Finishing up the Programmable Voltage Reference project …
Items that were left to complete from the last Programmable Voltage Reference V2.12 assembly post:
- mounting the board in the case — Done
- aluminum voltage control knob modifications — Done
- user interface decisions — Done
- possible shielding for the analog side — Done
- BOM documentation — Done
- and power supply decisions — Done
Mounting the board in the case was the easy part of the completion list. There was plenty of room compared to some of my past builds, and I could have easily gone with the 4.72″ long version instead of the 6.3″ that I used. The main reason I opted for the longer version is that it would be the same size as the Millivolt Meter and would easily stack on top of it.
I placed ferrite toroid cores on both banana jacks, and held them in place with kapton tape. To power the unit I will normally have it connected to the computer with a USB cable plugged into the Teensy 3.2 thru a rear mounted USB type B jack on the back panel. For the times that I need the unit to be isolated, it will use an internal battery pack with 6 AA cells connected thru the front panel switch.
I decided on a flat copper sheet shield the width of the circuit board and just over 3/4″ high, with notches on the bottom to clear board components. I may print a small plastic cover for the 4.096 volt reference IC to reduce air currents across it and improve thermal stability.
I modified the aluminum knob on the RGB LED rotary encoder with a light pipe panel termination. I currently have the indicator blue when in calibrated offset mode, red when calibration offsets are turned off, green when in manual offset mode, and violet when the programmed voltage differs from the actual voltage output.
To use an external voltage reference on the Teensy 3.2 board, a 470 ohm aref resistor tied to 3.3 V on the Teensy 3.X board needs to be removed.
I finished up the BOM documentation and added it to the previous
assembly post, with all parts being sourced from Digikey. I also included a CSV text file for the required board parts.
I performed an initial calibration on the 4.096 volt reference, and then stepped thru all 4095 settings and noted the difference from the actual setting. The software offset calibration written by uChip uses run-length encoded data stored in the eeprom. I only needed 8 offsets for the entire range, so I must have lucked out on a good DAC chip. After about 250 hours of power on time for the reference, I will check the 4.096 volt reference again and re-calibrate if needed.
So far I am pleased with the functionality and accuracy, especially when compared to the total cost of the board, components, and enclosure.
Total cost of the system using premium components, and custom front panel was around $200 USD. It can be built for a lot less, but my goal from the start was to build the best device I could afford, not the least expensive system possible.
A big THANKS to uChip for the initial design and post on the Sparkfun forum.
Link to uChip’s original project files with PDF documents of serial port command definitions and calibration notes.