Electronics > Projects, Designs, and Technical Stuff

Motor inverter as a battery to mains power inverter

<< < (3/4) > >>

Siwastaja:
The VFD expects significant inductance motor windings provide. Good transformers have little leakage inductance.

I wouldn't except it to work. You could use a VFD as a building block with an added output filter, and if you are very lucky, the V/f mode could work, although I wouldn't count on it. It's designed to drive motors.

richard.cs:
Essentially what you propose will work, but you will need a filter before the transformer to turn the VFD PWM into a sensible sine wave. If you rely on transformer leakage inductance to do that for you everything is a bit uncontrolled and the core losses in the transformer will also be high. This relies on the VFD being able to be set into a "dumb mode" where it ignores the fact that only two outputs are connected.

As an alternative ebay etc. is flooded with low-cost inverter boards that take a high voltage DC input produce sine wave outputs, often including the filter on the board. Many are built for 380 V in to 220 V out but would require minimal changing (just feedback components basically) to do 200 V to 120 V. e.g. https://www.ebay.co.uk/itm/2000W-Pure-Sine-Wave-Inverter-Power-Board-Post-Sine-Wave-Amplifier-Board-DIY-Kit/133197461127 - note that if you operate at 120 V the available current is not increased so the power rating is halved. All the cheap ones seem to use a EG8010 controller, unfortunately so far as I can tell it does not support syncing multiple devices.

NiHaoMike:
A better approach would be to get a second Prius inverter (~$100) and connect the battery input to 3 or 4 telecom rectifiers in series. Then use the built in boost converter to step the voltage up to about 400V and then use 2 of the output phases for 240V out. To allow that kind of control, replace the controller with a STM32.
https://openinverter.org/wiki/Toyota_Prius_Gen3_Board

It would also be helpful to know exactly what kind of load you need to power, for example electronics work great on a filtered square wave while motors run best with a PWM sine wave but only need minimal filtering to avoid standing wave issues, or no filtering if the motor is close to the inverter. It might be easier to have separate outputs for electronics and motors.

haxby:
Thanks Richard, the ES8010 is looking to be the most promising way to do this.... The pre-assembled ES8010 modules with MOSFET drivers on eBay look convenient. I'm surprised there is enough of a demand for a low level board like this to be available from so many suppliers. Any tips or links on what the filter between the MOSFET outputs and the transformer input should look like? I assume it will be dependent on the characteristics of the transformer used? If I design the power electronics with quality components I hope I'll end up with a reliable solution.

Regarding the EG8010 data sheet, on page 10 it shows a recommended diagram for a low frequency transformer inverter. There is no filter between MOSFETs and transformer. But there is a 2.2uF cap on the secondary windings. The note says "T1 needs to use low power frequency transformer. Transformer filters PWM high frequency signal by connecting its secondary turns to a 2.2uF/400V capacitor of CBB. After filtering, it outputs 50Hz/60Hz sinusoid".

 Is this enough or should there also be a filter on the primary side?

Thanks for your help!

richard.cs:
Those modules with the gate drivers turn up in all kinds of inverter products, both as a final-stage for an inverter with a high-frequency DC-DC, and as the first stage of an inverter with a low-frequency transformer. Be aware that by default those boards don't connect up all the protection features (like overcurrent) that the ES8010 offers. Boards like the one I linked that contains one of those modules and some output FETs are much less common, but probably won't do the power output you need anyway.

It can work without an explicit filter, but to do so well the transformer has to be carefully chosen (or more likely in a commercial product as this chip is intended for, designed) for the right amount of leakage inductance low core loss at the switching frequency, etc. If you add a filter it's everything (including the filter) is much less dependent on the transformer characteristics, the 3.3 mH / 2u2 filter on the preceding page is probably fine as a functional waveform-reconstruction filter.

EMC is a rather separate issue and may require additional filtering at input an output, but this would be at higher frequencies and could use more modest component values. Be aware that (with any inverter) unless you add a lot of bulk capacitance to the input, and maybe a low-frequency inductor, your DC input waveform will mostly consist of 100 Hz ripple. Prius batteries probably won't complain too much about that, but the rms will be higher than you expect so be careful when sizing cables.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod