A/C outflow monitor and control for grid tied solar.

[youmanskids]

hey all,  I have a control question for you guys.
I have 4 KW of solar grid tied with Enphase micro inverters,  I also have a back-up battery bank of 1050 Ah (with an 10 foot Dans trubine).  My goal is to use excess power from the panels to charge the batteries during the day and use it in the pm to run the house. THE PROBLEM is that I need something that can monitor the outflow of power to the grid and turn on the charger when flow is out, but shut it off when flow in inward. The charger/inverter I am using is a Magnasine 4024 (24v, 4 kw) which has and adjustable charge rate via it's control panel (not sure how to intigrate a control system into this function either)
  Does anyone out there have a way to accomplish this-  Ive tried contacting Magnasine company for an off-the-shelf add on to do this but get no response.
If anyone has used a back-up battery bank with a grid tied system like this I sure would appreciate any input or ideas to optimize my power use from the panels (the reason to do this is the fact that I pay the city 11 cents/Kwh for what I use, but they only pay me 3 cents/Kwh for what I produce)
I hope this makes sense to someone,  thanks for any help.  Roger

[DamonHD]

I have similar plans but have yet to solve the problem.  My input cost is ~15p and export 3p, so we're in a similar position, though my issue is about minimising carbon footprint rather than bills.

Rgds

Damon

[youmanskids]

At least I'm not the only one with this problem....  I am also wanting to limit carbon footprint, but why not save money too?..  But, ? nobody out there has anything that can monitor a/c flow direction from the house and use it to turn a charger on and off?     I've seen a few threads about using a transformer for this, but the discussion was over my head.  I'm not an electrical engineer (but consider myself reasonably intelligent) and need a simple solution/application.  Hoping someone can help me out.  thanks in advance for any help.  Roger

[Phil Timmons]

Sure.  This sounds like a good project.  I am off the road for a while and this could be a little fun.  Something like this may already exist, but we will not discover that until we start to build one, if this is typical.  

Thing we are looking for is the Voltage and Current to be in-phase (rising and falling together) either going upstream or downstream, and they should be 180 degrees out-of-phase when current switches direction.

So we could monitor in-phase v. out-of-phase relationship, as you noted, with a couple of small transformers -- one a small potential (voltage) transformer to step things down to where we can put an op-amp on it, and the other transformer a small current (measures current or amps) transformer, which again we would put to an op-amp.  

Then we compare the peak voltage to the current and if they are both above zero (in phase) we know current is going one direction, and if one is above zero, while the other below (out-of-phase) we know the flow is going in the other direction.

We can add a little slop in there to allow for some Power Factor issues (could cause slightly out of phase Lead or Lag) and only really compare the peak portions of the waveforms.

The comparison circuit could drive either a power transistor or more likely a relay to cut the charger on and off.  

Now let's sit back for a day or two and let folks tear that up, and then we can do the circuits after we see what of that concept survives the peer review of the peanut gallery.  

[boB]

After you determine which direction the power flow is and the magnitude of that power, then you can control the charging amount of the Magnasine charger by tapping into its communications.  Magnum Energy publishes their RS-485 protocol on their web site and it's not too difficult if you are into such types of design projects.

I am surprised that nobody answered your email.  At least you don't need to talk with anyone to get ahold of that protocol.

boB

[ghurd]

Let the Tearing-Upping begin!  

1050AH at 24V?
The battery will not take 4000W input very long before it is at regulation voltage.  Especially if the batteries are not being abused.

Seems like it could be done easier.  Not exactly as efficient, but I think I could argue pretty close.

Different line of thought:

-  Since the batteries are NOT being abused (assumed), it will not take a lot of power to get them up to float.

-  Simple Day/Night sensor circuit (LDR and 741) triggers when the sun comes up.  It operates a relay in the AC line to the charger.  The charger can only get AC power in the day time.

-  Something like a 'ghurd controller' on the battery?  It could operate a second (in series) relay in the AC line to the charger.  The charger would not be sucking AC power if the batteries did not need some charge.  
If the 10' has been making 750W for the last 36 hours, and the battery dump load controller has been dumping for the last 32 hours, no reason to have a fancy AC charger sucking up 75W to monitor the battery.

-  Solar is working best when you do not need the power.  Usually?  
Where are you at 11:00AM and 2:00PM?  At work?  Outdoors?  The 4KW of solar is chugging along at its best, and nobody is in the house, opening the fridge, makeing coffee, running water (well pump), 1850W hair dryer, with 3 computers, a TV, and 8 lights on.

Short version.  The AC charger Only operates in daylight, and Only if the battery needs a bit of charging.

Not exactly an "In vs Out" circuit, but it sure would be easier to make!
G-

PS-  boB posted while I was typing.

[Phil Timmons]

After you determine which direction the power flow is and the magnitude of that power, then you can control the charging amount of the Magnasine charger by tapping into its communications.  Magnum Energy publishes their RS-485 protocol on their web site and it's not too difficult if you are into such types of design projects.

I am surprised that nobody answered your email.  At least you don't need to talk with anyone to get ahold of that protocol.

boB

Very nice. 

I was at first thinking just hammer it on or off depending on whether power was heading upstream or down -- but then got to thinking at some marginal levels, whether charger was on or not might make the difference as to whether we would be sending power up or pulling it down from the grid -- so then it would sit in oscillation, turning the charger on and off, unless we put in a threshold for current the charger would draw. 

But, like you say, if we can vary the charger current, that would take care of that. 

Of course, actually monitoring the magnitude of the service current rather than just direction gets the circuit more complicated, but yunno around here this is going to turn into a full house energy management system with its own independent power source and backup Thorium Fueled reactor in the backyard by the time we are done, anyway.    

[Phil Timmons]

Let the Tearing-Upping begin!  

Yeah, I love this place.  

1050AH at 24V?
The battery will not take 4000W input very long before it is at regulation voltage.  Especially if the batteries are not being abused.

Yeah, what is the (shotgun numbers) expected max. current level (AC side of the charger) we are talking about anyway?

Seems like it could be done easier.  Not exactly as efficient, but I think I could argue pretty close.

Reason I wandered towards this is the application is a good deal broader than just the charger on/off feature this started with.  The same system that would shut the charger off -- depending on direction of current flow -- can be used to shut off other non-mission critical loads, as well.  Sort of "load shedding" at the household circuit level.

As you observed there are some predictable Time of Production matters.  There are also some Time of Use (for us still Grid-Tied) matters creeping across the country.  In some areas the summer daytime Peak Time of Use (TOU) is proposed to be somewhere around 20 to 40 cents per kWh.  So while a PV system or wind system may be cranking out during that same time, at those high rate peak prices, suddenly it matters whether we are always a net-producer in the "peak" time period, as well.  

The perceived economics are what is driving Roger (in the OP), as well.  His concern was because is only getting "avoided costs" (typically generation fuel) from his utility for the power he produces, while he has to buy at full retail.  

For the broader TOU application -- if we see we are drawing in current at the metered billing point of the service during those peak times, we start instantly dumping loads until we get to net production -- drop off the dryer, the stove, the pool pump, on and on, maybe even . . . . egad, the Air Conditioning.  

But like Bob observed, maybe this requires monitoring magnitude of current, as well as direction, and every time I have started wandering into this, I wind up with an entire paper energy monitoring and control system, and was just hoping for something faster, cheaper, quicker.  A girl can dream, huh?  

[TomW]

Ain't this place a treat!

Always a few methods to get the same[ish] result.

I do appreciate devices that are multitaskers. Seems more sensible.

Tom

[Phil Timmons]

Ok, a little help on the spec'ing this magic device?

(Assuming we are talking 240/120 volt single phase -- we can easily work up to three phase, but let's start simple).

Are the typical battery charger systems ever 120 volt (one "hot" leg + the neutral)  -- or are they all 240 volt (two "hot" legs + no neutral)?

Asking for whether we are tracking both incoming (or outgoing lines) and how to balance this monitoring system.

[Phil Timmons]

Chatted some more details on this over the week.

I have a ponderance that I was hoping someone with knowledge of billing/revenue meters to speak to.

Since we are designing this to work on a US Grid Connected US 240/120 Volt "split-phase" system, we should know how the billing meter really works and reads the energy use, and when getting into the detail portions, I know that I do not know those details.

Per Wiki, the billing formula is:

http://en.wikipedia.org/wiki/Electricity_meter

Wh = V(IA-IB)/2.

Of course, that being Wiki (which I like . . . .  however), it may or may not be the case, and your mileage may vary.

My concern is for an unbalanced load being used at the metering point, and then putting a balanced source (such as inverter) into the mix -- how does one cross from the overall billing consumption range and into the net generation range?

Let me try some numbers for an example --

=============================

Starting point -- No Local Solar/Wind/Inverter Generation.

On Leg A we are using 30 Amps, on Leg B we are using 10 Amps.

Per the equation above, the meter should be showing: 

Watts = Volts X (A Amps - B Amps) / 2

Or nominally -- 240 X (30 - 10) / 2  ->  240 X 20 / 2 = 2400 watts.

Does this seem reasonable as Leg A = 120 volts X 30 amps = 3600 watts and Leg B = 120 volts X 10 amps = 1200 watts, for a A + B total of:  4800 watts?

I think wiki must have been trying to say:  Watts = V X (IA + IB) / 2.

Otherwise, by the Wiki math, say we were drawing 100 amps on each leg, at 240 volts -- our power use would be metered as zero!?!?

==============================

But let's say W= V X (IA + IB) / 2, and go on with that existing unbalanced load and throw in our local generation.

We turn on our 5000 Watt inverter and start making our local generation.  That should be near 20 amps on each of Leg A and Leg B.

So now at the billing/meter point we have:

Leg A = 30 amps being used - 20 amps being generated = 10 amps still being used.
Leg B = 10 amps being used - 20 amps being generated = 10 amps being produced.

So the math for the metering comes out:

Watts = 240 X (+10 + (-10)) / 2 = Zero. 

So have we reached the (metering/billing) point of balance while we are still drawing 10 amps on one leg, if at the same time, we are sending up 10 amps on the other?

Any one have any idea if this is really how Billing/Metering really works?  Especially with the new electronic DRG (distributed renewable generation) Meters?

Thanks!

[Ungrounded Lightning Rod]

hey all,  I have a control question for you guys.
I have 4 KW of solar grid tied with Enphase micro inverters,  I also have a back-up battery bank of 1050 Ah (with an 10 foot Dans trubine).  My goal is to use excess power from the panels to charge the batteries during the day and use it in the pm to run the house. THE PROBLEM is that I need something that can monitor the outflow of power to the grid and turn on the charger when flow is out, but shut it off when flow in inward.

Does the power company credit you as much for the power you supply as it charges you for power you consume (net billing)?  Or does it pay a lot less for power you supply them than power they supply you (net purchase and sale)?

If it's net billing, why bother?  Just let the charger run whenever the batteries need it and use the grid as a buffer battery.  Hang the charger on the grid side of the inverter so it shuts down if the grid is gone and you're standalone (rather than trying to charge your backup bank with power taken from other batteries).

If it's net purchase and sale (charge you retail, credit you wholesale) then eating your own power when it's plentiful rather than selling it cheap and buying it back pricey makes sense.

Don't forget that you don't want to turn it on and off when buy/sell crosses zero.  That way produces oscillation and you need hysteresis to stop it.  You want to turn it on when the amount you're selling is a tad greater than the power drain of the charger and off when the flow is just getting back to "buy".

[youmanskids]

Thanks for the help guys,  I thougt no one else had my problem...   I'm you all can help.

"If it's net purchase and sale (charge you retail, credit you wholesale) then eating your own power when it's plentiful rather than selling it cheap and buying it back pricey makes sense."
 that is exactly the problem,  I buy for $.09/kwh and sell back for $.03/kwh, the city uses the difference in wholsale and retail to help fund the police and fire dept. (a tax of sorts which I am trying not to pay)

the idea of comparing inflow vs outflow and varying the charge rate of my Magnasine to match (with a bit of time delay hysteresis) is exactly what I need.   the inverter is 120 v single phase, and is wired to just one leg of the 240.  (the solar panels are 240 and wired to both).  it draws 110 amps at 28V during bulk charge, the absorptive stage holds it at 28.4v and the amp rate drops as the batteries fill up. The charger's charge rate can be set at any % of that max using the controller panel.

if the in/out flow direction and flow could be monitored and the charge rate adjusted (someone mentioned tapping into the communication wire) I would be in RE heaven!  that is exactly what I need.  when in house loads are high, charge rate decreases, when loads are down, send the power to the batteries for use at night.  this will maximize the power produced by the solar.
cheap and easy are also desireable 

thanks again for all your help
im excited to see some interest in my project

Roger

[youmanskids]

But like Bob observed, maybe this requires monitoring magnitude of current, as well as direction, and every time I have started wandering into this, I wind up with an entire paper energy monitoring and control system, and was just hoping for something faster, cheaper, quicker.  A girl can dream, huh?  

Hey phil, are you back to work?  still wondering how to put this thing together.  right now I'm just going outside to check the meter, then turning loads  on or off whenerver i can....this manual control is wearing me out! a computer would be much better at it and could follow multiple variables- I just need some way to measure magnitude and direction of flow and send input to the inverter/charger.  hope you can still help me in your spare time 

[madlabs]

Instead of transformers, a couple of optoisolators and bi-directional Hall effect sensors? Then feed that to a uC to switch relays etc.

Jonathan

[youmanskids]

Instead of transformers, a couple of optoisolators and bi-directional Hall effect sensors? Then feed that to a uC to switch relays etc.

Jonathan

can you give me more details?  Hall effect sensors don't work on AC as far as I know,  and how would optoisolators help me?  thanks for the input.  Roger

[Cleanergy]

We can know the power flow by measuring the power generated by the solar inverter, the power going to battery charger  and the power of the load.

if P_load => P_solarPanel + P_batt_Charger: Inflow
if P_load < P_solarPanel + P_batt_Charger: outflow
 The difference in power will set the percentage of Charging current going to the backup battery. This power ratio can be converted to a digital code or proportional voltage then send to charging rate controller of the battery charger or to an external charger as a charging rate reference. Measuring the power of the Inverter and grid will be difficult because you need to determine the ampliture and phase of the waveforms, thus needs a fast and accurate DAQ controller, unlike the power of the Solar panel it's a DC signal so power measurement will be very easy. The controller can be accurately and totally implemented in a single chip MCU ( ex.  pic18F2523 12bit adc for accurate power measurement at very low cost). A multi-channel bi-Directional high current FET switch that will work as DC Charging/Sourcing (multiplexer/Demultiplexer) will provide very flexible solution. It's an interesting project to start with, I'm in .

[youmanskids]

thanks for the reply cleanenergy,  I see what you are doing, but that will not work for my set-up.   my panels are already grid tied via a micro-inverter on each panel (no DC bus).  also the loads of the house are many and variable.   this is why I need a monitor of some kind on the main grid power comming in and going out to regulate the charge rate to the batteries.   It still seems to me there should be some simple (or off the shelf) way to monitor in- and out-flow of AC power which can control my charger-   I just havent found it yet and am hoping someone on this site has seen and solved this problem before.  thanks again for your help.   Roger.

[ghurd]

It still seems to me there should be some simple... monitor of some kind on the main grid power coming in and going out

I can't really see it at all, let alone being simple.

Monitor a leg (or both) of AC.
115W of power coming in, or going out, is still 1A that switches from one direction to the other.
My clamp-on meter will say it is 1A.
It will not say what 'direction' it is headed.

The only thing I can imagine is comparing the amps flowing from AC from the solar, to the amps flowing to the home's loads.
If the amps flowing to the home's loads is less than the amps from the solar, then you are in sell mode.
It doesn't sound simple to me, and I think it is what cleanenergy said?
A preemptive 'No' belongs here.
I have no idea how I would start working on that one.


My spit, bubblegum, and duct tape solution is starting to sound better, isn't it?
G-

[Phil Timmons]

But like Bob observed, maybe this requires monitoring magnitude of current, as well as direction, and every time I have started wandering into this, I wind up with an entire paper energy monitoring and control system, and was just hoping for something faster, cheaper, quicker.  A girl can dream, huh?  

Hey phil, are you back to work?  still wondering how to put this thing together.  right now I'm just going outside to check the meter, then turning loads  on or off whenerver i can....this manual control is wearing me out! a computer would be much better at it and could follow multiple variables- I just need some way to measure magnitude and direction of flow and send input to the inverter/charger.  hope you can still help me in your spare time 

Hi.

Back at it.  Back at this, that is -- not work, so much.

Finishing up an entire household and shop move.  Jebus.  Have not slept in three weeks.  Our old house (fortunately a rental) fell to pieces the same week as Kindergarten Graduation.  Septic Tank Blow Up, Air Conditioning, Roof Leak, and a crazy landlord converged for a perfect storm.  Nature's way of telling us to move. 

Have some really good J-men that do most of the real work, although our top guy just passed his Master's license this week, so I have to make sure our business folks do what it takes to retain him.  (or else I may find myself sweating in the Texas heat. )

But all-in-all, all is now better than before.  Our new place has a pool, better schools, acreage and a barn, is totally set up to do any kind of Renewable I want -- on a bu$ine$$ basis, and the local community loves that! -- and momma and the kids love it -- so I am a happy camper.

--------------------------

Back to your project.  Only monitoring 1 Line?  This gets simpler.

Towards ghurd's question just above -- Phase of the Current -- in relation to the Phase of Voltage -- should tell us "direction" of the Power Flow. 

Your amprobe example is correct -- you will only "see" the magnitude of the Current on an RMS meter -- not the direction.  Remember the "direction" of the Current changes 60 times a second -- our eyes can only see about 28 hz, max (called flicker fusion -- how movie frames work), so even if the simple meter tried to show a + / - direction, our eyes could not discern it.

So what we will be looking for is to compare the + (or - ) part of the Voltage waveform, with the + (or - ) part of the Current waveform.  When they are both the same ( + X +, as it were), we can say the direction of Power (which is V X I) is in the Positive Direction for Power Flow; when the Voltage and Current wave forms are in opposite ( + X - ), we can say the direction of Power Flow is in the Negative Direction.  And so, overall, we can always know if we are producing or consuming Net Power on a 1/60 of a second basis.

I know we all speak better by pictures, so if these still seems a little foggy to anyone, just say so, and I will do some waveform sketches of what we would expect to see on an O-scope and we can go from there.

[Ungrounded Lightning Rod]

Sounds like what you need is:
 - A watt sensor, reading the power you're drawing from or feeding to the grid.
 - A control input on the charger, and
 - Enough controller to use the first to adjust the second.
Set it to turn on the charger whenever you're "selling" to the grid and slowly adjust the charge rate so the selling drops to a trickle.

[youmanskids]

 momma and the kids love it -- so I am a happy camper.

--------------------------

So what we will be looking for is to compare the + (or - ) part of the Voltage waveform, with the + (or - ) part of the Current waveform.  When they are both the same ( + X +, as it were), we can say the direction of Power (which is V X I) is in the Positive Direction for Power Flow; when the Voltage and Current wave forms are in opposite ( + X - ), we can say the direction of Power Flow is in the Negative Direction.  And so, overall, we can always know if we are producing or consuming Net Power on a 1/60 of a second basis.


Sounds like you were in ... well you know where,  and now you're in paradise!!    a nice change!

so what's next?  if you can give me a schematic for the A/C flow monitor, I can build it.  then I still need to know how to tap into the communication wire from the control panel to the charger/inverted so it will adjust the charge rate. (could I use an old lap-top computer dedicated to the system to do this?)

[Cleanergy]

grid tied inverters are sync to the power line voltage, the power line voltage will be used as the reference current profile for the inverter. The inverter will operate in CSI  mode( Current Sourcing Inverter), since power line and inverter are at sync, the phase difference might be very small and difficult to detect. Since the inverter is operating in current mode, the power flow can only be adjusted by increasing or decreasing the HVDC Bulk voltage of the Inverter and it's internally controlled, what you can see at the output voltage AC  will be same as the power line waveform. Measuring the current of the DC bus ( if accessible) will determine if the inverter is sourcing or not. It can be done but not as easy as + or -.

[Phil Timmons]

Sounds like you were in ... well you know where,  and now you're in paradise!!    a nice change!

Yeah, it is often only pain that moves us -- or at least me -- forward.

So here's to pain.



so what's next?  if you can give me a schematic for the A/C flow monitor, I can build it.  then I still need to know how to tap into the communication wire from the control panel to the charger/inverted so it will adjust the charge rate. (could I use an old lap-top computer dedicated to the system to do this?)


[/quote]

I will have to play with parts a little to get a working system.  Should have the shop set up (from the move), this weekend.

Since you are a test-site / test-case, I will just wind up giving the parts to you, if that is good?

Can you tell me where this is at?  PM me if you do not want to say on the open internets.

Per some of the observations, on the thread so far, this would be two projects tied into one -- called I&C (Instrumentation and Control) in the industry side.  The instrumentation side is easy (from my point of view), and controls -- if we are going to make them match the control input to your charger -- may be a little more complicated.  We will probably have to study those specs, a bit.  Otherwise, we can have a Plan B for the control side that just shuts the charger off when it is running close to producing near Zero Net amps on the Leg being used.

For anyone tracking along -- maybe help me on some specification bounds for design concepts? 

========================

60 Hz, only, or 50/60 Hz?

Input Voltages -- 120, 208, 240, 277, 480 -- or is this overkill?   Maybe just 120 on the first one?

Measured Currents -- 0 to 200 amps?  Or maybe make ranges select-able by various Current Transformers?

Display on measuring instrument, itself?  Or only readable by a control unit?

[Phil Timmons]

grid tied inverters are sync to the power line voltage, the power line voltage will be used as the reference current profile for the inverter. The inverter will operate in CSI  mode( Current Sourcing Inverter), since power line and inverter are at sync, the phase difference might be very small and difficult to detect. Since the inverter is operating in current mode, the power flow can only be adjusted by increasing or decreasing the HVDC Bulk voltage of the Inverter and it's internally controlled, what you can see at the output voltage AC  will be same as the power line waveform. Measuring the current of the DC bus ( if accessible) will determine if the inverter is sourcing or not. It can be done but not as easy as + or -.

Yes. The line Voltage and the inverter Voltage should match exactly -- but the AC current from the inverter will go up and down from zero to max of the inverter output.  And you are also correct that the inverter Power Factor (phase angle between Voltage and Current) should be at or very near 100% -- or totally in phase. 

H O W E V E R -- that part is not the object of this exercise.  What youmankids needs to know is what is happening at the Revenue or Billing Meter.  And then switch his battery charger (and/or other loads) on-or-off to make sure he stays on the power producing side of things. 

While the voltage at all these points should stay the same (giving us a good phase reference point),  the current in or out of any of these nodes can vary widely.  So we will to measure the current at the billing meter to make sure he stays in a net power producing mode at the billing meter.

We should be able to tell the magnitude of the current by measuring it with a Current Transformer, (a common practice) and then also the direction of the current by comparing it to the phase of the Voltage.   For example, we can use 0 degrees phase angle to mean incoming current and 180 degrees phase angle to be out-going current.

In all that the inverter, itself, will likely remained untouched, and just doing its thing.

[Cleanergy]

the phase of the current can't be use as a good reference since it will vary depending on the local loads and the grid loads, it's possible but the variation might need very fast response and accurate measurement and aquisition system.

[Phil Timmons]

the phase of the current can't be use as a good reference since it will vary depending on the local loads and the grid loads,

You follow we are talking about changing all the way from consuming power to generating power?  That is a 180 degree (massive) phase shift for the current. 

But you are very correct about local power factor constraints when it comes to just looking for a simple or slight difference in between the voltage and current such as motor loads may create.

Most extreme case I have run across this is in West Texas oil fields.  They are motor loads in groups of 50 or more wells, water pumps and other motors. 

Since it is the end-of-the-line for the grid in those areas, and loads are so wild we see power factors of up to - 40% while a typical 20hp motor is drawing up oil -- but then on the down-stroke, the whole rig becomes a generator, and the power factor can jump all the way across to the +40% side. 

Makes the metering and billing go wild.

it's possible but the variation might need very fast response and accurate measurement and aquisition system.

Yeah, if the initial pick up comparison is analog, we will be able to compare the phase of the voltage and current in analog as well, and that should have the effect of sampling at 1/120th of a second, if compared to the digital control system modeling. 

The feedback/control system and response of this particular battery charger is something I do not know about, yet.  If we have to settle for a hard ON / OFF relay, that should have a response of better than 10/60th of a second (relay open/close time), I would think, but I guess we can make a "vibrator" of the system to see its response time. 

[youmanskids]

hey Phil,  sounds great, If you can get a monitoring system together, I think "boB" mentioned in an earlier response than the Magnasine protocols are published on-line and we could "tap in" to them for control of the variable charge rate.
If boB is following along, would appreciate more input/education (how is it done (it has standard phone cord and jacks between the control panel and the charger)?, usb plug?, do I need a dedicated PC for this? what programming language?...),  thanks  roger

[oztules]

This is an interesting problem.

Could it be easier than we think?

What if we put a current shunt in the AC circuit..... or even measured the V drop over the main fuse.  If we measure the voltage  neutral to line on both ends of the shunt or fuse. we would see a voltage differential. When the voltage on the grid side of the fuse or shunt is higher than the voltage on the house side..... we are using the grid. (only millivolts... but thats plenty to work with)

When the voltage on the house side of the fuse or shunt is higher than the voltage on the grid side.... we are exporting. The differential between the grid side of the fuse/shunt against the house side of the fuse/shunt, will give us direction and amplitude....(mv or volts, depending on the choice of shunt ) and some small op amps acting as DC amps should lift the levels to  a usable level to work with. If we use two small (tiny) transformers to bring the voltage down to  say the 9v level, we can be isolated, and still get accurate results.

This looks like a workable idea to me.




.............oztules

[TomW]

Oz;

Looks like a good way to me. I am convinced many folks "overthink the plumbing" these days simply because so much technology is available.

Good idea, IMHO

I am not sure how the old common mechanical KWH meter as used here deciphered which way the power flows but they would run backwards if flipped in the socket mostly been upgraded to digital now. I doubt they have much of anything "electronic" in them but must be some method with transformers, shunts, etc?

Tom

[oztules]

Hi Tom,

The old meters were simple eddy current motors attached too a gearbox and clockwork dials. The motion was induced by a voltage coil (across supply) directed at the flat round plate. This created emf in the plate, that is shorted in the plate, which creates  eddy currents which create  magnetic fields emanating from the plate.

A second electromagnet is directed at the plate from a current coil (in series with supply). So the plate has two coils, out of phase acting on it. It is now direction sensitive, so spins one way or the other depending on current flow in or out.

To stop it spinning like a top, it has drag induced into the plate by an adjustable distance permanent magnet. Closer the magnet is placed, the higher the drag.... slows it down.... we now have rpm (interaction of fixed voltage coil vs variable current coil) and braking torque created by eddy brake permanent magnet.....so   torque x rpm........we can now measure power, and direction of power flow simply.



.............oztules

[boB]

I wonder how much real power one of those spinning power meters draws, itself ??

Probably not very much but it has to draw ~something~

boB

[Tritium]

I wonder how much real power one of those spinning power meters draws, itself ??

Probably not very much but it has to draw ~something~

boB

Mine states 3W 240V on the name plate.

Thurmond