50Hz Sinewave Inverter Filter, 24VAC

Go To Last Post
6 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi All,

I have created a 24VAC 100W sinewave inverter that uses an ATMega8535. The Mega, is used to control, amongst other things, a boost converter from 24VDC to arround 35VDC, which is then used to create the 50Hz sinewave output via a H-Bridge.

The Sine PWM section is via OC1A, running at 12.288MHz/510 ~24KHz. There are 480 steps in the initial Sinewave table which varies from min 5, to max 250, to allow for refresh of the bootstrap capacitors on my HIP4080A H-Bridge driver IC.

The problem is the Output Filter LC. Without the output filter I can drive loads to over 150W.

The original filter, uses 470uH inductors, 2.2uF to 0V from each output, and 4,4uF between each output.

With the filter in place, the inductors either run Hot, or cause the inverter to fail. Going up to a 1.57" iron powder torroid (Amidon T157-26mix) seems OK, but why is the output filter size so big?

What if this was a 240V application?

I would be happy to share my code, and perhaps hardware for a valid solution to this problem.

Regards,

Ron.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Interesting problem, Ron.

The first thing that comes to (my) mind is resonance. Apparently the inverter-output from the H-bridge contains frequency components that match the resonance frequency of the output-filter. Serial resonance means zero ohm impedance. And as a result, each of the two outputs from the H-bridge see a short for a significant part (frequncy-wise) of the signal.

Two things need to be done:
1. Finding out what the spectrum is of the un-filtered inverter-output
2. Create a filter-design that "shapes" the sine.

Btw, 24V or 240V, the issue remains the same.

Do you have tools to analyze the spectrum ? If not ..... I think I can come up with a work-around.

Nard

Edit: added schematic

Attachment(s): 

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

Last Edited: Wed. Jul 25, 2007 - 05:46 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Nodding my head vigorously in a greement with Nard. I recently went through a similar problem. Just as a test you can shift the corner frequency of your filter and see if the problem goes away (of course this will be at the expense of noise but for a test it will at least help to illuminate)

Go electric!
Happy electric car owner / builder

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Nard, I don't have tools to see or find the spectrum. The cutoff frequency for this filter is around 1500Hz. RL ~6 ohms.

BTW, the filtered output is a nice clean sinewave, it just causes problems as the load current is increased, core saturation comes to mind.

Wrt to 24/240V, I think there is a significant difference. For example, when the duty cycle is 50% the time that the inductor is powered is about 20.75us. With "I" rising at V/L A/s from a 35VDC source this produces about 1.5A in the inductor, and for the 240V example, lets say DC_Bus=350V, "I" would rise to 15+ amps in the same time.

Any thoughts?

Ron.

Edit: Yes, Nard, your Schematic is correct.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

First things first: good grocious .... are you still awake ?! I didn't expect an answer from you this evening (at this side of the rock)

The load: is it pure resistive or is it a more complex load ?

When the load current is increased, the inductance of the filter-coils change when they are coming close to saturation. And as a result the resonance frequency changes. Which, on it's turn, may bring the filter to resonate.

About the 24/240V matter: I prefer to leave that for a moment (it was a long day)

I assume you're using Phase&Frequency correct PWM on 24 kHz. Please correct me if I'm wrong.

Okay, the trick: since you don't have equipment for frequency analysis, I suggest to use a laptop, that is running from battery, and with no more wires connected to it than a mouse at the max ! :)
You can use the audio-line-input to connect to the attenuator I added in the schematic. Use a program like Audacity to make a recording of the two outputs with the inverter running, with filter, without filter, with load etc. I guess you get the picture.
Choose the maximum sampling frequency the laptop-audio can provide; usually 44100 Hz

Next you can use that same Audacity to make a Fourier analysis of the two outputs (we are lucky with the stereo-facility, aren't we ?)

I will PM you for my email-address.

Now it's bedtime at this side of Mother Earth.

Nard

Edit: I just realize that it's not possible to record the inverter-output when there are no filters .... pitty huh ?

Attachment(s): 

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Attached a second test-setup ...

Could you post (links to) datasheets of the used filter-components ? That would help ;)

Cheers

Nard

Attachment(s): 

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips