- switches to use are DK CW102-ND

- move switches down box so they aren't spread out into gasket rim area

- add another 2-pin header to PA8 (MCO) so we can sample the master clock to
  calibrate the SOC oscillator, which means moving the RED RF strength LED to
  some other gpio line

2019.10.31
- The more I think about this project, the more I want to go back to a board
  that just does one each load cell and pressure sensor.  I just haven't ever
  actually used anything else.

  Discussion with Joe Bowen made it clear that if I were to design such a 
  board so that it could take option cards, and/or had just a pile of basic
  analog inputs and maybe switch outputs available, that doing other stuff
  could be fun, and we'd "grow the system as we need to".  Every time I tried
  to pin him down on specs, all I got was vagueness.

2020.01.18
- Maybe the right answer is to move to the ADS131A02 which is the 2-channel
  version of the acquisition chip, and costs less.  That would support one
  pressure sensor and one load cell.  Then make it possible to add on other
  acquisition boards somehow.

2020.10.20
- sticking with the ADS131A04 and routing the extra two inputs to the header
  means that supporting 2 burn wires for the strand burner daughter card is
  dead simple.  it also means that adding a second load cell and pressure
  sensor if required on a daughter card would also be dead simple.

- for strand burner, I need a daughter card that provides a high-side switch
  for pyro power (since I need a common ground through the system), and has
  resistors for 2 ADC inputs hooked to burn wires.  Bring this all out through
  a 5-pin connector to the outside that hooks to the strand burner end.  We'd
  also use the existing pressure sensor channel, but not use the load cell.

  So, firmware wants to log all 4 analog inputs all the time.  For a test
  stand run, only the pressure sensor and load cell are relevant.  For strand
  burner, pressure sensor and the two burn wires are relevant.

- for Jamie / Estes, I need a daughter card that adds at least 4 thermistor
  inputs, and support for their Sony camera trigger thing.

- 40 pin header and mating female connector for daughter cards w/q1 pricing

	header	Sullins PRPC020DAAN-RC	DK S2011EC-20-ND	0.90	
	mating	Sullins PPPC202LFBN-RC	DK S7123-ND		2.38	

- high-side switch for strand burner daughter card.  This is an 8-SO part
  good to 28V and 21A, 50 mOhm on resistance.  Should be trivial to make 
  work on a 2-layer PCB

	ST's VN7050ASTR, DK 497-15993-1-ND, 1.58

  Means we need multiple header pins carrying pyro voltage to get enough
  current, since the header spec is 3A per pin.  2-4 pins each for v_pyro
  on the pin 1 end, and gnd on the other end should do.

- for strand burner daughter, use CUI TB003-500-P06BE 6-position 5mm pitch
  screw terminal strip, supports 14-26 AWG wire, 10A per pin continuous,
  DK 102-6165-ND, $1.01 q1

2020.11.15
- thinking about connectoring again.  for the base TeleStatic, we need to
  have DC power in for battery charging, pyro output, pressure sensor, and 
  load cell.

  For battery charging, use the same powerpole and 3d-printed part we use on
  TeleFireEight.

  For pyro output, let's try a panel-mount single 3-wire AC socket:

        Schurter 4300.0701, DK 486-1082-ND, $1.48 q1, needs a complex routed
        opening, has lugs for soldering wires.  Sides squeeze to remove if
        if needs to be replaced.  Should be easy enough to CNC mill opening.

  For the pressure sensor, a 3-conductor 1/4" audio plug seems like a great
  plan, use metal-shell Switchcraft parts.  Couple that with shielded twisted
  pair cable.

  For the load cell, we need 4 conductors plus ground, so a 5-pin XLR is an
  economical possibility.  

	male chassis IO-XLR5-M-BK-JL	DK 1937-1135-ND		4.65
	female cable IO-XLR5-F-BK	DK 1937-1035-ND		5.25

  Gary's load cells use Amphenol PC04E-10-6P, which is $64/ea from Newark!
  Could offer to deliver him a cable with one of those on one end and what
  we want on our end?

  Cable that's 2 twisted pairs, shielded with drain wire, would suffice for
  the load cell and could be used ignoring one pair for the pressure sensor
  if desired.  Here's one with grey PVC jacket and foil shield:
	Alpha M4473 SL005	DK M4473SL005-ND	51.79/100ft

2021.02.09
- Grainger carries Rust-Oleum "Industrial Choice Spray Paint in Gloss OSHA
  Safety Purple for Masonry, Metal, Plastic, Wood", Item #6KP15, $6.60 per
  12oz can.  The idea of test stand electronics being purple with orange
  logo decals vs TeleFireEight being orange with purple decals might make
  sense?

2022.05.22
- looking at firmware for v4 board.  

	ads131		has driver for ADS131A0X, not quite working?
	bdale-ui	work on teststand ui

  - created new branch telestatic-v4 from master
  - merged bdale-ui, resolving conflict in altoslib/AltosCSV.java
  - merged ads131, resolving conflict in src/kernel/ao_data.h

- improper gain setting or higher than expected mV/V on load cell can easily
  drive INA125 output above the 5V allowed by ADC input.  adding a 5.1V, 500mW
  zener between INA125 and ADC to clamp that and protect the ADC


2022.09.17
- and now, for something completely different...

  What if I re-think TeleStatic to use an ESP32-S3 chip with integrated WiFi
  radio?  Imagine a board in an ammo box with an SMA female and a cheap WiFi
  whip on top of the box.  Looks over the air like an access point with web
  server.  Take any client device (laptop, tablet, phone) with a web browser
  and use it to talk to the box, check status, initiate a burn, etc.  If we
  include an SD card, then data is stored on board and can be downloaded via
  WiFi, or pulled out and parsed later.

  One really big win with this approach is that ALL the software involved in
  operation lives on the board.  Only post-processing tools might want to live
  on a separate computer.  And, frankly, if we make "raw" data avaialble in 
  CSV form and have on-board plotting with pdf output .. we'd be about done?

  The onboard 12-bit ADC seems pretty lame, but I've been planning an off-SOC
  SPI attache ADC anyway, so who cares?  The pyro circuit, LED outputs, 

  The ESP32-S3 is under $3 each at Q1 in stock at DK, a 56 lead QFN'ish thing.

2022.09.19
- sadly, the ESP32 series requires a pile of binary firmware blobs to make
  the wifi and BT stuff work, and because those execute on the same processor
  and must be linked to the firmware, it just isn't legally possible to write
  GPL'ed code for this processor series.  Sigh.

  So, Keith suggests I use an available SAMD21 part with 256k flash and 32k RAM
  and an external standalone wifi peripheral, like the uBlox NINA-W131-03B.
  You talk to it with an AT command set, so we don't have to deal with a binary
  firmware blob.  Amusingly, this seems to have an ESP32 inside!

- bought a NINA eval board to play with .. should be able to hook it to one of
  my existing proto boards?

2022.11.10
- looks like I have 25 pieces of STM32L151VCT6A, which might be well used for
  an initial TeleStatic board run?  256k flash, 32k RAM

2022.11.22
- NINA-W13 module expects an RGB led to report status.  Uses 3 pins for R/G/B,
  so we could use 3 distinct LEDs to share LED parts with TeleLaunch?  Might
  also help color-blind users?  Docs talk about blended colors, so start with
  a single 5mm ultra-bright RGB LED.

2022.11.24
- I keep pondering including the strand burner interface on the main board
  somehow.  the BTS7008-1EPP high-side power switch is only about $1.88 at
  q100, but selling it as a daughter card isn't stupid?

2024.02.06
 DANGER DANGER DANGER DANGER DANGER DANGER DANGER DANGER DANGER DANGER 

- I noticed today that the LPC11U12 parts we bought MIGHT be the 7x7 package
  instead of teh 5x5 package we use for the LPC11U14?  Why in the hell they
  have to package sizes with the same pin count is beyond me...  The easy fix
  is just use one of the LPC11U14 parts we have in stock .. might ought to 
  just sell the 11U12 inventory when I sell the CC1310 inventory?