An interesting method of making a worm drive is to use a flexible threaded rod turning against a gear. The gear has teeth formed by repeatedly running or pressing the threads into its surface.

This idea was obtained from an issue of Sky & Telescope for use as a polar right ascension drive.

An method to convert the linear motion of a linear actuator to rotary motion.

Virgil Vinz found this interesting window actuator from a Pontiac 6000.

The actuator is a flexible plastic strip riding in a steel track. The strip is punched with holes that fit over a motorized sprocket. Think of this as a mutant sized 35mm film sprocket. Neat Huh!

act01

Proposed but untested, well somewhat tested now, SCR/MOSFET H-Bridge.

act023p

SCR/MOSFET H-Bridge three ø

act021p

SCR/MOSFET H-Bridge single ø

ledhelo1

The next two trackers are based on the ultra low power "Beam" technology pioneered by Mark W. Tilden.

This controller is essentially 2 pairs of LED sensors in an H-bridge configuration. One H-bridge pair is for up/down motion and the other for east/west motion. This configuration has completely independent motions.

A feature of this controller is the ultra low power consumed during idle time. It's probably around 10uA or so. The voltage monitor circuit shuts down the drive and lets a low power source such as a small PV panel charge the energy storage capacitors. When sufficient voltage is detected the energy stored in the capacitors is dumped to the H-bridge circuits. Quite high currents are provided to the motors making motion is short bursts.

ledhelo2

3Ø LED Heliostat Drive I am currently working on.

This controller is essentially 3 pairs of LED sensors in a 3Ø-bridge configuration. The driver is capable of operating 3 actuators. The motors are connected in a Delta configuration. The apexes of the delta are connected to the center connection of each half-bridge. Of course only 2 actuators are required as the third is stationary and fixed.

When a pair of LEDs turns on the associated top SCR the bottom SCRs of the other half bridges is turned on and movement of the shadow mask is toward this LED.

In the absence of any turn on conditions all SCRs are off and no movement results.

If 2 or more LED turn on conditions occur multiple SCRs in both the top and bottom will fire. In many circuits this would be a disastrous condition, however, in this circuit the only thing that happens is there is no movement and the storage capacitor is discharged. Actually this condition won't happen in normal circumstances when the light is bright.

The LED sensors are arranged in a circle of 6 with a shadow mask sized to exactly illuminate each LED 50%. They are connected in pairs at opposite sides of the circle.

US4342501

Solomen US4342501 Radiant energy collector with focal point between the plane of the frame and earth
Note, there is a U-joint at the base of the reflector.

US4930493

Sallis US4930493 Multi-lever rim-drive heliostat
Note, there is a U-joint at the base of the reflector.

US5945961

Price US5945961 Satellite Dish Mount.
Note, there is a U-joint at the base of the reflector.

US4251819

Vickland US4251819 Satellite Dish Mount.
Note, there is a Ball-joint at the bottom of the reflector instead of a U-joint. Tipping of the reflector is resisted by the cross brace. A disadvantage is the motions are quit complex. OK, the U-joint has complex motion too but not as complex as the Ball-joint mount.

US4798949

Wilcox US4798949 Linear actuated optical concentrator
Note, the specific detail of using a universal joint where the lower, rigid arm attaches to the mirror. This U-joint prevents tipping of the mirror left or right. The two linear actuators provide tilt and rotation to the mirror.

US3945015

Gueguen US3945015 Satellite Dish Mount.
Note, there is a U-joint at the center of the reflector.

3phtam
3Ø Tamiya model. (Hey, it looks kind of like a little robot!)
A picture is worth way more than a thousand words. As US4798949 and US3945015 teaches us that a U-joint keeps the mirror from tipping I have a rigidly mounted U-joint at the top. The other U-joints could be ball joints if one can find ball joints that can take high angles.

US4798949 needs the other 4 U-joints because of the use of linear actuators. These generally require that the far end be prevented from rotating. The U-joint solved this problem for him. But technically, only the single lower U-joint prevents the mirror from tipping.

Note, US4342501, US4930493, US5945961, and US3945015 only use 1 U-joint.

I worked out a method where I can eliminate the need for limit switches in the actuator. If a crank is used the motion can't run into the end and jam as in a linear actuator. The protection limit switches can be eliminated. My model uses a Tamiya Twin-motor gear box model 70097. The crank mechanism can be scaled to large sizes with no foreseeable problems.

Q. What if the controller gets confused and runs over the end and the crank moves to the upper part of the motion?
A. Nothing bad happens. The controller just continues to move the crank in the same direction, (the control action is reversed), until the crank is back on bottom and normal control is resumed.

If course, there is nothing wrong with using limit switches.

In this type of mount pairs of actuators are generally operated:
1. For up/down motion the motors are driven in the same direction.
2. For East/West motion the motors are driven in the opposite directions.

tamiya
The model heliostat drive uses hobby gear boxes by Tamiya.
Model 70097
These guys have it for $9.98us. And another with good specifications.

como

MFA/Como Drills

Ondrives Ltd.

ledcook
DCB asked an interesting question. He was wondering if I could add a temperature control feature to the LED3 solar tracker. He wants to use the tracker to run a solar cooker.

He said that with a dish concentrator the temperature in the oven can get to high. He suggested an input to the tracker could move it off axis to control the temperature.

I have bread boarded a couple of circuits to try this. I also will try a thermostat version. However, the thermostat seemed kind of high in cost compared to the electronic version. Especially when settings of 350F are needed for baking.

One could get carried away with this with oven timers and such.
Thermocouples could be used here but they require opamps and cooking gust doesn't require that precise of a temperature accuracy.
I will look at positive temperature coefficient sensors such as platinum or copper types. If a sensor resistance of about 1KOhm were used these could be more rugged than the diode sensors.

Each of these circuits is connected to the LED3 5V regulator. The 2 diodes on the left are connected across R2 & R3. The essentially the circuit grounds the LED sensor pairs fooling the tracker into heading for home.

The lower diode is the temperature sensor. It probably should have a glass case to take the temperature. OK, I know 350F is way above the normal operating temperature of semiconductors, but it works in this case. You will need to mount the sensor inside the oven and make connections with high temperature wire. I have some 30 gauge twisted pair wire wrap wire that has Kynar, (?), insulation. I twisted and crimped this to the diode to make a good mechanical connection. I also added solder. The solder may melt but still makes a good electrical contact.

Other diodes may be used such as the emitter base junction of a power transistor.

The choice of components is not critical at all. The resistor values have been chosen to allow operation to about 350F.

led3cok1

Temperature control for the LED3 solar tracker. First try.
The 1N914 at the bottom is the sensor.

led3cok2

Temperature control for the LED3 solar tracker. This works much better.
The 1N914 at the bottom is the sensor.

small
The Solar Cooking Archive.
http://www.solarcooking.org/tracker1.htm
They have a neat solar cooker tracker described and sold by:
Small Power Systems
http://www.pacificsites.com/~sps/oven.html Small Power Systems
It operates as a lazy susan with a drive wheel on the side.

Differential Temperature Sensors

I've posted a couple of differential temperature sensor circuits. These are preliminary and should be used as the starting point for more in depth circuits that control output power.

pv-i2v

PV current to voltage converter.

This circuit, when used with a small single PV cell, will do a good job of measuring the solar influx after calibration.

It's based on the primary property that the short circuit current of a PV cell closely matches the incident insolation.

Tilden H-Bridge
http://www. ????
by Mark W. Tilden
This is a very neat design. Well worth studying how it works.

BEAM Solar Engine
http://nis-www.lanl.gov/robot/html/plans.html
by Mark W. Tilden

Micro Power Solar Engine
http://www.xs4all.nl/~sbolt/e-lezersbijdrage.html
by Ken (K.G.A.) Huntington

Alf
http://wollongong.apana.org.au/~ben/alf/schematic.html
by Ben Hitchcock
A solar powered Alfa Romeo Matchbox car.

Fred
http://wollongong.apana.org.au/~ben/fred/schematic.html
by Ben Hitchcock
A little robot that searches for light.

D2 Robotics
http://www.panunzio.demon.co.uk/circuits.htm
by David & Dean
Many tiny robots.

Dave
http://www.geocities.com/SoHo/Lofts/1121/beam/dave.htm
by Justin Fisher
Tiny photovore.

Solarbotics Ltd.
http://www.solarbotics.com/

Skitterbot
http://www.working-ideas-co.com/
Skitterbot.
Cheap PV cells.

Forever Flasher
http://www.tripoint.org/kevtris/flasher.html
Kevin Horton

suntrek
Suntrek
http://www.suntrekenergy.com/
Small solar panels.

pagermotors
PagerMotors.com
http://www.pagermotors.com/
Small solar panels and pager motors.

cdsrely1

This is a non electronic analog tracker. It's very simple using only two relays and two Cadmium Sulfide, CdS, photoresistive cells.

The relay is from Radio Shack 275-249A. The coil is rated at 12VDC and 200 ohms.

The CdS cell has a full sunlight resistance of 50 to 75 ohms. This resistance is a little low for reliably dropping out of the relay when in shadow. I adjust the resistance by using a Sanford Sharpie Permanent marker. If you get too much marker on it can be removed with fingernail polish remover or acetone.

This tracker is not as accurate as the electronic ones. It's perfect for flat panels thermal or PV panels.

ledshex2

This is a greatly improved version of the LED sensor circuits with Schmitt Trigger Logic and MOSFET power drivers. This circuit fits onto a .7"x1.4" circuit board. The MOSFET drivers are capable of delivering about 10A to the motor load. They can do this because the power on duty cycle is about 10%.