DIY 230v/2kw HV inverter

Hi.
I just wanted to share this open source inverter project; https://github.com/jaw0/littlebox-power-inverter as I guess it could be of interest for some people here.

This pic shows the size and layout;

These are the specs descried in the repo which are inherited from the google little box challenge back in 2014;

• Must be able to handle up to 2 kVA loads
• Must achieve a power density of equal to or greater than 50 W/in3
• Must be able to handle loads with power factors from 0.7–1, leading and lagging in an islanded mode
• Must be in a rectangular metal enclosure of no more than 40 in3
• Will be taking in 450 V DC power in series with a 10 Ω resistor
• Must output 240 V, 60 Hz AC single phase power
• Must have a total harmonic distortion + noise on both voltage and current of < 5%
• Must have an input ripple current of < 20%
• Must have an input ripple voltage of < 3%
• Must have a DC-AC efficiency of greater than 95%
• Must maintain a temperature of no more than 60°C during operation everywhere on the outside of the device that can be touched.

I'm not sure but I guess this project fullfill the specs above.

Just ordered some pcb's to try to put one together myself. My idea is to mainly use it powered from the HV bus (about 405-320v) at a nissan leaf, but hopefully also powered from other sources. As I understand it could also be powered from lower voltage sources but probably with a quite poor efficiency, looking forward to investigate that at different voltages.
My other ideas are to look into syncing 3 pcs for 3phase output or maybe elaborate with adding some code for MPPT, maybe with some code/inspiration from this project; View: https://www.youtube.com/watch?v=ShXNJM6uHLM
, as the topology seems to be similar to most standard transformerless grid tie inverters, with a boost converter and then an H-bridge.

Regarding the little box challange, there is a really interesting report, describing several of the approaches, difficulties and solutions the participants ran into, well worth a look, here;

Hope I can get back with some more info in a bit.

Guess I should also warn about lethal voltage/amp/power levels and clarify that I'm not the creator of the mentioned project.

Defenitly a neat idea! I eventually wanted to diy a high dc voltage to ac inverter myself.
But why 240v 60Hz? Most of the world runs on 50Hz and 240v is the upper limit of 230v lines. Being able to adjust the output would be great.

Also as a side note, AC voltage is given in RMS and the actual peak voltage in the sine wave is for 230v ac at about 325v. So your battery has to be above that voltage at all time to get good efficiency.

I'd love to use a battery pack, with 128 cells in series (537.6v-384v if using the default 4.2v-3.0v limit) That allows for staying above the peak voltage needed.

If you need any help or have any questions, feel free to ask.

Oberfail said:

Defenitly a neat idea! I eventually wanted to diy a high dc voltage to ac inverter myself.
But why 240v 60Hz? Most of the world runs on 50Hz and 240v is the upper limit of 230v lines. Being able to adjust the output would be great.

Click to expand...

Good point. Not sure but guess it is cause the challenge was initiated in US, even if 120v is more common at households, 230v is used at the low volt sections of the grid, right?
But its a minor change I would say, the sine wave is not calculated on the fly but taken from a lookup table in this file; https://github.com/jaw0/littlebox-power-inverter/blob/master/src/sin60_350.h
And refered to at line 29 in this file; https://github.com/jaw0/littlebox-p...060c87bd240e555c413af368cc95d0/src/inverter.c

I just put together a new array and made a pull-request for it, lets see if it gets merged https://github.com/jaw0/littlebox-power-inverter/pull/2

Oberfail said:

Also as a side note, AC voltage is given in RMS and the actual peak voltage in the sine wave is for 230v ac at about 325v. So your battery has to be above that voltage at all time to get good efficiency.

I'd love to use a battery pack, with 128 cells in series (537.6v-384v if using the default 4.2v-3.0v limit) That allows for staying above the peak voltage needed.

Click to expand...

As you can see on this schematic below the inverter starts with and step-up, as most grid tie inverter does, so lower dc input would not be a major problem but likely affect the efficiency. The DC-bus is rated for 1kv and the SiC transistors for 1,2kv, to be able to get 230v RMS out I'm quite sure you'll need 650v dc, maybe 325v is good for 230v split phase..

Oberfail said:

If you need any help or have any questions, feel free to ask.

Click to expand...

Next task for me is to tidy up the BOM to be able to also order components:
Seems like some of the parts used are not available anymore so will be a little effort to update with presently available equivalents.
Started here but there is quite a bit left to do;

I think the total part cost would be around 200eur.

Another thing that would be nice to have is, the pcb files translated from eagle to kicad, there is an import function in kicad for eagle files but I think there are a little more work to do at least to verify that all looks correct also in kicad.

One step closer.
Still have some components to order though.
The large blue card is the power stage, the smaller blue is the logic board, the 4 larger green are different output filters and the small green is logic voltage supply.

That thing is a death trap.
If I'm not mistaken, the output voltages are not isolated from the input voltages.
It looks like the output sine wave would have a nominal DC offset of 500V (ie half the 1kV rail).
The positive going part of the sine curve would be approx 500+325V min
It's using & making some very lethal voltages.
If you connected this to a house mains system with neutral earthed, your battery bank & associated solar panels, control electronics, etc, etc would be at highly lethal voltages.
The git site says "This device does not comply with consumer electrical safety requirements."

Summary: this is a research/academic project. DON'T build it & expect to use it, you'll kill someone.

Redpacket said:

That thing is a death trap.
If I'm not mistaken, the output voltages are not isolated from the input voltages.
It looks like the output sine wave would have a nominal DC offset of 500V (ie half the 1kV rail).
The positive going part of the sine curve would be approx 500+325V min
It's using & making some very lethal voltages.
If you connected this to a house mains system with neutral earthed, your battery bank & associated solar panels, control electronics, etc, etc would be at highly lethal voltages.
The git site says "This device does not comply with consumer electrical safety requirements."

Summary: this is a research/academic project. DON'T build it & expect to use it, you'll kill someone.

Click to expand...

Good point. Correct it's not isolated. Mainly what I want to try out is connecting it to an isolated battery setup from an electric car. Do you know how PE use to be handled in standard grid-tie inverters as they're usually transformer less and have a similar topology?

laasrocket said:

Do you know how PE use to be handled in standard grid-tie inverters as they're usually transformer less and have a similar topology?

Click to expand...

Not sure, sorry.
Might be (marginally) less dangerous if you were making each half sine with an inverting (-ve) & non-inverting (+ve) buck or flyback topology directly in the first "stage".
Even if you've got experience/qualifications with HV stuff, only then a very cautious maybe for one test only. Otherwise don't.
Isolating the battery wouldn't be safe, whole idea is a nightmare.

What I've found about standard transformerless grid-tie inverters is that there are higher level of quality requirements on the PV side cables, also safety switches and over current protection seems to be convention or law. On EV's with HV batteries isolation tests between chassis <-> HV+ and chassis <-> HV- is standard. I guess such procedure would also be a good idea for this kind of setup, probably would even be possible to jack into an exsisting isolation test on an EV.

Hey, how far are you with the project?