I've been letting this project lie fallow for the last couple weeks. I intend on getting a dedicated lock in amplifier board built. here is the todo list for that effort.
- Characterize components:
- Put up a resistor / pot bridge, see what sensitivity we get and whatnot…
- Just do the lock in portion.
- Add back in pre-amp - don’t worry ‘bout rail protection for today - monitor
- Change midpoint resistors!
- See if ad8429 works with shorted Rg for 10,000x gain? Should be like 0.6 ohm (check math)?
- Possible re-design of input amp?
- Work noise budget!
- Test point selection
- 100 mil header
- 0603 probe pads - https://www.digikey.com/product-detail/en/te-connectivity-amp-connectors/RCT-0C/A106144CT-ND/3477801
- Get rid of ground pigtails somehow?
- Mmcx connectors - $$$
- Connector selection
- BNC - edge mount
- SMA - edge mount
- 100 mil headers
- Guard traces
- Possibly a can? Pick an optional can and footprint.
- https://www.digikey.com/product-detail/en/leader-tech-inc/SMS-201C/1798-1176-ND/7203462
- Transimpedance amp w/ switches for single ended, differential, and current inputs
- Component selection
- Switches - spdt
- Switched gain
- common numbers for measurement (1x one of the amps)
- Switch selection
- 2xN 100 mil header jumpers
- N pole N throw slide switch - $$$? Real estate?
- DIP
- amp cascade - LNA + “normal” amp
- Nominal goal for 1 nV sensitivity with long enough time base; 1nV -> ~1V
- 1e9 v/v system gain.
- Ad8429 does 1e4 v/v, leaving 1e5 remaining
- Run noise budget trade; friis.
- 1e2 + 1e3 cascade (order)
- 1e4 , 1e1 cascade?
- Run LNA at lower gain (i.e. 1e3 for each stage?)
- Datasheet + friis to make trade
- Multiple 1e3 stages would be more user friendly. Just a chain of amps w/ a single switch to enable (1e3 v/v) or disable (1 v/v).
- Noise budget for additional esd protection at input?
- Diode noise acceptable?
- Or just not do it and label as esd sensitive? Cheap enough components to blow...
- LNA selection
- Reason not to pick AD8429?
- Secondary amp selection
- PLL w/ divide for lock in?
- F, 2F modes
- PLL lock / unlock indicator LED
- Component selection, again...
- Line filter at input - concerns with saturation
- Switched enable - dip or jumper
- Power supply
- LDO selection, +5, -5, +3v3? Unless others necessary
- 7805 serries
- others
- Line filter - maybe?
- Pi filters from current design should be good.
- Beef up bulk caps
- Move indicator diodes to clean side of supply
- Reverse bias protection w/ fets - clean side / dirty side?
- Supply connector w/ bias protection
- Output integration / lpf
- Switched time constant / output buffer
- Long time constants (~10s) desireable
- Output monitor connectors
- Amplified signal
- Demodulated signal, I/Q components
- LPF output
- Phase shifters - include or not?
- How to even implement?
- Indicators, dials, connectors, etc. start pretending at being a “real instrument”
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.