Callisto Low Voltage Modification

Lower power consumption not only reduces the need to harvest additional power under adverse conditions in order to function correctly, it opens up opportunities to operate effectively from DC power accumulators which are not tied to, or directly affected by noisy chargers, MPPT controllers, DC-DC converters, etc... and other common RFI sources.

LVM Callisto Receiver

While the nominal voltage of the +12Volt power buss within the receiver is indicated as +12V, the actual input can be anywhere between +9VDC and +15VDC.  This is indicated in the Power Supply schematic and on the PCB at KL1.  A simple and easy way to reduce the power consumed by the receiver would be to reduce the input voltage toward the lower limit of +9VDC, rather than closer to the upper limit.
The original receivers employ Qty 2,  7805 linear voltage regulators to provide +5VDC derived from the +12Volt power buss to the RF signal (IC2) and digital (IC6) sections.  The 7805 regulators have a minimum dropout voltage of 2.0V and a nominal output uncertainty of ±0.25V.  When we add the 800mV drop (210mW) of the polarity protection diode (D1), in order to maintain regulation the original receiver's input voltage must remain above +8.1VDC.

Our LVM solution is to replace the two 7805 regulators with low dropout alternatives.  The 7805 regulator has only three terminals and while many LDO devices have five or more terminals, three terminal replacement LDO devices are available.  For the analog RF circuits (IC2) and digital logic circuits (IC6), our device of choice is the MIC5209-5.0YS from Microchip, the YS denotes an SOT-223 case.  The MIC5209 regulators have a maximum dropout voltage of 500mV, meaning that in order to maintain regulation the device's input voltage must remain above 5.6VDC.  Like most LDO regulators, this device requires an additional 22μF tantalum capacitor across the output terminals.

A further improvement would be to replace the silicon polarity protection diode with a schottky barrier device.  For polarity protection (D1 & D2), our device of choice is the 1N5819.  This reduces the voltage drop across the diodes from 800mV to approximately 350mV (90mW) at +25°C and means that our modified receiver's minimum input voltage must remain above +6.1VDC.

Phase 2 Spectrometer Implementation

The Low Voltage Modification option for our Phase 2 receivers will eventually be incorporated by replacing the two off-the-shelf receivers with our own LVM option receivers.  After all, ASSA may not appreciate us fiddling about with their assets.

Phase 3 Spectrometer Implementation

 The Phase 3 Spectrometers use several DC/DC buck Converters to reduce the DC voltage from our nominal +12VDC down to around +7.0VDC for the two LVM Callisto receivers and around +6.0V for the two MWA differential Low Noise Amplifiers.  When Phase 3 was first envisaged, the high efficiency DC/DC converter of choice was the LMR66320 from TI, but this component was impossible to obtain during the COVID-19 pandemic.  Because of the lack of parts at the time, LM2596 adjustable voltage output buck converters were enclosed within diecast boxes with some RFI suppression filtering to clean up the outputs.  There is some evidence of RFI on many of our MWA Spectrometer images and improvements are currently being tested and implemented, as opportunities to reduce this noise present themselves.  Simple solutions may include replacing un-shielded network cables with shielded cables and better grounding schemes within the enclosure.  A draft image of the LMR33620 PCB now in pre-production is presented here.  We're hoping to deploy these onsite as soon as practicable.

Phase 4 Spectrometer Implementation

Initially planned for the Phase 3 Spectrometers (COVID-19 got in the way), our Phase 4 receivers and/or LNAs will have the option to operate powered by either an isolated 3S5P LiFePo4 battery (10.4V), or a 100Ah 12V lead-acid battery/solar controller/charger arrangement.   For the isolated power mode, the "on" time powers the instrument(s) for the daily measurement cycle and the "off" time returns the LiFePo4 battery back to it's charge cycle, drawing upon some of the days solar harvest from the 100Ah lead-acid battery.  We chose the LiFePo4 battery technology because we could not guarantee that the battery operating temperature would be below +60°C at any point over summer.

An LFP 3S battery (nominal 3.2V series) combination has been used to power an LVM Callisto Receiver for in excess of 5.5 hours continuous measurement operation.  Starting from a fully maximum charged state at 10.35VDC @ 230mA (2.4W) to a minimal charge state at 8.47VDC @ 225mA (1.9W). Extrapolating these figures out to a 3S5P combination should give us a continuous measurement cycle in excess of about 26 hours.

In order to prolong the operating life, we should avoid excessive discharge of the battery at any time.  The physical size of the battery was largely determined by the stainless steel enclosure that the Phase 3 and Phase 4 Callisto spectrometers would be housed within, fifteen cells was the maximum number possible.  A daily measurement cycle of 14.5 hours is usual around the summer solstice, therefore a 3S5P configuration is both adequate and practical.  Such a battery has been built for testing purposes.
This battery has been constructed using 1600mAH recycled 18650 cells, a suitably sized BMS sourced from eBay and a very simple prototyped charging circuit.