Linear Technology Harvesting Chips

For harvesting there are many different situations and domains. For many of these domains Linear Technology has specialized chips.

We have used three of these:

  • piezo harvesting (buck circuit) LTC3588
  • solar harvesting (boost circuit) LTC3105 and LTC 3652 (solar buddy Sparkfun)

Buck1: LTC3588: This circuit takes care of positive/negative input (piezo).

Picture of the DIY Beamboard harvesting (enormous (- tiny)) amounts of energy from pinching a piezo element: (loading a 22microF Cap), every push produces 30.0 milliV !!!


For the connaisseur: this LTC3588 is a buck circuit (not a boost) because the Piezo produces lots of Volts (eg 14V). These Volts must be DC-DC-ed downwards to 3.3V.

I bought the LTC3588 from Farnell, but I had to solder the connections with a magnifying glass. You need an inductor and some capacitors to go with it.

You can also buy it at Sparkfun on a board:

The chip can be used to harvest energy to a certain point, at which a microcontroller can be waked up from sleep mode. Then you have a bit of energy to do something before going into sleep mode again. You can send a sensor value with a transceiver for example.

This is a movie where a LED is on after waking up, then going to sleep again:

Buck 2: LT3652

The second IC, LT3652 is incorporated in the Sunny Buddy board of Sparkfun. This chip has different properties than the LTC3105. It needs more voltage and power (starting only at 6V up to 20V) and its ambition is to charge a lipo battery. For harvesting this amount of energy in a lipo is enormous! It will take quite some time to charge the lipo. This chip has automatic MPPT.

DSC09052 DSC09053 DSC09054 DSC09055

Boost: LTC3105

This IC is for use with solar cells. It takes care of clouds: once started it stays on producing the set voltage between 1.2 and 7V by using a boost circuit. MPPT is functioning on a preset voltage. There is also power to feed a Attiny85.

MPPT stands for Maximum Power Point Tracking and tries to reach the optimal energy point in exploiting solar cells.

This optimal point is found maximizing the P=V.I relation for solar cells. This starts with lot’s of measuring and making plottings of the V, I, V*I values. From this the MPPT of different solar cells is established. This MPPT point was reasonable stable in different lighting conditions. Every solar cell can have different V, I and so MPPT values.


Below a picture of manual MPPT point exploring. There is an arduino with 2 inamps connected to analog pins. With a serial connection the arduino sends data to Processing, which plotted the above graphs. The load can be regulated using a potmeter. For the graph you have to use the computer. But there is a stand alone possibility totally self powered! The LCD connected to the arduino (both powered by solar cells) shows the maximum value. Still, a diagram is more convincing, showing a real maximum.


Setup with a self powered MPPT analysis: the arduino is powered by the left solar cells, the display by the lower cells, and the cells on the right are investigated.
This is an important video explaing the use of the harvesting MPPT IC’s:, because we don’t have an internet connection in Paillard, please look at it before!



You cannot buy this chip at Sparkfun’s. I bought it at farnell:

like the ltc3588, and i have soldered it using a magnifying glass and made my own board:(again, add one inductor and a few caps, besides two potmeters)

2014-08-12 16.24.27 as you can see there is an attiny85 attached which can indicated if the power provided is sufficient for the chip to function (amongst others).

Page with all the harvesting chips from Linear technology: