Author Topic: Turbine current control using a buck converter. (Read 14768 times)

Caleb · « **on:** November 26, 2013, 05:32:44 PM »

This is my first post here. I wanted to share a circuit I've been working on for a while. I have a small VAWT and I want to control the current coming from the turbine because the power curve is quite narrow. I've tried using buck converters, but most control output current. I wanted something that would control input current (that coming from the turbine) and let the current going into the battery be what it is. Anyhow, this is a prototype circuit for a buck converter using a TL494 based on a paper "Average Current Mode Control of Switching Power Supplies" by Lloyd Dixon. I thought it may be of interest to someone. Although, there have been many posts regarding buck controllers so perhaps this is a repeat.

www.ti.com/lit/an/slua079/slua079.pdf

boB · « **Reply #1 on:** November 26, 2013, 10:57:15 PM »

Lloyd Dixon really knows his stuff ! Went to his seminars in the 1990's.

Good luck !

boB

Caleb · « **Reply #2 on:** December 09, 2013, 09:54:15 PM »

I have the schematic in EagleCAD. Now I just have to cram it all onto a PCB.
I added a rectifier section and one leg goes off to an LM2907 for an RPM indicator. The RPM sensor also feeds one of the amplifiers on the TL494 to serve as a speed limiter; the MOSFET will connect to the battery to slow the turbine in an over-speed event. We shall see how it goes.

Caleb · « **Reply #3 on:** December 13, 2013, 10:17:41 AM »

The PCB is designed and out for fab at OSH Park. This is a rendering of what it should look like (sans components). The design could use some work, but I want to see if it will function. I think a couple chips could be eliminated and components positioned better, but if one picks nits too long nothing gets done.

boB · « **Reply #4 on:** December 16, 2013, 01:43:49 AM »

You're system looks like it's working pretty well actually !

What is your battery voltage and what is the highest input voltage allowable ?

boB

Caleb · « **Reply #5 on:** December 16, 2013, 08:21:23 AM »

This is meant for a 12 volt battery. The input voltage can go up to 60 volts (in theory, I haven't tried it) with around 50 Watts of input power. It is for a small wind turbine I've been working with for a while. My thought was it should cut in at around 3 m/s and cut out at 12 m/s. So if it is producing 15 volts at 3 m/s (open circuit. There would be very little current at that wind speed), it should be producing around four times that much at 12 m/s, or 60 volts open circuit. The actual input voltage would be less because there would be more current and therefore more stator loss (I have a 9 ohm stator), but that is how I came up with that design goal. The plan is to turn on the MOSFET 100% of the time and stall the turbine if the voltage gets too high. This is for a VAWT with a narrow power curve. It can stall pretty easily.

Caleb · « **Reply #6 on:** December 16, 2013, 08:39:49 AM »

This is my wind turbine by the way. It is a 0.35 m^2 VAWT. I have not been able to demonstrate more than a few watts of power production. My thought is that the problem is the narrow power curve. If I connect it to a 12V battery, the RPMs do not go up high enough to produce good power in a decent wind; it stalls. That is why I designed the controller. The plan is to measure the RPM and program a current that is proportional to RPM^2 to get power production that is proportional to the cube of the wind speed. It isn't MPPT, but I'm hoping it is close enough to get some power out of it.

Caleb · « **Reply #7 on:** December 24, 2013, 11:36:04 AM »

Here is the board all soldered up. I am short a few diodes which will have to be ordered so testing will be delayed. I hope it doesn't smoke when it comes time to apply power.

Turbine current control using a buck converter.

Bruce S · « **Reply #8 on:** December 24, 2013, 02:57:12 PM »

My guess is that you forgot Murphy's 3-day law about putting up a turbine.

Caleb · « **Reply #9 on:** December 26, 2013, 08:19:28 AM »

The PCB is populated and seems to be working (minus a few tweaks). It woks on the bench. Now I have to finish my turbine so it can have a proper test.

Caleb · « **Reply #10 on:** December 31, 2013, 10:02:36 PM »

I was having issues with starting up the controller so I added a couple components. It's a little clunky looking, but I think it will do the trick.

Caleb · « **Reply #11 on:** January 04, 2014, 06:42:21 PM »

Just documenting more modifications. I replaced the frequency to voltage converter with a transistor and I watch for a state change on a pin of the ATTiny85 for RPM monitoring. I was using a simple resistor divider to turn the 12 volt regulator on and off. That has been changed to a circuit using transistors for better performance. I may end up getting rid of the op-amp too before all is said and done. Anyhow, I think it is ready for testing using an actual wind turbine.

boB · « **Reply #12 on:** January 04, 2014, 08:26:12 PM »

Caleb, you are being extremely thorough on your design !

I am impressed !

Did you post your mosfet drive circuitry ?

I'll look...

boB
K7IQ

Caleb · « **Reply #13 on:** January 04, 2014, 09:22:00 PM »

The drive circuit is a number of posts back. There is also a transistor circuit shown that is driven by a 10% duty cycle pwm signal from the ATTiny85 that acts as a charge pump to get the voltage on the driver up to around 24 volts to bootstrap the driver. I should read the application notes to see if there is a condition where the driver will latch up.

I'll have to update the whole circuit diagram after I record the changes. I am going to be teaching a class (Intro to Thermodynamics) starting in February, so I hope this works and I can focus on the class.

boB · « **Reply #14 on:** January 05, 2014, 01:05:42 AM »

I looked at the circuit. Looks OK.

I think that latchup is rare in reality (maybe I've had this problem and didn't know it ?)

With a non-synchronous buck, you want to be careful that the boost cap always charges
if the inductor goes discontinous and Vs goes to battery +, but at least there is a UVLO
in these things.

How high of voltage do you think you will run this ?

Also, it certainly looks like you are of the teaching type from watching your video !

boB

Caleb · « **Reply #15 on:** January 05, 2014, 08:20:20 AM »

I've been keeping 60 input volts as a target; although, I don't think it will see more than 45V from my turbine. The output current should be able to get up to around 4 amps.

The circuit lacks a provision for over-speed protection other than suddenly calling for more current than the turbine can put out. The plan was to get to around 30 input watts and then start rapidly ramping up the current draw to control the turbine speed. Seeing how I haven't seen more than 3 watts from the turbine, I'd be happy with 30 watts. The main point of the circuit is to test the theory that the reason the turbine has performed so poorly is the narrow power band.

Teaching is fun. I didn't finish my PhD so I can't get tenure. I fill in once in a while when the regular faculty are on sabbatical. Part time works well with my schedule right now since my main job is to take care of the kids.

Caleb · « **Reply #16 on:** January 09, 2014, 04:38:25 PM »

Here is the new schematic and rendering of the PCB board that is out for fab. It might be a few weeks before it comes in.

Caleb · « **Reply #17 on:** January 25, 2014, 04:14:40 PM »

This is how I plan to calculate the RPM based on the sensed value.

http://youtu.be/W4lsdi7a4Nw

Caleb · « **Reply #18 on:** February 06, 2014, 05:33:19 PM »

Here is the latest circuit working with the PMA. Next step is to try it in the wind.

http://youtu.be/QwJ0j94BuXs

Caleb · « **Reply #19 on:** February 13, 2014, 08:02:54 PM »

Here is the buck converter in action connected to a small VAWT. It made a peak of 10.6 Watts (0.78 A) but not for very long. One of these days we will probably have a steady wind to test it out with.

http://youtu.be/VS-bCZyi7xE

gzglad200 · « **Reply #20 on:** May 30, 2014, 05:59:02 AM »

I have complished the contoller have the buck-boost funciton. i used the buck-boost circut. the mcu generator 4 pwm to controle the 4 mosfet.

Caleb · « **Reply #21 on:** October 16, 2014, 08:04:24 PM »

Here is the latest board. I tried to simplify things a bit and just went with an MC34063 Buck regulator. It is also designed for about 1.5-2 amps instead of 4 amps which is closer to what is needed anyway. I like this board a bit better because it will regulate the output voltage so the battery will not overcharge. It incorporates an SCR to shut the turbine down if the voltage gets too high, and the input current control appears to be adequate. So far the Watts-Up meter has shown a peak output of 22 Watts which I an quite happy with given the size of the turbine. There are still a few things I'd like to improve on, but this has been working out pretty well.

http://youtu.be/Y39E-B-IGds

http://youtu.be/l4RjroLX8uU

Caleb · « **Reply #22 on:** November 23, 2014, 09:26:03 PM »

The MC34063 board failed, and I think it was the controller chip. My guess as to what happened: a low frequency (500 Hz) PWM was turning the MC34063 on and off with a varying duty cycle. Every time the chip would be "on" it would run right up to it's max current (set for 1.5 Amps) and I think that burnt out the switch. Anyhow, it wasn't quite up to the job. So it was either add an external switch to handle the higher current, or use a different controller. I chose the latter and put in an LM2596 chip that is rated up to 3 amps. Also, after looking at how current control was handled in one of those cheap Chinese LM2596 buck converter boards, I added an LM358 dual op-amp and mimicked their current control strategy. Here is a little video about the board. I just put it on the turbine and we are supposed to get some wind tomorrow so we shall see how it goes.

http://youtu.be/rgr3P4fQjeE

News:

Author Topic: Turbine current control using a buck converter. (Read 14768 times)