windgen.org vawt discussion

JeffD wrote:

---

Hi GoVertical,

I can see that you are going through the same problems I went through.  Don't give up, eventually you will get it.

Yes, the circuit must share the same ground but control/signal ground needs to be kept seperated/isolated with only one common point which does sound contradictery but that video posted by definitionofis gives a good example.  IRF and Fairchild have some really good application notes on how to setup your grounds so go through their documentation, its free on their websites.  A few years ago there were some good images on the web showing ground layout for boost and buck circuits but I am having a hard time finding them now.

A note about using a driver chip that uses a floating bootstrap.  Vs has to be able to go down to ground (or close to it) and then back up in order to charge the bootstrap capacitor.  On startup with the battery connected this will never happen since the battery keeps Vs at battery voltage until the buck starts bucking.  But how does the buck start bucking if the bootstrap cap is not charged?  Sounds like a vicious circle but it is solvable.  I talked about this in one of my earlier posts and why I went to a synchronous configuration just so that I could use the low side fet to pull Vs down to ground with a quick pulse on circuit startup in order to kickstart the bootstrap.  I have to admit I never understood this method until I tried it in Spice and finally saw the light.

There are several ways to solve this problem according to the application notes put out by IRF, TI, and Fairchild but I don't know which is the best way or if there is a best way, maybe just comprimises.

In one of my first test buck circuits I also used a 555 for generating the pulse signal to the driver chip.  I was able to use the same supply voltage for both the 555 and the driver chip as long as I had caps (0.01uf) at the power pins of each chip.  I can't see why this would not work in your setup as long as your grounds are setup properly ie common ground.

That video of the MPPT charger is very similar to my setup.  I also used an Arduino board for the controller.  I used a 4x20 LCD display instead since they were on sale for the same price as the 2x20 LCD displays ($20 instead of $60).  I like the use of the Dremel tool, a very versatile tool.

---

 I saw an application note or some such document that suggested charging the bootstrap capacitor through a resistor connected to the high voltage that you are bucking down so that a milliamp or so of current flows into the cap.  Then put a zener diode across the capacitor to prevent the voltage from going too high.  But I just can't find that document again.

http://www.youtube.com/watch?v=1L86_fdsFKY 

Hi, I found a video that show another MPPT project.

I have only tried resistive loads so far. I will start searching for another reference paper. 

Maybe you can slow the oscillator down to ridiculously slow, like 100Hz, just to see the MOSFET switch off on, before you swap diodes and capacitors for higher cycle speeds. It's just a thought to help debugging.

GoVertical wrote:

---

Hi, I checked the MOSFET it is still working. I used it to turn a 12 volt light on and off in a separate test circuit. 

 

I do not have a fast recovery diode. I will have to order some. I am trying to use switching diodes because it is what I have available. This maybe the problem.  

 

At this time it appears the boot strap capacitor is not discharging when Vs from the IR2117 is connected to the Source of the MOSFET. 

 

The control signal frequency from the 555 is about 10kHz. I need to calculate the size of the boot strap capacitor. I may be using the wrong value for this frequency.

 

Thank you for the assistance, I will post results when available.  

---