The idea was requested by Mr. David. The following email conversation between him and me details the main specs of the proposed Synchronized 4kva Stackable Inverter Circuit.
Firstly I wanted to say thank you for your contribution to the world at large, the
information and most importantly your willingness to share your knowledge to help
other people in my opinion is invaluable for many reasons.
I would like to enhance some of the circuits you have shared to suit my own
purposes, unfortunately whilst I understand what is going on in the circuits I lack
the creativity and knowledge to make the amendments myself.
I can generally follow
circuits if they are small and I can see where they join/connect into bigger
If I may I would like to try to explain what I would like to achieve, though I am
under no illusion that you are a very busy person and would not like take up your
precious time unnecessarily.
The final goal would be that I would like to build ( assemble the components ) of a
multi-source renewable energy micro grid, using Solar PV, Windmills, and bio diesel
The first step is the PV solar inverter enhancements.
I would like to use your 48 volt pure sine wave inverter circuit capable of
maintaining a constant 2kW 230V output, it must be capable of delivering at least 3
times this output for a very short duration.
The key modification that I want to achieve it to create a number of these inverters
units to work in parallel and connected to an AC bus bar.
I would like each inverter to independently and constantly sample the AC bus bar for
frequency, voltage and current (load).
I will call these inverters slave units.
The idea being the invert modules will be “plug and play”.
The inverter once
connected to the AC bus bar would constantly sample/measure the frequency on the AC
bus bar and use this information to drive the input of a 4047 IC such that its clock
output can be advanced or retarded until it exactly clones the frequency on the AC
bus bar once the two wave forms are synchronized the inverter will close a contactor
or relay which connects the invert output stage to the AC bus bar.
In the event that the frequency on the bar or the voltage moves outside of a
pre-determined tolerance the inverter module should open the relay or contactor on
the output stage effectively disconnecting the inverter output stage from the AC bar
to protect its self.
Additionally once connected to the AC bus bar the slave units would go to sleep or
at least the output stage of the inverter would sleep while the load on the bar is
less than the sum of all of the slave inverters.
Imagine if you will there are 3 slave inverters attached to the AC bus bar, however
the load on the bar is only 1.8kW then the other two slaves would go to sleep.
The reciprocal would also be true that if the load on the bar jumped to say 3kW one
of the sleeping inverts would instantly wake (already be in sync) to supply the
additionally required energy.
I imagine some large capacitors on each of the output
stage would supply the energy required whilst the inverter has the very short moment
whilst it wakes up.
It would be preferable (only in my opinion) not to directly connect each inverter to
each other but rather that they be independently autonomous.
I want to try to avoid
micro controllers or the units error or fault checking each other or the units
having ‘address’s’ on the system.
In my mind’s eye I imagine that the first connected device on the AC bus bar would
be a very stable reference inverter that is constantly connected.
inverter would provide the frequency and voltage that the other slave units would
use to generate their own respective outputs.
Unfortunately I can’t get my head around how you could prevent a feedback loop where
the slave units would each potentially end up becoming the reference unit.
Beyond the scope of this email I have some small generators I would like to connect
to the AC bus bar synchronizing to the reference inverter to supply energy in the
event that the load exceeds the DC max output capacity.
The overall premise is that the load presented to the AC bus bar would determine how
many inverters and ultimately how many generators would either autonomously connect
or disconnect to meet the demand as this would hopefully save energy or at least not
The system being completely built of multiple modules would then be
expandable/contractible as well as robust/resilient such that if anyone or perhaps
two units were to fail the system would continue to function all be it at reduced
I have attached a block diagram and excluded the battery charging for the time
I plan on charging the battery bank from the AC bus and rectifying down to
48V DC this way I can charge from the generators or the renewable energy sources, I
do recognize that this is perhaps not as efficient as using DC mppt but I think what
I lose in efficiency I gain in flexibility. I live a long way from town or the
For reference there would be a minimum constant load on the AC bus bar of 2kW’s
though the peak load could rise by as much as 30kW.
My plan is for the 1st 10 to
15kW to be provided by the solar PV panels and two 3kW (peak) windmills the
windmills are wild AC rectified to DC and a 1000Ah 48 volt battery bank. (Which I
would like to avoid draining/discharging beyond 30% of its capacity to ensure
battery life) the remaining infrequent and very intermittent energy demand would be
satisfied by my generators.
This infrequent and intermittent load comes from my
I have been thinking that it may be prudent to build a capacitor bank to handle or
pick up the system slack of any inductive load start up currents such as the motor
on my air compressor and table saw.
But I am not sure at this time if there is not a better/cheaper way.
Your thoughts and comments would be greatly appreciated and valued I hope you have
time to get back to me.
Thank you for your time and attention in advance.
Sent from my BlackBerry® wireless device
I have read your requirement and have hopefully understood it correctly.
Out of the 4 inverters, only one would be having its own frequency
generator, while others would be running by extracting the frequency from
this main inverter output, and thus all would be in sync with each other
and with this master inverter's specs.
I'll try to design it and hope it works as expected and as per your
mentioned specs, however the implementation will need to be done by an
expert who should be capable of understanding the concept and modify/tweak
it to perfection wherever it might be required....otherwise succeeding
with this reasonably complex design could become extremely difficult.
I can only present the basic concept and the schematic....rest will need
to be done by the engineers from your side.
It might take me some time to complete, this since I already have many
pending requests in the Queue...I'll inform you as son as it's posted
sub circuits one that looks at the frequency on the AC bus bar and this
unit is used to create the clock pulse for the inverter sine wave
that would feed back into the inverter sine wave generator clock
pulse to advance the clock signal or retard the clock signal until the
output from the sine wave generator exactly matched the sine wave on the
frequency of the AC bus bar there would be an SSR that would close
connecting the output stage of the inverter onto the AC bar preferably
at the zero cross over point.
carry on functioning. the purpose of the master inverter was that of all
the inverter modules it would never go to sleep and would provide the
initial AC bar frequency. however if it failed then the other units would not be affected as long as one was 'online'
mechanical and electrical engineer 🙂 I work with big plant items
like chillers and generators and compressors.
would you be wiling/open to accept a money gift? I don't have much but I
could perhaps gift some money via paypal to help suport your website
Basically you want the inverters to be in sync with each other in terms of
frequency and phase, and also each one having the ability to become the
master inverter and takeover the charge, in case the previous one fails
due to some reason. Right?
I'll try to fix this with whatever knowledge I have and some common sense
and not by employing complex ICs or configurations.
as the load increases off or increases they wake to meet the demand.
As requested by Mr. David, the proposed 4kva stackable power inverter circuits need to be in the form of 4 separate inverter circuits, which can be stacked up appropriately in sync with each other for supplying the correct amount of self-regulating power to the connected loads, depending on how these loads are switched ON and OFF.
Synchronizing the Inverters
And in case the second inveter also fails, the third inverter takes the command and plays the role of the master inverter.
All the above needs to be executed without losing the control over frequency, phase and PWM even for a split second, and with a smooth transition.
The basic design for fulfilling the mentioned criteria is shown in the following diagram:
However even with the precisely matched attributes, the inverters cannot be expected to run perfectly in sync unless these are tied up in some unique manner.
This is in fact done by integrating the the "slave" inverters through an opamp/optocoupler stage as indicated in the above design.
Initially, the master inverter#1 is switched ON, which allows the opamp 741 stage to get powered and to initialize the frequency and phase tracking of the output voltage.
Once this is initiated, the subsequent inverters are all switched ON for adding power to the mains line.
As can be seen the opamp output is connected with the timing capacitor of all the slave inverters through an opto coupler which force the slave inverters to follow the frequency and the phase angle of the master inverter.
However the interesting thing here is the latching factor of the opamp with the instantaneous phase and frequency information. This happens since all the inverters are now delivering and running at the specified frequency and phase from the master inverter, which implies if in case any of the inverters fail including the master inverter, the opamp is able to quickly track and inject the instantaneous frequency/phase info and force the existing inverters to run with this specifications, and the inverter in turn are able to sustain the feedbacks to the opamp stage to make the transitions seamless and self optimizing.
Therefore hopefully the opamp stage takes care of the first challenge of keeping all the proposed stackable inverters perfectly synchronized through a LIVE tracking of the available mains specification.
In the next part of the article we'll learn the synchronized PWM sinewave stage, which is the next crucial feature of the above discussed design.