Sunday, August 20, 2017

HAWT DIY Aug 2017 Data Collection 1

This is finally showing a real set of believable repeatable data, this is very gratifying.
This has been many years in the making and is simply typical of my relentless Hammer and Tong approach to Making and discovery.

My latest method of data collecting is by using my video camera. 
Since reading all three numbers simultaneously is so difficult I realized that all I needed to do is video the gauges and step thru them reading the values and inputting them into XLS

Of course a direct upload into a graphing software would be the real answer however the complete set up and remote connection is a huge project.

 

The 1st set of points I took with the V2a code below



The results (especially the steps) were perplexing for a moment but then I realized that these steps were actually the relay on - off points
The 1st step at 13v and the 2nd at 18v

Then I decided that I didn't like the particular relay steps in V2a so I made a change to the numbers below






In V2b I have all of the relays on at 6.5v and only turn them off at the rated voltage for each of the bulbs

The data collection results show a different step point


Of course we are only seeing a very small range because the wind speed was in a really small range

What we are seeing is the switching off of the 1st relay at 13v resulting in a swift rise in voltage and a dramatic drop in Amperage

Thursday, August 17, 2017

HAWT DIY August 2017 Light Array Experiment

After the July "proof of concept" with the relay-based light array I went for a more 2017 version.

Looking at the Arduino stuff there are a huge amount of options out there; I got a compatible 4 relay board for less than I paid for one relay on the 1st system.

The Uno is in the foreground and the relay pack is behind it
From the other end are the different bulbs and the power pack for the Uno

Tupperware enclosure is the cheapest and most secure for out doors
Added On-Off switches for both the Uno power and the generator output
The cable connection to upload to the Uno has a port in the side and a rubber plug

The main problem is that uploading a new code is time consuming (not the Uno part)
I have to figure out how to make the box easily removable to bring it onshore because I am not comfortable bringing the laptop out to it
Verification of the system is super easy on the bench; the power supply acts just like the generator
Went on-line to find a code guy to write the Uno C++ code.
Found this sourcing site Guru.com
From there I submitted to two different companies, and EmbatronicX got back to me in a day

The reason I went this rout is because I wanted to have the new system up and running from start to finish in just 8 days, and it worked.
Suraj contacted me I went thru the payment system and they made it on time, I had a couple of change requests and they took care of them promptly. From this experience I recommend working with them, I hope to go on from here and will need their help in the future.

I have studied the code a little and I can and do write G code for CNC so I get the concept, but learning a new code and debugging it would have taken a little more than 8 days
I also needed to order all the parts and build-out the entire assembly so I was doing that while waiting for the code to arrive

This is the original set of variables per my instructions that were written into code:


Relay 1 Relay 2 Relay 3 Relay 4





Bulb Volts7.0013.00 18.00 28.00
Voltage ON 2.80 7.30 13.00 17.80
Voltage OFF 6.80 12.40 17.20 27.00

These result in a very unsatisfactory experience when the winds aren't steady (wind is never steady)
My original idea was to have a 1/2 volt between the off of one bulb and the on of the next; in gusty winds its just a lot of flashing.

So now I'm going to change the bulbs, the voltages, and the on-off set points, to get the reaction I like best from the ever changing generator output
My new theory is to have 2 or 3 bulbs on at any given time




Now I have also made light management adjustments
I blocked out most of the translucent cover except for a 1/4" slot at each bulb center, then to improve on that I made a close-off between each bulb to limit light bleed; much better.
Since the bulbs are too close together for the distance they are from the viewer these changes help the on-shore viewing experience by separating the light more discretely

The new On-Off sequence:





V2a 081717 Relay 1 Relay 2 Relay 3 Relay 4
Bulb Volts 13.0 18.0 28.0 37.5
Voltage ON 6.5 8.0 10.0 13.5
Voltage OFF 13.0 18.0 25.0 37.5



The new code makes certain that at least 3 lights are on at a time in the middle of the array, this is visually more "readable" allowing your eye to track the direction of the volts, and increases the watts thru the middle.
Have had a 12hr period of sustained hi winds from 0 - 17 with some long blasts and gusts of 24mph
video

The shows voltages going from 5 - 25+ the system I think, shows that I have the capability of charging 12v
I need some nicer weather to put all my data collection gear (mph, volts, amps) out and get new numbers with the Uno

The goal for this summer was to have a real time, any time view of the working reliable system
The 8 AA batteries lasted about 40hrs, this is of course not acceptable.
Obviously the turbine should be charging these batteries, I think that will be the next step for spring 2018


Wednesday, August 2, 2017

HAWT DIY July 2017 Light Array Experiment

Trying to have an "its doing something" machine I looked long and hard for an LED array voltage meter but to no avail. I need / want something that covers around 4 - 35 volts DC and most everyone is using them for battery charge indicators; they just don't have a wide range.

The object is to have something real time active that can be seen from shore.

I am going to try the LM3914 but getting the range I want is still a challenge for me.

I went old school electro-mechanical, like might have been found around WWII

As voltage rises the little 9mm bayonet bulbs slowly brighten and then the relay shuts down and lights up the next one.

Its working perfectly within the component constraints

The 1st relay is a 6V which switches at around 3.5vdc
That lights up the second relay, a 12v which switches at 7vdc
And that lights up the 3rd relay, a 24v which switches at 13vdc
That turns on the last one.
 And I learned that using relays like this is troublesome; their switching voltage is different from their rated voltage (need spec sheet for each) and they remain switched until the voltage drops substantially below the switching voltage.

The incandescent are great because they go dim to bright and back unlike LED.

They are using some serious amperage too so the start up wind speed is around 3 - 4 mph

video

You can hear the little stepper motor grinding away its pretty cool. I can't get video at knight but it steps thru each bulb as designed, mostly stays on bulbs 3 and 4


video
The people at BulbTown were very bad at notifying me regarding my order, I needed a progression of bulbs to make this contraption work and they didn't get them out to me yet.
when I get a 3 volt bulb the 1st stage will light up to the max, and when I get a 24 volt bulb the 4th stage will not burn out!

Its looking like I get very roughly 1 volt per MPH

I am going to try to get amperage / wattage numbers next

I started to look at the wagon-wheel effect too, I should be able to get a more accurate TSR by videoing the windmill and the anemometer at the same time. I think that when the blades appear to stop with 6 blades showing that means 60 RPM

Convinced now that I like the incandescent bulbs because they put a heavy load on the machine.
I have figured out where to get a reliable series of 3v, 6v, 9v, bulbs (Mag-light)
And am going to figure out the Arduino controller with relay pack some how.

Best results from a day of readings