Anchor

Anchor will be an electronic current/voltage sink, for testing/characterizing power supplies and other similar tasks.

Status

• 2011-10-13: Built first Anchor PSU board, works as designed, putting out 3.3v! There are a couple very minor issues which have been noted here.
• 2011-10-12: Parts for Anchor PSU and logic board have been received.
• 2011-10-09: Anchor PSU PCB received, order placed for all PSU and logic board parts.
• 2011-09-22: Tagged as rev. 1 and sent to Laen for PCB manufacturing. Parts still need to be ordered.
• 2011-09-21: Added speaker and completed the rest of the routing. Added silkscreen to all connectors/test points/button/etc. One final check tomorrow, then will be sent to Laen.
• 2011-09-19: Completed rev. 1 schematic, started working on board layout and got about 80% completed, remaining to complete are the USB port, one of the two expansion slots, and GND.
• 2011-09-18: Worked on schematic, completed the two ADC sections and the DAC section for measurement and load driving.
• Pre-September 2011: Put together majority of desired specs and started putting BOM together, put together UI part of the schematic.

Specs/Brainstorming

• Desired features/specs:
  • Constant current and constant voltage options
    • Maybe constant power too?
  • Ramp-up/down automatically and track outputs
  • Pulse load to test transient response
  • 5A/100V maximums
    • 500W dissipation in a FET is a bit nuts, so it won't be 5A *at* 100V
    • 100W continuous, 500W pulse might be reasonable? Would be nice anyway.
  • Some kind of computer interface for more complicated tests/more detailed data analysis
    • USB would be easy to implement
      • Implemented on rev. 1
    • RS232 would be even easier to implement
      • Implemented on rev. 1
    • Ethernet could be cool, but that seems overkill here
      • The LXI standard seems useful for this, could be implemented on an expansion card
    • GPIB would be the traditional choice for test gear, but I've never done any work with it
      • Could be implemented on an expansion card later if there is any need
  • Add-on module system for adding extra monitoring channels for things that might be expensive to implement or not commonly used
    • Temperature monitoring (a couple channels worth) would be good, as you often want to monitor chips/pass elements/etc. while load testing supplies
      • Everything to do with thermocouples is expensive, so making this an add-on module would be good
  • Direct connection to mains, as power bricks are annoying in hackerspaces where gear is moved around a lot
• PIC24FJ256DA210 will be used as it has all the graphics/LCD drive/acceleration built in to enable a nice colour TFT display
  • Inputs
    • Digital
      • Load On/Off
      • Event trigger switch
      • CC/CV mode?
      • 3 for UI quadrature encoder
      • 2 for touchscreen
    • MCU-integrated 10-bit ADC
      • Heatsink temperature
      • 2 for touchscreen
  • Outputs
    • Digital
      • CC/CV mode relay (if required)
      • Scope trigger output
      • Load disconnect relay
      • 3x SPI /SS outputs
  • Communication Busses
    • I²C
      • AR1020 I2C resistive touchscreen controller (will try to eliminate it and do it in the PIC for revision 2)
    • SPI
      • 2x LTC2450 SPI 16-bit ADC for current and voltage measurement
      • 1x MCP4921 12-bit DAC for driving the MOSFET
• Want a graphical LCD for the output/UI, and the PIC chosen will drive a nice TFT
  • WQVGA is a nice size, can get a menu/buttons on the side and still have room for the info display
  • The Newhaven NHD-43-480272MF-ATXI-T-1 from Mouser seems the best option for size/touch/resolution

Revision 1 Issues

PSU

• Fixed on rev. 2 - Fuse holder pins are bigger than holes
• IEC connector ground tab would interfere with caps if anything was connected to it
• Fixed on rev. 2 - Up on power switch is off, down is on
• Fixed on rev. 2 - Plastic on output connector covers URL
• Added on rev. 2 - No power LED on the board (as the project was going to be a single board, there was one power LED which ended up on the logic board when it was split)