Week 5: June 10th - June 16th

Individual Log - Prem

The logic controller design was finalized this week so that efforts can now be directed at assembling the water interfacing parts for the upcoming week. Modifications and additions to the logic controller are as follows.

The 555 timer circuit was redesigned to reduce circuit complexity by replacing the voltage divider network with a potentiometer and increasing the resistor values to reduce the capacitor size in order to minimize power consumption.

The discharge resistor was also removed since the duty cycle can now be manually tuned.

The leak detection circuit, thermistor, and generator detector now all run on the MCU 3.3V power output since it uses a high efficiency buck converter, and the circuits also dissipate less power since they are now operating at a lower voltage.

The logic controller now has the ability to switch in and out of low duty cycle mode. This command, which is labelled as "trim_mode" can be activated by the MCU whenever the water temperature gets close to the requested value, and disabled otherwise.

The final complete valve actuation circuit is shown below with a brief explanation below it.

If no cool signal is received, relay X2 will apply 12V continuously to open the hot valve and close the cold valve. This mechanism is made possible by the fact that the valves will ignore commands and turn off whenever they are at their maximum positions.

If the MCU sends a cool signal, the logic will be inverted and the hot water valve will be closed while the cold valve is opened.

When the MCU sends a trim signal, relay X3 will power the valves with the low duty cycle signal instead of 12V to increase the precision of the system.

Lastly, whenever a leak is detected, three relays will be activated to close the valves.

• Relay X5 will intercept the cool signal which forces the cool valve to close while preventing it from opening.

• Relay X4 will intercept the hot open signal.

• Relay X1 will send a manual close signal to the hot valve.

The only caveat to this system is high current consumption during leak detection, and the fact that the system is never at rest and will always seek. At the low duty cycle however, this behavior does not lead to substantial increase in power consumption, and results in less burden and overhead for coding the MCU since it only needs to provide a status.

The power consumption of the system in various modes is shown below

At rest: 31.5mA (9.26 days of standby time)

During Use: 32.89mA (137.11mA left of 170mA budget for battery charging)

Leak Detection: 98.7mA (2.95 days in fault mode)

There are brief transients of 145mA, but these will exist in 3 second intervals whenever the system is turned off.

The final logic controller board is shown below.

There were also some minor additions to the board such as input for an external charger, and a buffered input to drive the recirculation pump.

A table for respective bit values at each temperature were also generated to assist Conner with the MCU interpretation of the thermistor data since it is not linear and cannot be extrapolated from a singular value.