-and-evccbasic.html

This is basically an EV charger that can charge 48VDC or even 24VDC batteries.

I spoke to the engineer and they support the full EVSE protocol meaning charge power can be controlled through parameters.

The only question is how quickly charge power can be updated (Emporium updates their charge power every 2 seconds) but the 48V charger consumes a maximum of 2kW which means startup power should be 200W and it should be able to adjust in 20W increments.

I don’t want to spend $500 just for a dump load, but this charger may form an important part of my future Zero Export system…

 

[fafrd]

As I’m going through the various options here, the winning solution has become a no-brainer.

I’ll activate my battery as a dump-load and when the charger moved out of Boost and into Float, I’ll use another current sensor to activate a second relay controlling an air conditioner.

So ‘Battery-Charger-Dunpload-ON and Battery-Nearing-Full = Turn on Air Conditioner as Dump Load.

We were thinking about getting a small portable air conditioner anyway, and the battery only has any chance of getting fully-charged on a clear summer day when the house will be hotter than we’d like anyway.

It makes much more sense to dump energy by cooling the house than heating the basement.

Only little detail I need to work out is how to control the Air Conditioner outlet from a relay in the basement…

The unit I’m considering consumes 1200W, so more than enough dump load to avoid over-export.

In fact, at that level of load, the AC may drain the battery enough to restart the charger, so I need to think about adding hysteresis or a timer to avoid thrashing…

 

[fafrd]

As I’m going through the various options here, the winning solution has become a no-brainer.

I’ll activate my battery as a dump-load and when the charger moved out of Boost and into Float, I’ll use another current sensor to activate a second relay controlling an air conditioner.

So ‘Battery-Charger-Dunpload-ON and Battery-Nearing-Full = Turn on Air Conditioner as Dump Load.

We were thinking about getting a small portable air conditioner anyway, and the battery only has any chance of getting fully-charged on a clear summer day when the house will be hotter than we’d like anyway.

It makes much more sense to dump energy by cooling the house than heating the basement.

Only little detail I need to work out is how to control the Air Conditioner outlet from a relay in the basement…

The unit I’m considering consumes 1200W, so more than enough dump load to avoid over-export.

In fact, at that level of load, the AC may drain the battery enough to restart the charger, so I need to think about adding hysteresis or a timer to avoid thrashing…

It’s a kludge but a least it’s a baseline solution:

I can run speaker wire from the basement to the living room where we’d want the AC to run, so a small switched power box containing a relay should be all I need…

Of course, I still need to figure out how to run a vent to the outside (but one problem at a time ).

 

[McKravitts]

It sounds like a reasonable approach you're taking.

Since I have both a grid tie and an off grid tie inverter, and I need to maintain a reasonable soc because of the frequent outages I have in my area.
My goal is to do the following.



If battery soc is above 90 % switch to grid tie.

If grid tie approaches net export, switch to battery charging using off grid inverter charger.

If soc rises to 95 % switch a circuit on from transfer switch driven by off grid inverter.

If soc rises to 97 % switch on another circuit.

If soc rises to 100 % and still at export levels, consider selling some panels or buy an EV.



I’ll keep you updated.

 

[fafrd]

It sounds like a reasonable approach you're taking.

Since I have both a grid tie and an off grid tie inverter, and I need to maintain a reasonable soc because of the frequent outages I have in my area.
My goal is to do the following.



If battery soc is above 90 % switch to grid tie.

If grid tie approaches net export, switch to battery charging using off grid inverter charger.

If soc rises to 95 % switch a circuit on from transfer switch driven by off grid inverter.

If soc rises to 97 % switch on another circuit.

If soc rises to 100 % and still at export levels, consider selling some panels or buy an EV.



I’ll keep you updated.

When you ‘switch to grid tie’ you mean switching panels over from SCC or off-grid inverter to Microinverter or String Inverter, correct?

At that point, our two systems are very similar.

How are you determining battery SOC?

And how are you determining when you are ‘approaching net export’

 

[McKravitts]

When you ‘switch to grid tie’ you mean switching panels over from SCC or off-grid inverter to Microinverter or String Inverter, correct?

Yes

How are you determining battery SOC?

I have a Thornwave powmon ds5 shunt that has an output that can switch a relay based on soc.

And how are you determining when you are ‘approaching net export’

My Y & H grid tie inverter has a limiter that shuts down the inverter before it reaches export. This drops the ac current on the line from the sub panel to a small amount. Right now I'm just using the voltage output of an ac clamp meter, rectifying that output and feeding a modular comparator which has a 10 amp relay which provides the necessary switching logic.

 

[fafrd]

Yes

I have a Thornwave powmon ds5 shunt that has an output that can switch a relay based on soc.

Nice. I was planning to use battery voltage which is a poor-man’s substitute for a shunt but I’m now thinking about using A/C current consumption of an A/C charger, since consumption drops dramatically once the charger exits CC Mode and enters CV mode…

My Y & H grid tie inverter has a limiter that shuts down the inverter before it reaches export.

I’ve got a couple GTILs from Y&H as well. Are you just referring to the CT sensor input or some other feature / setting?

This drops the ac current on the line from the sub panel to a small amount.

I’m not understanding your overall system architecture - is ‘subpanel’ grid-tied or tied to your off-grid power? I understand how a GTIL inverter can offset most all power consumption of a subpanel, but don’t understand whether that subpanel is tied to the grid or tied to your off-grid power system…

Right now I'm just using the voltage output of an ac clamp meter, rectifying that output and feeding a modular comparator which has a 10 amp relay which provides the necessary switching logic.

And this last section is very interesting to me but I don’t understand it.

Is this ‘AC clamp meter’ the standard CT sensor that came with your Y&H GTIL or a seperate CT sensor you purchased?

How are you ‘rectifying’ the output of that CT sensor? Using an OTS product or did you rig up a rectifying circuit yourself?

I’m assuming you rectifier circuit just takes the sine-wave AC current output of your CT sensor and turns that into a square wave, correct?

So I’m not understanding whether it’s just a 50% duty-cycle output or there is some sort of adjustable current threshold or voltage threshold involved so that you are getting a duty cycle that reflects some measure ‘closeness to zero export / current’???

And I’m also not understanding what the ‘modular comparator’ is - is that a comparator that has a 0 or 1 output depending on whether input duty cycle is above or below a tunable duty-cycle threshold?

From a high or low signal controlling a 10A relay, I’ve got no questions but I’ve been struggling to understand whether there is an easy way to use CT sensors for power control short of putting together a full-blown Rasberry-Pi-based programmable energy monitor.

Sounds like you’ve got some of this control working using relatively easy-to-find control boards and I’d like to understand in enough detail to possibly build something similar…

[EDIT: OK, I followed the link to the comparator so I understand how it is both an adjustable voltage comparator and a 10A relay integrated together. So are you putting the AC current of the CT sensor through a resistor to generate a dime-wave whose amplitude is proportional to AC current and then using this adjustable voltage comparator to duty cycle the relay output based on a voltage threshold?

If there is any tutorial on the approach that inspired you, a link would be great. And if it’s an original idea you came up it’s in your own, a few waveforms would be worth a thousand words ]

 

[McKravitts]

Nice. I was planning to use battery voltage which is a poor-man’s substitute for a shunt but I’m now thinking about using A/C current consumption of an A/C charger, since consumption drops dramatically once the charger exits CC Mode and enters CV mode…

I tried using battery voltage but I encountered a few problems.
I want to keep my batteries at a high soc since I need them for power outages. The voltage between on/off points is very small. Switching off the scc, will drop the voltage enough to cause the scc to switch back on. The system goes into a fairly rapid on/off loop.
If you're Ok with a deep discharge, the method will probably work, but remember battery life is shortened by deep discharge cycles.

I’ve got a couple GTILs from Y&H as well. Are you just referring to the CT sensor input or some other feature / setting?

It's the limiting feature. It uses a CT sensor to determine if it is in net export and reduces output power. I'm adding my own sensor to the line to measure when the output is near zero. This could be caused by things other than limiting and using an integral of the power output might be more appropriate. I'll have to think about that.

How are you ‘rectifying’ the output of that CT sensor? Using an OTS product or did you rig up a rectifying circuit yourself?

I’m assuming you rectifier circuit just takes the sine-wave AC current output of your CT sensor and turns that into a square wave, correct?

So I’m not understanding whether it’s just a 50% duty-cycle output or there is some sort of adjustable current threshold or voltage threshold involved so that you are getting a duty cycle that reflects some measure ‘closeness to zero export / current’???

I had an electronics business many years ago, so I have boxes full of discrete components that I still use for simple circuits. I just used a diode, a capacitor, and a couple resistors. The input of the comparator is high impedance, so the signal is just a slowly varying DC that tracks the AC current level.

And I’m also not understanding what the ‘modular comparator’ is - is that a comparator that has a 0 or 1 output depending on whether input duty cycle is above or below a tunable duty-cycle threshold?

I'm using the word comparator to mean an over the counter ic chip. It isn't hard to build one from transistors, diodes, resistors etc, but ic comaprators are a dime a dozen and have a very sharp on/off curve.
These days you can purchase an adjustable comparator circuit complate with switching relay.
I bought this one on amazon: CZH-LABS DIN Rail Mount Voltage Comparator Relay Module, DC12V, SPDT 10Amp Relay.
They refer to it as a module, so I called it modular.

 

[fafrd]

I tried using battery voltage but I encountered a few problems.
I want to keep my batteries at a high soc since I need them for power outages. The voltage between on/off points is very small. Switching off the scc, will drop the voltage enough to cause the scc to switch back on. The system goes into a fairly rapid on/off loop.

Yeah, thrashing is no good. I’m currently just using voltage to determine when to shut down the inverters at night. My restart voltage is high enough that I never need to worry and the inverters turning back on after bounce-back.

If you're Ok with a deep discharge, the method will probably work, but remember battery life is shortened by deep discharge cycles.

Backup is not a concern for me, so my cycle consists of charging the batteries starting with morning sun, turn on the GTIL inverters at the start of peak period (4pm) and continue to the end of peak period or when the battery depletes to low voltage cut off.

So I discharge my battery every night but a ‘deep discharge’ (below 25.6V) is relatively rare.

It's the limiting feature. It uses a CT sensor to determine if it is in net export and reduces output power.

Yes, I’m using the same feature.

I'm adding my own sensor to the line to measure when the output is near zero.

So you mean you are going to use GTIL output approaching 0W / 0A to determine when you are nearing the start of export, correct?

This could be caused by things other than limiting and using an integral of the power output might be more appropriate. I'll have to think about that.

Monitoring GTIL output power sounds like a good approach and the easiest way to go so may be to use a current switch: https://www.amazon.com/gp/aw/d/B07W...d2lkZ2V0TmFtZT1zcF9waG9uZV9kZXRhaWwp13NParams

I had an electronics business many years ago, so I have boxes full of discrete components that I still use for simple circuits. I just used a diode, a capacitor, and a couple resistors. The input of the comparator is high impedance, so the signal is just a slowly varying DC that tracks the AC current level.

So just to understand, the diode allows you to clip off the negative current so then the half-sine-wave is put through a resistor and filtered by a capacitor to give a crude estimate of AC current in the form of a voltage, correct?

I'm using the word comparator to mean an over the counter ic chip. It isn't hard to build one from transistors, diodes, resistors etc, but ic comaprators are a dime a dozen and have a very sharp on/off curve.
These days you can purchase an adjustable comparator circuit complate with switching relay.
I bought this one on amazon: CZH-LABS DIN Rail Mount Voltage Comparator Relay Module, DC12V, SPDT 10Amp Relay.
They refer to it as a module, so I called it modular.

Understand. You use ‘modular’ to mean what I mean when I say Off The Shelf (OTS).

The current to voltage transform is very crude but since you are only interested in the current approaching zero and since the comparator gives you a 1/0 output based on a settable threshold, I see how it is effective.

I’ve got an export limit of 3.5kW and am interested in ways to trigger a dump load when export approaches that limit, but that’s a tougher problem and I don’t the same technique can work.

 

[McKravitts]

So you mean you are going to use GTIL output approaching 0W / 0A to determine when you are nearing the start of export, correct?

Yes.

So just to understand, the diode allows you to clip off the negative current so then the half-sine-wave is put through a resistor and filtered by a capacitor to give a crude estimate of AC current in the form of a voltage, correct?

Prior to the invention of switching power supplies, diodes and capacitors were used extensively in power supplies. The amount of ripple depends to a great extent on the load. For a high impedance comparator input the output voltage is very stable, unless the capacitor is leaky.

By using a string of series capacitors separated by inductors, you can get a very steady dc output voltage. It can be further stabilized by using a full wave bridge rectifier and a single transistor. For what I'm doing the single diode and a 1000 uF capacitor are more than adequate.

Understand. You use ‘modular’ to mean what I mean when I say Off The Shelf (OTS).