Epic fail. My circular PCB will not work. I did the testing using 5V from the arduino nano board. Without thinking through what the cell batteries will actually deliver. Looking up the datasheet for the TCS34725 I remembers that I needed at least 3V3 for the supply, also the Blue LED on the HSMF-C114 of typically 3V4. No problem - I had designed the circuit to run on 2 cell barreries, which deliver 6V.

But The cells delivery between 3V and 3V7. In reality I am getting above 3V with one cell, in fact enough to run the Green LED at 3V4. However if I run the ATTiny with one cell the voltage drops are such that the Green and blue LED are not lit.

Ok two cells.

But The cells delivery between 3V and 3V7. So this gives above 6V, which mean the ATTiny cuts out.

Just one cell - direct to Green LED

Just one cell - via ATTiny no Green

Two cells - ATTinys cuts out

Ok Give me Solutions

So how do I get around this problem ?

I discounted zener diodes, current leakage too high. Also Voltage divider for the same reason. There must be a small form IC .... After many minutes googling.

LTC3212 - RGB LED Driver and Charge Pump

From the product page

  • Power and Current Control for Driving RGB LEDs
  • Individually Programmable Current Sources
  • 1x or 2x Mode, Low Noise, Constant Frequency Charge Pump
  • Single Wire Enable Control for All LEDs
  • White Mode Adjusts R, G, B Currents for White Light
  • 25mA Maximum LED Current
  • VIN Range: 2.7V to 5.5V
  • Automatic Soft-Start, Mode Switching and Output Disconnect in Shutdown Mode
  • Available in 12-Lead (3mm × 2mm) DFN Package

I have requested some samples, and whipped up another breakout board in KiCAD order with https://oshpark.com/.

Love the costs

Final notes

https://oshpark.com/ has just emailed within a few hours....

We've sent the panel containing your boards to the fabricator. We expect to get them back around May 16th. In case you're interested, there are 82 other orders on the panel along with yours, adding up to a total of 450 boards. Neat eh?