SparkFun Forums 

Where electronics enthusiasts find answers.

Have questions about a SparkFun product or board? This is the place to be.
By skunkwerk
#101428
Hi,
I'm making a solar tracker that needs pretty high torque to move the solar panels (I'm estimating 500 Nm of torque to move all 5 arrays that are in a line). I don't need high speed at all, since it's just the sun we're tracking here... but I've been having difficulty figuring out what the cheapest option would be.

1) I initially wanted to just take a hobby servo, attach a geared sprocket, modify for 360 degrees rotation, and attach a gear reducer to it, along with a rotary potentiometer (angle sensor) somehow attached to the shaft/gear. I don't really need the gear to be in-axis with the motor, as I have lots of space. Most of the gearheads I've seen are expensive because they use planetary gears, etc. So I'm having two issues:
- I can't seem to find a large gear on McMaster, and
- I'm not sure how I'm supposed to connect a rotary pot to my gear. Something like a 500:1 gear ratio should do the trick I think
- are these reliable at all?
http://www.pololu.com/catalog/product/69
I've read that worm gears with high gear ratios are ridiculously inefficient (<20%) due to friction...

2) Would a stepper motor be a better option here? I'd still have to use gears, though...

3) If I could somehow turn the panels from the sides, I could just use the mount in the center as a lever, which would require much less torque. But how do I position and attach the motor to be able to do this?

any help much appreciated
thanks!
By lyndon
#101497
At 500Nm (over 300 ft-lbs!) of torque I think you're outside the RC servo range :-)

Are you sure about that number? Like mac said, you're in the range of industrial motor control and should probably start looking at fractional-horsepower motors and heavier gearboxes.
By esklar81
#101509
mac wrote:500 Nm? Are you sure? What speed? For 18rpm, it means 2Hp motor reducter. And output gears need to be in the industrial ballpark, no toys allowed!
Perhaps, but for solar tracking, he needs about 1 revolution/day to point the array due east at sunrise and due west at sunset.
(1 revolution/day) * (1 day/24 hours) * (1 hour/60 minutes) = 6.94e-4 rpm
so the 2 hp is reduced to ~1.4e-3 hp of delivered power. Given the required gearing, though, I wouldn't be at all surprised if the drive train losses exceeded the delivered power.

That said, I think a stepper motor is a good choice for this application. Stepper motors are usually "torquey" and allow reasonably good open-loop control of angular position. From an industrial automation perspective, this application isn't particularly demanding. From a hobby perspective, it will be a challenge.

It may be that a multi-stage chain drive (Think salvaged bicycle parts and ~5:1 per stage.) would be worth some consideration. The stepper motor SparkFun offers is spec'd at a holding torque of 0.23 N-m, so that would require a gear ratio of several thousand to one. Much more powerful stepper motors are available, often as salvage parts.

Eric
By SFE-Robert
#101665
MichaelN wrote:A satellite dish positioner might be ideal for this. Plenty of force, can be cheaply obtained, and designed to last for years outside Here's some examples:
Rotary type: http://cgi.ebay.com/SG9120-Heavy-Duty-F ... 335c4ec341
Linear actuator type: http://cgi.ebay.com/24-Satellite-Dish-A ... 5887848684
Good idea. I was going to suggest that. Beat me to it!
By winston
#101833
As it happens, I've been mulling the same problems. For vertical control, I'm considering using a jackscrew arrangement. I was looking at the gearboxes used by a radio controlled tank at the model shop and I suspect it'd be good enough to drive the jackscrew, and the Y axis doesn't really need to move very much.

For the X axis, I was considering a set of RC car wheels running on the flat surface the panel is connected to and more tank gears, or perhaps another jack screw arrangement (probably the latter since there wouldn't be potential problems with slippage, just the jack screw would need to be quite long and would need a bit more articulation at each end than the Y axis jackscrew). There would be a fixed spindle around which the whole arrangement swings some distance from the panel itself, so the wheels or jackscrew wouldn't need to generate particularly massive amounts of torque. I hadn't considered satellite dish positioners never having been into satellite TV, but that sounds like a better way to do it given that they will already be designed to last years outdoors and will probably do the Y axis as well.
By evarobotics
#101903
Electric cylinders like the ebay links above are the go.

We have 20 large solar panels - in two clusters of 10 - which track using one of these.

The sensor box (just 3 pyro sensors) and the controllers to run off your battery bank are fairly easy to find at any solar shop.

I haven't looked but I'd bet there are lots of home-brewed designs online as well.
By Grimfox
#101998
If you can find a motor with the power you need, using a POT on the appropriate axis and micro you can easily build your own servo.

I think that if you remove the weight load from the motor by balancing the solor panel (on a trailer hitch or similar ball in socket or gimbal) then you could drastically cut the power requirements.

Finally i don't know if the above calculations take into account that the panels don't need to rotate a full 360 but only 180 to get from due east to due west. Though i've never seen a panel that goes full verticle to catch first ray in the morning so something like 120deg of rotation is all that is required. (30E-30W) And that it's not rotating 24 hours a day but less even in higher latitudes. I'm jsut being nitpicky. Although i think based on the calculations all that does is further prove your point that only a tiny fraction of HP is being delivered.

Continuous motion is also not really an issue updating every hour to half hour should be suffcient, then again i'm just going with a logicial approach and no real research into the subject. Perhaps there's a study to show efficiency of PV's based on how often they update their angle and during what times of day you need to update more often to achieve higher efficiencies. Along with things like angle accuracy and multi axis setups.

All that said satelite tracker is probably a great option! making the above block of text moot...
By esklar81
#102006
Grimfox wrote:Finally i don't know if the above calculations take into account that the panels don't need to rotate a full 360 but only 180 to get from due east to due west. Though i've never seen a panel that goes full verticle to catch first ray in the morning so something like 120deg of rotation is all that is required. (30E-30W) And that it's not rotating 24 hours a day but less even in higher latitudes. I'm jsut being nitpicky. Although i think based on the calculations all that does is further prove your point that only a tiny fraction of HP is being delivered.
Yes, I did take the need for only half a rotation into account. I did this by calculating on the basis of one full rotation in 24 hours. That there's no value in pointing the panels at the sun through the Earth at night is a good reason for not building the mechanical portions to allow it, but it doesn't change the power calculation. Put another way:
(360 degrees/24 hours) = (180 degrees/12 hours) = (120 degrees/8 hours)
Furthermore, if this is built as a low-mechanical-power, slow-moving system, then it may take all night to get the panels back to their sunrise orientation.

Grimfox wrote:Continuous motion is also not really an issue updating every hour to half hour should be sufficient.
I agree that continuous motion has a trivial advantage in terms of PV performance. However, continuous motion is the mode that requires the lowest peak motor power.

Eric