Solar Garden Light Refresh - Replace the Cells and Extend Run-time.

by holiday_david in Circuits > Reuse

1605 Views, 10 Favorites, 0 Comments

Solar Garden Light Refresh - Replace the Cells and Extend Run-time.

PNG image.png
IMG_3197.JPEG
IMG_3196.JPEG
IMG_3209.JPEG
IMG_3248.JPEG
A2DB06E0-3768-4FFF-A32D-83CD17B28AF9.jpeg

I moved in to a home with some old solar lights in the garden. The design seemed nice, and they looked to be in good condition other than the solar cells and photo resistor on top. The bright sunlight had yellowed and cracked the resin over the cell and it no longer charged the battery and couldn't light at night.

My goal was to find a circle cell with a cutout like the originals, but this proved difficult. Solar cells that fit were not common (58mm diameter) and the ones I did find were 5v, not the 2v needed for a simple single cell NIMH set up, which these were. Looking at the design of the light and circuit I was able to determine the photo resistor was not needed, so I purchased square solar cells (40x40mm) and set out to adapt the circuit and 3d model a spacer to hold the solar cell.

Supplies

Supplies needed:

  1. Putty Knife
  2. Wire Cutters
  3. Soldering station with small tip
  4. Screwdriver (PH1 head)
  5. Solar cell
  6. 1 mohm resistor
  7. Sealant type adhesive
  8. 3d printer - to print the model (optional)

Disassembly and Testing

CEED9CFF-9B4F-4961-8D8E-1377E34F8C61.jpeg
3DCED57E-DA78-4DF6-BFD0-323D65D52339.jpeg

Start by separating the solar cell and led portion of the unit from the stand and reflector. In my case a simple twist of the decorative top (1/4 turn) let me easily remove the portion containing the solar cell and battery.

On this unit the internal circuitry was held in a plastic compartment secured by 4 Phillips screws. The solar cell was secured by a sealant adhesive which was difficult to remove. Using a putty knife i was able to gently pry around the edges and slowly loosen the cell. Heat may help here as the adhesive was silicone and likely more pliable when warm.

After detaching the cell I de-soldered the wires from it. This is optional, you can just clip the wires near the cell as you will not be re-using it. At this point I also removed the internal circuit board as the leads to the solar cell were the only thing preventing its removal.

With the unit in a disassembled state, inspect the interior for water and other damage. A good method to test a unit that will not turn on is to install a freshly charged battery and cover the sensor. With the switch in the on position, the led should turn on when the sensor is covered and off when exposed to bright light. If it doesn't operate as expected simply replacing the solar cell is not likely to resolve your issue.

Modify the Circuit

IMG_3251.JPEG
IMG_3250.JPEG
IMG_3346.jpeg
IMG_3345.jpeg

The circuit in my solar lamp used the common QX5252F in TO94 packaging (QX Micro devices or a clone). The available datasheet for that chip had two circuit diagrams, one with and one without a photo resistor. A few simple modifications to the circuit would allow the solar cell to pull double duty as both the source of power in the day and illumination control at night.

You may not wish to remove the photo resistor but the complication of a larger solar cell, 3d printed holder and maintaining water tightness of the refurbished unit led me down this path.

Explaining this modification is difficult at my novice level but it came down to the following steps.

  1. Remove the photo resistor. (clip, de-solder, whatever you like)
  2. Remove the (62kohm?) resistor connecting ground (-) to pin 1 (SBAT) on the IC.
  3. Add the new resistor (1Mohm) Remove the Diode.
  4. Add connection from solar input to the same pin 1.
  5. profit??

So, some will ask why I did this, and others will correct my resistor values which as a novice I just accept and learn from. Why do I think I did this? Simply removing the resistor and connecting solar to pin1 didn't work. The reason, I assume is that the manufacturer has a diode from solar in to battery. This shouldn't be necessary as the QX5252F Chip has a diode integrated here as well, and I assume the additional diode allowed leakage current high enough to trigger the lighting detection. Re-installing the 62k resistor was too aggressive at draining that current and nothing but intense sunlight could turn the led off. Rather than wander through many values (or doing the math!) I found a 1mohm resistor and tried that. The solar cell is now fairly sensitive turning the light on under dim dusk like conditions.

I might be bold enough to pull both the resistor and extra diode on one of the next 4 I have to fix.

Replace Solar Cell

IMG_3252.JPEG
FG95KSGL2HLBYTG.jpeg

You have many options in replacing the solar cell. The important specs to match the original cells from my unit were as follows:

  1. Cell should fit inside a 58mm diameter circle. (with 3d design skills this is arbitrary)
  2. Cell should provide 2v. (for 1.5v nimh battery)
  3. Cell should output similar ma to the original.

The cell in the original, under all that yellow had 4 7x25mm elements(unknown rating). The new cell has 4 8x30mm elements and carries a rating of 120ma @ 2v. The QX5252F is rated to be used in this voltage range with NIMH cells, so this fits the bill. The batteries in the devices are 600mah NIMH and will charge fully in a day of bright sunlight with this solar cell. The measured load on the battery with the led on was 20ma allowing a max 30 hour(!?) runtime on a charged (600mah) cell. I plan to do some testing, additional leds and larger batteries may not be out of the question if the cells get sufficient daylight.

Installation:

I had considered a few methods to install the cell and settled on a 3d printed mount. (STL attached) Simply gluing the cell to the top would certainly have worked, but it is metal and the cell solder pads are on the back requiring space/insulation to sit flat and not contact the metal. You'd also have to cover/seal the photo resistor hole.

Instead:

  1. Use sealant glue to attach the cell to the printed holder.
  2. Solder the solar input wires to the cell.
  3. TEST OPERATION.
  4. Glue the mount to the top of the unit again with sealant type adhesive.

These will be outside, so the use of sealant type adhesive is relatively important. Water entering the unit will be destructive even to low voltage electronics like these. Testing operation should be as simple as installing a fresh charged cell, switching the unit on and blocking light to the solar cell. The light should turn on/off as appropriate.

Let it dry 24 hours, re-assemble and you are in business.