Grid Tie Inverter Schematic

[selimy]

uploading worked

hi there,
this timed I tried using zeners and minimal components. It worked great. Nothing to set or adjust. And the square wave looked pretty neat on the sine.

I was able to upload around 40W onto the grid (limited by cables and supply). I noticed when I increased above 36Vpp for a short period, the power factor increased and uploading exceeded 100W.

To make it work this time, (thanks to tcmtech) I connected both12V secondary coils in parallel and supplied them w/approx 36Vpp through a 15A buck converter (19VDC out) connected to the load-side of the mppt-charge-regulator, which outputs a constant 26.6V, utilising the combined power of the 24V lead-acid battery and mppt.

Followed Tcmtech's steps to the letter and all went smooth. pictures attached, but a bit messy, sorry

Setup:
+Mppt charge controller 5-60VDC in 30A
+15A dc dc buck converter, up to 36VDC in
+2x 7VA/12-0-12 control-transformers toroidal (magnetically shielded)
+8x15V zeners 5W
+1200V/25A IGBTs NGTB25N120IHLWG (they are not the best, will try others soon)
+300VA output transformer 12-0-12
+used a 15A DC feed in diode instead of X1 (used 2 within GBPC1506), they get around 70C hot, would need a better diode here... would a SCR be better suited here?

+skipped X2 as there is no center-tap due to paralleling of the 12V coils. What would X2 do exactly?

+I haven't connected any LC/RC tanks yet. How can I improve the power factor in the most simple way?

... thanks for looking

 

[tcmtech]

Nice going! This is about the most simple yet effective GTI circuit you can make and I guarantee that once it's tweaked in properly it is avery durable and reliable circuit at that!

X2 going to a center tap gives you a optional half voltage input range. That is if you designed your system around say a 24 VDC input for the full bridge input through X1 the center tap input through X2 would give you a 12 VDC input range as well.
Mostly it's just an option for if you have a input power source that has a very wide input range and you don't have a buck or boost converter MPPT stage this would give you a way to feed power back at lower than average input voltages.

As far as using a SCR instead of a diode the SCR would give you a way to keep the input power turned off to the switching devices entirely until the SCR is turned on.

The simplest power factor improvement I can suggest is to use common AC power factor correction capacitors (Motor run capacitor) on the AC side For C5 and or C2. Just find what size gets your power factor up where you want it at whatever load point you want.

 

[selimy]

Thanks a lot for your inputs, tcmtech.

just got back from Lyon, France. Antoine de Saint-Exupery once said:

In anything at all, perfection is finally attained not when there is no longer anything to add, but when there is no longer anything to take away.

Drawing parallels, I really like this Zener approach, and it looks simple, rock solid, and stable.

There is a bit of work ahead, actually two ideas that keeps me busy for now. One, is to connect the 98% efficient buck converter straight to the battery. I was thinking of your ZR431/555 approach to cutoff battery supply below 21V, avoiding damage to battery.

The other idea is to move away completely from mppt/battery and connect the variable power source (solar panel best output at 31V 8A) directly to the buck converter. 22VDC is required to achieve around 20V at the buck converters' output. I believe this would cause on/off cycles, which would be OK if uploading of energy was achieved with only a small loss, which I doubt.

Your previous circuit using ZR431/555 was quite clever as it uploaded energy when the voltage was above a certain level. When it dropped below set value it simply waited a period in sby . I might go this path, even though I feel that the energy below that set voltage is lost.

I tried to find a simple mppt circuit on the net, not much luck. Most of them are micro controlled and have battery charging logic. Is there a way to turn my buck converter into a quasi-mppt, let's say, if we added big capacitors between the solar panel and the buck converter, extracing constant 20V DC to feed the H bridge.

Haven't worked on the power factor yet. Will read a bit more on this as I get close to uploading around 200W.

 

[tcmtech]

The 555 based standby mode could be expanded to include a second 555 that shuts down the actual switching devices after a slightly longer standby period.

At that point the whole GTI is basically off other than the voltage monitoring circuit. Very little power is wasted that way.

Its just a thought.

 

[-nikos-]

hallo
someone blew out the plan that I upload...

in video this time ...

 

[selimy]

Hi Nikos,

looked at your video and liked your idea. Will try this circuit on a bread board and let you know. I noticed you retrieve the sine from he mains directly, which could be a bit tricky.

I saw somewhere a very precise and simple to build zero-crossing detection circuit that creates a 0.5V pulse every time when the wave alternates. Would it be possible to feed TL494 with a "perfect" sine wave (eg from a 8038), however sync-triggered from that ZCD circuit. This way, we wouldn't have to worry about distortions, harmonics aso. Always have a nice 10 ms sine hump at the right time. What do you think?

Hi tcmtech, I ordered a 300W mppt boost converter eval board which should hopefully arrive in January ( https://www.st.com/web/en/catalog/tools/FM116/SC1078/PF251965 )

rgds

 

[selimy]

Hi Nikos,

seems to work very well (see pics). Will do a bit more testing using other mosfets/igbts and transformers under load. Ignore the centre bit on the bread board, which has the IR2113 and is not connected. I noticed an increased delay using this chip so I used the transistors instead. The BD transistors are working much faster.

rgds

 

[tomizett]

Hi All,
I've been reading this thread with interest for a while (although it seems to grow quicker than I can read it some times). Although I've no cause to actually build a GTI, I thought I'd have a go at designing one in simulation only - here is the result.
It seems to me that one of the earlier contribtors (was it MrRB?) was right to suggest a constant-current aproach, with a circuit that drives the grid with a current proportional to the grid voltage (although it does strike me that such a circuit does nothing to improve the line waveform if it is already distorted).

I've opted for a line frequency transformer, putting all the electronics on the low-volts side. I prefer this because a) it's safer, b) high-side switch drive is easier, c) you can use low-voltage MOSFETS with low Rds(on) d) you can use the transformer as part of the line filter and e) it's marginally less likely to explode during development.

The design uses a full bridge with PWM control to aproximate a sinusoidal current drive onto the grid via a main power transformer. Two smaller transformers are used for control an power supply purposes.

I was aiming to come up with something that could easily be built by a hobbyist like myself. It's by no means a complete design, but I'd be very interested to see what people here think.

For the immediately curious I've attached a .png schematic; the zip file contains the schematic plus some graphs from the simulation and a circuit description

 

[Ng Jing Xi]

Why must the GTI be in phase with the power grid?
Is it because that if the GTI voltage is lower than the power grid, it will sink current instead of sourcing it?

 

[rrranjit]

Hi Robby here, I read your article with interest, and I would like to build a 250 watt
230 volt, 50 hrtz Grid tie inverter, but I need your help with design and parts list please, I am a very interested newby, Not sure if I should put my e mail address, but I gess someone will tell me,thanks,

 

[Tony Stewart]

A lot of good energy discussing this topic. From my experience a design is only as good as the specs it can meet, the ones you need to verify.

This is cover , all intended and unintended inputs and outputs and functional responses including detection/protection specifications. Power Line Transients may exceed 6kV for very short durations as the power meter arc suppression gap is set 6kV has a finite ionization time.

Those who have specs , please contribute.

Communication Mode: Power Line
Power transmission mode: Reverse transfer，load priority
Electromagnetic Compatibility EN50081.part1 EN50082.Part1
Grid disturbance EN61000-3-2 Safety EN62109
Grid detection DIN VDE 1026 UL1741


RATED POWER
PV INPUT (DC)
Maximum DC Power
Nominal DC Voltage
Maximum DC Voltage
Working DC Voltage Range
Start-up Voltage / Initial Feeding Voltage MPP Voltage Range / Full Load MPP Voltage Range
Maximum Input Current
Isc PV (absolute maximum)
Max. inverter back feed current to the array
GRID OUTPUT (AC)
Nominal Output Voltage
Output Voltage Range

Output Frequency Range

Nominal Output Current
Inrush Current/Duration
Maximum Output Fault Current/Duration Maximum output Overcurrent Protection Power Factor Range
AC INPUT
AC Start-up Voltage
Auto Restart Voltage
Acceptable Input Voltage Range
Nominal Frequency
AC Input Power
Maximum AC Input Current
Inrush Input Current
BATTERY MODE OUTPUT (AC)
Nominal Output Voltage
Output Frequency
Output Waveform
Output Power
Efficiency (DC to AC)
BATTERY & CHARGER (Lead-acid/Li-ion)

DC Voltage Range
Nominal DC Voltage
Maximum Battery Discharging Current Maximum Charging Current

230 VAC (P-N) / 400 VAC (P-P)
184 - 265 VAC per phase
47.5 ~ 51.5 Hz or 59.3~ 60.5Hz
14.5 A per phase 17 A per phase / 20ms 51 A per phase / 1ms
51 A per phase
0.9 lead – 0.9 lag
120-140 VAC per phase
180 VAC per phase 170 - 280 VAC per phase 50 Hz / 60 Hz
10000VA/10000W 40 A
40 A / 1ms
230 VAC (P-N) / 400 VAC (P-P) 50 Hz / 60 Hz (auto sensing)
Pure sine wave
10000VA/10000W 91%
40 – 60 VDC 48 VDC 275 A
200 A

 

[rrranjit]

A lot of good energy discussing this topic. From my experience a design is only as good as the specs it can meet, the ones you need to verify.

This is cover , all intended and unintended inputs and outputs and functional responses including detection/protection specifications. Power Line Transients may exceed 6kV for very short durations as the power meter arc suppression gap is set 6kV has a finite ionization time.

Those who have specs , please contribute.

Communication Mode: Power Line
Power transmission mode: Reverse transfer，load priority
Electromagnetic Compatibility EN50081.part1 EN50082.Part1
Grid disturbance EN61000-3-2 Safety EN62109
Grid detection DIN VDE 1026 UL1741


RATED POWER
PV INPUT (DC)
Maximum DC Power
Nominal DC Voltage
Maximum DC Voltage
Working DC Voltage Range
Start-up Voltage / Initial Feeding Voltage MPP Voltage Range / Full Load MPP Voltage Range
Maximum Input Current
Isc PV (absolute maximum)
Max. inverter back feed current to the array
GRID OUTPUT (AC)
Nominal Output Voltage
Output Voltage Range

Output Frequency Range

Nominal Output Current
Inrush Current/Duration
Maximum Output Fault Current/Duration Maximum output Overcurrent Protection Power Factor Range
AC INPUT
AC Start-up Voltage
Auto Restart Voltage
Acceptable Input Voltage Range
Nominal Frequency
AC Input Power
Maximum AC Input Current
Inrush Input Current
BATTERY MODE OUTPUT (AC)
Nominal Output Voltage
Output Frequency
Output Waveform
Output Power
Efficiency (DC to AC)
BATTERY & CHARGER (Lead-acid/Li-ion)

DC Voltage Range
Nominal DC Voltage
Maximum Battery Discharging Current Maximum Charging Current

230 VAC (P-N) / 400 VAC (P-P)
184 - 265 VAC per phase
47.5 ~ 51.5 Hz or 59.3~ 60.5Hz
14.5 A per phase 17 A per phase / 20ms 51 A per phase / 1ms
51 A per phase
0.9 lead – 0.9 lag
120-140 VAC per phase
180 VAC per phase 170 - 280 VAC per phase 50 Hz / 60 Hz
10000VA/10000W 40 A
40 A / 1ms
230 VAC (P-N) / 400 VAC (P-P) 50 Hz / 60 Hz (auto sensing)
Pure sine wave
10000VA/10000W 91%
40 – 60 VDC 48 VDC 275 A
200 A

 

[DerStrom8]

rrranjit Hijacking other peoples' threads is not allowed here. Please start your own. Additionally, do not post your email address in the public forum. That's just asking for trouble.