SamoPower you were right on...

[Murlin]

Sliprings must be eliminated.

I might as well make a precision machine instead one where the servo/stepper motors loose ticks and or have intermittent servo power loss from using sliprings.  The advantages of using controlled yaw are huge.  U da man.

It won't be that hard. and you really wouldnt have to use a brake but like you say, its only a couple hundred more bucks to put one up there.

I have run across a ton of stand alone micro processor motion controllers out there that incorporate PID algorythms.  I kinda like a PC interphase but it looks like Microprocessors would even be easier.

Some of them store quite alot of set points and would probably work just fine for the pitch control and yaw.  The same companies usually have control cards for PC interphase too.  I only have to move about a linear distance of 6" to pitch 90 degrees.  But I like the things a pc console will do for me so I am going that way.

Servo controllers are about twice as much as steppers. I like the quiet motion of servos but steppers are alot cheaper.

I guess you could control yaw by interphasing with a weather vane, a magnet reed switch, and a couple of switches to set motor direction.

You might even have to have a 3-axis system so you wouldn't keep going back and forth between 1 and 259 degrees.

I have to go read your stuff and see what your solutions are.

Keep up the great work,

I figure about $1,500.00 extra for high tech wind turbine. Not too bad actually.

Now I know for some, that price would not be acceptable, but it is ok with me.

I love hi-tech toys  

Maybe I can figure out some way to make Uncle Sam pay for it lol....

Murlin

[scottsAI]

Hello Murlin,

I will be using:

http://www.analog.com/en/prod/0,2877,ADF7020,00.html

Look no wires!

900Mhz ISM Transceiver.

Processor: EVAL-ADUC7020MK simple $30 evaluation board.

http://www.analog.com/en/prod/0,,ADUC7020,00.html

For the price, 62k FLASH, 40Mips, 12bit ADC 1Mspe some I/O sweet. 3 ph motor control.

Much more powerful than the PIC and may of the others.

Requires C or ASM programming, other languages available.

Just great for me. I know.

Tie the two together, make a pitch controller.

Second unit at the house to send and receive info. Engage the brake...

Connect to PC serial port.

Range should be 50-100m. Bigger antenna farther.

I like ADI parts, and have used them before.

Have not picked the motor, so not picked the power electronics to go with it yet.

To deal with lightning, MUST be in metal box.

Twisted pairs for all the wires coming into box.

Connector into box, each wire must have 0.1uf cap to ground, case ground. Short wires.

If 0.1uf causes a signal problem then use what you can with series resistance after the cap.

Some additional reading would be in order. This will give a good start.

Have fun,

Scott.

[RP]

"You might even have to have a 3-axis system so you wouldn't keep going back and forth between 1 and 259 degrees."

I don't understand the part about a 3 axis system to drive a single axis servo.  

As to the wrap around 360 issue:  Plan on a 720 degree range of motion.  A simple algorythm can be employed to prevent hunting around the turnover point.

[commanda]

To deal with lightning.......

Look up MOV (Metal Oxide Varistors). Opto-couplers also come in very handy.

I've got a large manual at work by some company that specialises in lightning protection. I'll try to remember to look it up tomorrow and point you at their website. Lots of good info.

Amanda

[SamoaPower]

Thanks for the flowers murlin.

One of the best answers to the question of a good motor for yaw comes from the ham radio world and has been around for a long time. That's a surplus aircraft prop pitch motor/gear drive. When hams want to rotate extra large antenna arrays, the prop pitch is often the first choice even though they require some adaption. Years ago, they used to go for less than 50 bucks but they're now getting scarce so the going rate is more like $200-$400. They have house-turning torque and excellent braking torque and are pretty hard to break. Typically, they run on 24 Vdc.

Commercial antenna rotators are plentiful, but way over-priced. You'd pay over $1000 for a new, large one but they do come with a controller.

You want something that rotates at 0.5-1.0 rpm and has high braking torque.

I don't think the 0/360 problem is a major issue. As RP points out, some logic in the controller can handle it but a simple approach is to orient it so that point is in the least prevalent wind direction.

One thing I thought I'd try is to mount a horizontal 10' pole on the machine axis, pointing upwind. This is on a downwind machine. A wind vane would be on the end to indicate relative wind direction. This would simplify the controller since you then only have to adjust to a null, when the machine axis aligns with the relative wind.

For the pitch servo, I still think the best is a cut down satellite dish positioning linear actuator. Cut down since we only need inches of travel, not feet. This is essentially your lead screw to drive a push rod to the pitch collective. They usually have built-in limit switches and a positional feedback output. Best part is you can find them all day for less than $100.

It's true that getting complete control of our machines comes at a price. I think it's worth it.

[Murlin]

"I don't understand the part about a 3 axis system to drive a single axis servo"

My first knee jerk soulution would be to have the 3rd axis work to a limited range.

Just let the tail push the turbine around as usual, but limit the pivot angles with stops.

Then you could just use micro processors and program the set points.

The 3rd axis would be be controlled by the wind...no motors.  I am not sure if characterizing it as the 3rd axis is correct, because it pivots on the second axis, but I'll just call it that.

Murlin

[wdyasq]

Link to 'Surplus Center' search on 'linear' shows many actuators in the type and range one might be able to use.

http://tinyurl.com/epn5s

They also have a controller module one might want to look at as it has a 'preset' that might be able to be rigged to move the blades to a 'safe' position if power was lost for some reason. Although, it is more entertainment to have a large mill doing a runnaway.

Ron

[Murlin]

"Thanks for the flowers murlin."

ROFL!!!

Hey this this is not Brokeback Mountain...lol...

But seroiusly....

"For the pitch servo, I still think the best is a cut down satellite dish positioning linear actuator. Cut down since we only need inches of travel, not feet. This is essentially your lead screw to drive a push rod to the pitch collective. They usually have built-in limit switches and a positional feedback output. Best part is you can find them all day for less than $100."

Man I think you have something there...

Yep... you sure have put alot more thought into all this than I have, but I am with you on the satalite gizmo.

I am all for, as little fabrication as possible, and that sure looks like it could work great.

Man this is getting better and better.

Murlin

[Murlin]

Thanks for the link.  This project is comming together nicely.

I am getting excited now.  All the peices of the puzzle are comming together.

It is great to be able to have a place like to come together and have these types of discussions...

Something good is bound to happen.

Thanks everyone, ane especially to Dan for a great site...

Holy smokes....I have alot to do.....

Murlin

[scottsAI]

Hello Murlin,

Yaw:

Controlled rate change. Electronic would be nice.

Hi-tech is fun, the wind gen already is!

I like it simple, reliable, no maintenance, low power...

My first thoughts were complicated things. Like electronic yaw control!-)

Realizing the fundamental issue took me a few days. To state the obvious.

The tail must be of sufficient size to rotate in light winds, too small it may not track.

Variable size tail was the next direction. Taller shorter!

Scissoring tail. Too complicated. Cool.

To limit the rate the tail can rotate, need to limit the force the wind can push on the tail!

Like I said obvious. The word FORCE helped me understand. We all have our triggers.

I designed a motorized rotating tail, had a force sensor, tail could rotate flat to limit or stop the wind pushing. Too complicated.

Another day, looked to mechanical solutions. Tail with hinge, magnets hold up, wind force can break the hold and bend over, spring to return. Spring would have to be tweaked to hold up somewhat so in high winds the tail would rotate. Better than the electronic solution, not KISS.

I have had springs break and hard to buy just what you need.

Upon further thoughts.

Tail mounted on a bushing rod, so it can rotate. (pipe in pipe works too)

Most of the tail surface area above the rod, some below to balance it.

Bottom edge have a counterbalancing weight. Consider mounting the weight on an adjustable rod for up down movement to adjust how much force is allowed to push and rotate the tower. Electronic remote control up/down to tweak it. Hi-tech geeky factor. Breaks after adjusted who cares.

Light winds, full tail area. Bends over very little.

Med winds, bends over limiting the area.

High wind, bends way over almost flat.

Gusting winds, bends over instantly, no waiting! No pushing.

No gyroscopic force to damage your wind gen!

The tail structure can be weaker than before, we are limiting how much force it should see.

Well now I have a yaw rate limited tail design. Hope you enjoyed the story.

What do you all think?

Have fun,

Scott.

[Murlin]

Hey Scott,

 That is pretty good thinking and sounds like it will work great since you are using a gear at the bottom of your yaw axis and don't need any wires up on the rotation axis itself.

After I get this first Axiel Flux genny made, I may just do an induction one like what you are doing.  I need 2 turbines anyway to do what I need to do at the farm.

Keep up the good work,

Murlin

[stephent]

Also gaseous discharge tubes (the Cat's Meow for this) , or a real simple, close "spark-gap" will limit voltages across wires for lightening "suppression" (if that's even possible even though it's a trade buzzword).

A simple fuzzed out section of bare stranded wire placed here and there (grounded to the metal and the free end sticking out a few inches and "fuzzed" up a bit) on the tower or support mast will help bleed off static buildup....the precursor of a lightening hit.

Using opto-isolators can limit the damage to the tower/genny sections and "maybe" keep it out of the living quarters--but that's a .3" gap across the chip's legs and just compare that to the 3 miles the bolt has already traveled.

Dealing with lightening there just ain't no guarentees, even with using wireless links across a few hundred foot gaps...

[Murlin]

"A simple fuzzed out section of bare stranded wire placed here and there (grounded to the metal and the free end sticking out a few inches and "fuzzed" up a bit) on the tower or support mast will help bleed off static buildup....the precursor of a lightening hit."

A sacrificial node al la electrical  heh neat trick.

When the weather gets real bad, I an shutting my turbine down and firing up the deisel.  But sometimes things just might catch us un prepaired or you might have a bad case of dain bramage from the night before.....That is a perfect example of how a computer console comes in handy.

Sign on to mypc.com and do it all by remote.  Then shut the pc down until you get home, all from Starbucks or whereever you can find a public internet terminal, you dont even need to have a laptop. I'm going to network a p III into my home PC to run it.  The software will also be installed on my home PC, in case of console failure.

I would like to hear more on lightning protection also.

At my shop, I have what they call a lightning surpressor on one of the 3-phase legs in my j-box. I dont really never knew what it was, just used it to protect my maching from lightning.

It kinda looks like a huge cap, but maybe it's one of these gaseous discharge tubes you are talking about?

I never even thought about a turbine developing a static charge....

Food for thought to be sure...

Murlin

 

[Murlin]

Hey Sammo, one thought had occured about using the linear actuator.

Scott thinks that the blades need to pitch @ 20 degrees per sec.

This can be done with the satilite gizmo, but to pitch the blades using a see/saw double lever with a fulcrum, the lever that is doing the driving force, has to be 1/3 the length of the lever opposite the fulcrum, to get that kind of response time.

The travel speed is only .5" per sec on the linear gizmos.  So you would need 4 times the driving force then you would if you used the longer lever.  Note that is only if you were trying to hold a steady RPM.  20 degrees in 3 seconds may be quite acceptable, that part of the equasion is unknown.

A cheap servo can travel at 150 IPM.  If you used the longer lever to get more power, you could use a very small servo and get more power out if it.

I was going to use equal length levers and a huge servo to maximize servo life.

I bet I could update pitch so fast, it would seem instantanious with a servo/encoder/frictionless/ball-screw/nut.

Scott...I think you could do the same thing equally as well with gears, but gears would be alot more costly, because to be frictionless you would have to have more bearings.

Another consideration for using gears is the weight of rotating mass of metal.

A simple screw has far less inertia than a large gear mass to start up and slow down. Your servo would only be working in very small operating bursts because you would not be turning the shafts very many turns.  Acel/decel algorythms would not work as good.

A screw on the other hand will distribute the wear along more of an area than gears, where the wear on the gears would be more localized to one small spot in where they mesh.

Rotating 90 degrees would not use up that much of a gear if you wanted to keep the servo size down.  The gears on the blades would be large and the gear on the servo would be the small one.  And be prone to wearing out.

The thing about the gears on a car lasting so long is that they are always traveling in the same direction and not oscillating back and forth. You get a little worm gear and move it back and forth alot and it is going to wear very fast, even if you use prime materials.

You cannot temper the gears hard enough to keep them from wearing because they would be too brittle.

I worked on an old Monarch CNC machine a few years ago and it had a gear/rack design incoroprated in the tool change arm.  It was always wearing out. The small gears always seem to be the achellies heal on machines that use it alot.

The worm gear works very well when they just do intermitent jobs.

Just a couple more thoughts about things...  

Murlin

[stephent]

What you have is what commanda has suggested--a MOV.

What you NEED there is more of them to protect that machine.

one between each phase and each phase to ground.

A gaseous discharge tube is what radio/tv stations use to shunt the lightening voltage to ground off of the tower or feed lines. They aren't foolproof, but go along with about everything else you can do to protect that stuff and the wiring going into closer to home places.

Smaller MOV's usually give their lives up with a good hit from a decent sized line glitch or surge--afterwards you may or may not know if they are still working. They have been known to just slowly go into overheating just sitting on the normal AC line voltage and suddenly just flame up....you know if they are good then of course, but your carpet may suffer a bit, and possibly even your whole home.

Bigger MOV's can take quite a bit of surges--but then go poof with a very loud voice.

This is their job--they usually do it well--but don't just blindly trust them. Look at them often, and if ANY brown spots are on them--get rid of them--replace them--etc.

A gaseous discharge tube will work (over and over) until it's blown apart by substancial overload--it's readily apparent then.

Sparkgaps are easily done--just 2 sharp points just slightly farther apart then the normal voltage will jump under normal circumstances.

It's still just a part of a package to protect what you want...You need a lot of different parts and pieces and even then it may not work as planned.

Lightening does what it does. It's a horrendous amount of energy at times that HAS to go somewhere. Best we can do is make the best path the least destructive and least costly to fix.

Those Static bleedoffs (fuzzed wires) bleed the static buildup off before it develops into a "feeder" for the main bolt to follow--or so it's assumed to do.

Just a sharp turn in the wires coming from the genny will "usually" make the bolt not make the turn and jump off--most times--maybe.

The feller who can make a bolt jump off--go away--or around anything they want to protect 100% of the time will be able to hire Bill Gates for a houseboy real quickly.

[commanda]

It's pretty near impossible to protect against a direct lightning strike. What you can protect against is a nearby strike which induces voltages into that long cable run from the tower to the house.

I'm sorry, but I forgot to look up that manual at work today. Having too much fun with neo magnets and resonating wires.

Amanda

[ghurd]

"Just a sharp turn in the wires coming from the genny will "usually" make the bolt not make the turn and jump off--most times--maybe."

I believe that helps a lot, but don't hear mention of it often.

Also, several tight coils in the line, like an air core inductor, is said to help.

I do both (a couple times along the way if possible) and never had a PV hit.

Knock on wood.

G-

[Murlin]

What about trying to cancel out the electro motive force that is running around the input lines to protect from far away lightning strikes?

I have heard that by taking 2 strand romex and twisting the wires to a calculated pitch, you can cancel out the field around the wires itself.

I am guessing it is this field that is another doorway into the system.....but this is purely spectulative of course, as I have no idea what I am talking about...just tossing around hearsay   :-)...

MOV's....don't leave home without them.....hooked up.. ;-)

Murlin