PlannedHardware

This page contains hardware projects that haven't yet become professionally-designed parts of Sinara yet. These projects range from ideas waiting for further specification/funding to well-advanced physicist-designed projects that haven't been professionally developed/tested yet.

Locker

• PDH lock
• Form factor: 2 channels on a Eurocard?
• 2 channel DDS for modulation + demodulation (allows digital phase shift)
• RF out path: filter, digital step attenuator, pre-amp (copy from Urukul/BaseMod)
• Input path: mixer, active low pass filter/output buffer. Design for low latency, so this can be used for locks up to a few MHz (utilize max bandwidth of diode laser)
• Cheap FPGA + ethernet to program DDS etc

Stabilizer

• General purpose lock box
• 2 or 4 channels on a Eurocard?
• 16-bit ADC + DAC with simple filters for each channel. No PGIA etc (assume AFE is on a separate board)

• 1MSPS. 10MSPS? Ideally, aim for a few MHz of loop BW

• ADC could potentially be the nice LTC 4 channel 5MSPS LVDS.
• Cheap Artix FPGA for loop filter, communication over ethernet
• Maybe have an option so that the input can either be an analog voltage or a digital signal (frequency). Ideally, set things up so that the input is routed both to a TTL-LVDS translator/comparator (for the digital path) and to the ADC. Otherwise, have a solder jumper to switch between the two paths.

Ingester

• Fast uTCA ADC

• =100MSPS. 1GSPS?

• 8 channels?

Blaster

• 12 GSPS SAWG RTM card for direct microwave synthesis for sc qubits (issue #183)
• Details here

Thermometer: Temperature Logger

• Non-real time (?)
• Eurocard size
• Controlled via ethernet: RJ45 jack on front panel; microprocessor (same rust/smoltcp stack as IonPack?); simple, SCPI-like text-based interface
• Same power connector/power supply as Kasli (connector on front panel)
• Same EEPROM as EEMs for serial number, etc
• Maybe combine with temperature controller board
• what kind of sensor: 10k thermistor? PT100? DS1820
• What accuracy/stability?
• 4+ inputs

Laser diagnostics

• Port Oxford WaND to Kasli. (Old version here https://github.com/ljstephenson/wand/)
• Multi channel laser diagnostics
• Multiple lasers connected via a multi-mode fiberised optical switch (switch controlled via ethernet by Kasli)
• Digitises Optical Spectrum Analyser outputs (OSA)
• Logs wavelength on wavemeter
• Simple feedback loops to lock laser to wavemeter
• Nice GUI
• Detect lasers starting to go multi-mode?

Coil current stabilisation loop

• Non-real time
• Eurocard size
• Controlled via ethernet: RJ45 jack on front panel; microprocessor (same rust/smoltcp stack as IonPack?); simple, SCPI-like text-based interface
• Same power connector/power supply as Kasli (connector on front panel)
• Same EEPROM as EEMs for serial number, etc
• External sensor (e.g. LEM IT-400S ULTRASTAB) used to measure current flowing through a coil (typically 50A-400A)
• Feedback loop controls current in small bypass transistor to remove drift/noise in the coil current
• Complex loop-filter implemented in DSP to give maximum loop bandwidth: 2 zero-frequency poles to remove low-frequency noise; notch filter to remove flux-gate noise; compensation for power supply output impedance/coil inductance.
• Bypass transistor also used for feed-forwards to modulate coil current 50Hz/60Hz and harmonics to cancel ambient magnetic field noise in lab
• Version of Oxford high-current stabilisation loop, but do loop filter with DSP.
• Link to paper when published

Pylon: HV piezo driver

• Intended use-cases: ?
• Should this be an EEM to provide real-time control? No. Let's make it ethernet and PoE. 4 outputs on 4HP.
• Based on https://github.com/JQIamo/hv-piezo-driver and https://arxiv.org/abs/1609.03607
• 0-250V output in present design, higher voltages with minor modification.
• < 100uV RMS (1 Hz - 100kHz)
• All digital, no external analog input
• Sensible output current limit to avoid excessive power dissipation (means this design will have limited BW when driving large signals into large capacitive loads).

Possible improvements:

• Existing design has onboard MCU which handles frontpanel interactions (pushbutton rotary encoder; SPDT switch; LCD display). I2C exposed on backplane. Probably remove all of this and just provide an ethernet interface.
• remove MCU and control DAC directly over SPI (AD5663R; 16bit) or swap for better DAC.
• Add onboard fast DAC to drive analog modulation input
• Add more channels (how many? what connectors to use?)

Current controller

Low-noise current controller for (low-power) laser diodes. This is an iteration of the Libbrecht-Hall design, via Dallan Durfee (https://arxiv.org/abs/0805.0015).

Should this also be an EEM? (Kasli-based EEM laser controller? when combined with TEC and PZT units?)

• Based on https://github.com/jqiamo/current-controller/tree/v2-updates (branch needs to be rolled back into master)
• Current feed-forward from piezo driver
• Existing MOSFET supports 0-180mA; modifications might be necessary depending on diode voltage. Overlapping footprints for higher current MOSFETS (untested, but on the list of things to do).
• Control via ethernet and/or EEM connector?
• ~ 10MHz analog modulation bandwidth summed onto output
• cf https://arxiv.org/abs/1604.00374 (reduce noise when operating near current rail)
• low-frequency 1/f noise from voltage reference temperature instability -> use different reference.