This is a follow-up from a previous topic: Compact Voltage-Controlled Current Source Array.
What I asked originally was some proposal for a current source array controlled by a multiple channel DAC schematic. The main application is an optical phased-array antenna, constituted by resistor elements. The requirement was to have at least a 16-channel board (but up to 64) with each current source channel should provide up to 50 mA. I ended up doing a modified version of the design in this Texas Instrument application note and now I'm drawing the PCB. Each channel is a double stage amplifier in a current-mirror fashion scheme. In this modified version I replaced the DAC with a 16-channel one and doubled the components count.
After some discussion with few colleagues, I decided, as a first prototype, to make an Arduino shield for the Arduino UNO board (to save some time in design) that can be stacked up to 4 boards (in order to have 4x16 channels). Since I do not have much experience in PCB design, I would kindly ask you few problems to address:
- Power supply for the DAC and the op-amps: do you think that the +5 V from the USB port of the Arduino board is able to power up also the 4 nanoDACs and the 64 dual op-amps? According to specs, the +5V pin should not exceed 0.5 A, but I think we are very far from that. If not, could you suggest a different power scheme? Should I use a voltage regulator on each board?
- Power supply for the current mirror: since each board might sink up to 0.8 A, I don't think I can use any Arduino board pin to deliver this power supply. What is the best way to power up all boards with one connector?
- Voltage reference: at the moment, the DACs rely on the internal reference and it should be accurate enough. What if in the future I wish I could provide an external reference to minimize noise? It has to be directly on the board or could it be provided from a pin?
EDIT Details on the PCB: The schematic for 2 channels is shown below.
The DAC is this IC. The PCB layout closely resemble the one described in page 5 of the TI app note, with the differences being the double count of channels, the west connector and the Arduino-compatible pins. The output west connector will be 2x8 ribbon connector lateral socket with 2.54mm pitch. This will be used for connection on another PCB having the chip to be tested (or for the multiprobe test equipment).
The main goal of the board is the generation of periodic current ramps (sawtooth waveform), with each ramp having a different slope. The load of each channel is constituted by a 100 ohm resistor, but some chip have 600 ohm load resistors. Noise will limit the accuracy of the system, but for the time being, we don't know if noise on the Vref will be the limiting factor. The laser RIN, supplying the chip, might be a bigger source of noise. Nevertheless, the less noise the better.