Maker Pro
Maker Pro

Design of electric load controller for a turbine.

Jeroen Humasol

Nov 8, 2015
68
Joined
Nov 8, 2015
Messages
68
Hi everyone,

I'm a 3rd bachelor engineering student and this summer i'm going to Peru to construct a turbine with a whole team of students. I'm responsible for the electric part of this project. The turbine will generate power for an isolated grid. Therefor we need to find a way to balance production and consumption. We will do this by using an electric load controller. This device works as follows: it measures the net frequency, once this frequency goes above 60Hz it starts dumping energy in dump loads. These dumploads should dissipate up to the total 5kW that the generator can generate.

The brain of the device is an arduino mega (because we need a lot of in out pins). The frequency measurement already works and is not an issue. The dumploads will be activated by using PWM (pulse width modulation) at 5kHz. We want to use PWM because it's better for the turbine if we're able to continuosly dissipate energy. By continuosly i mean that we should be able to dissipate 500W as well as 1467.3W for example. This is (as far as i know) not possible by using TRIAC's so thats why we opted for the IGBT.

Right now we first transform the current to DC as you can see in the attached picture and then we dump it in the dumpload. However we need a quite large capacitance (100µF) before our simulations give a descent result. And we do not really trust our simulations completely. We are going to try this circuit because we already built it but, suggestions for improvements are welcome.

The part where i really need your help for is the overvoltage protection. Normally the arduino should activate all the dumploads when the frequency and the voltage (the rise proportionally) rises. However we also want to have a hardware overvoltageprotection in case the arduino wouldn't work. This protection should activate all the dumploads as soon as the grid reaches for example 250V.

Right now you can see that there is such a protection. This works as follows: The voltage of the net is measured by a resistive divider and is called Vref (this is done at the input of the 12V regulator). The maximum voltage to compare with is created by using a resistive divider at the output of the 12V linear voltage regulator (because this maximum voltage may never change). Then at the left of the picture you can see an opamp that compares the 2 voltages and activates the IGBT if the Vref rises above the maximum voltage. However we found out that there is a problem with this system: if there is for some reason a very short voltage peak then it already activates. Of course we don't want this because the sudden increase of load damages our turbine. So we would need some overvoltage protection with a small delay to avoid those voltage peaks.

I hope anyone managed to get all the way to this sentence: any ideas?

Thanks in advance!

PS: LT-spice file of the circuit can be found in the attachements
 

Attachments

  • Schema.PNG
    Schema.PNG
    784.6 KB · Views: 95
  • Electric circuit PWM 9-mar.asc
    15.5 KB · Views: 118

Jeroen Humasol

Nov 8, 2015
68
Joined
Nov 8, 2015
Messages
68
Here are some pictures of the current highpower part of the circuit (dumpload part)
 

Attachments

  • 20160326_144320.jpg
    20160326_144320.jpg
    138.8 KB · Views: 70
  • 20160326_144335.jpg
    20160326_144335.jpg
    174.1 KB · Views: 97
Top