Concept nRF52 Target with FRAM
Design is based on Squeeze
Assumptions
Things to cut away
power supply - complete block
i2c routed to connector (i2c to RTC left)
INT to max-IC
all 7 pwr-supply signals to connector
energy management - first in parts, then completely
RTC-PWR-Switch ⇾ is a bad Idea, as we have no way to initially power MCUs (even during programming)
shared pwr-IO (mostly to management) ⇾ enough shared pins already
Things to keep
RTC
programming-pins
shared spi, i2c, max-int, rtc-int
Things to change
Shp_Bat_OK should go to some gpio (=9)
add current-protection between shared io-pins ⇾ all ICs (levelchanger, RTC, MSP, nRF) are comfortable with >10 mA drain/source so we limit at 8-10mA for ~4V max (~500 Ohm) which results in 240 Ohm for each connection of a star-topology
some more debug LEDs
RTC gets more “battery”, + 100 uF
capacity on MSP_SBW_IO is ref-design (probably debounce for RST) ⇾ set DNP
GPIO-Pins to cape-Port, special functions bound by MSP
0:1 are UART
0:3 are SPI (with target as Controller)
9 signals BAT-OK from Shepherd
10 is missing (is now LED)
Misc Questions
open source OK
do some of D0:D10 have hidden dedicated functionality? 11 Pins to 9+1 ShpPins have to be mapped somehow. D2:3 seems to be A_In, D0:1 are uart?
one of D0:10 influences RF ⇾ D6 is LF-IO, P1.03
there are 6x free HS pins, why are they not used? space constraints?
NFC pins are special care (pwrgd H0)
really just put CLK on nRESET of nRF-Module? should be ok after reprogramming
to clear up
MSP has a lot of compromises because of smaller package
MSP GPIO 0/1 is UART (This will be routed to SHP-GPIO 7&8)
bauteile sichten (BGA, QFN), advantages QFN
simpler design and footprint
pin-def incompatible to squeece
already locally in stock
if our qfn stock runs out we could switch to BGA without changing pin-def
QFN (-1IRGZT) = 7.80€@100n; BGA (-IZVW) = 9.42€@100n
msp reset C ⇾ DNP
CoPi
spannungsbegreuntzung diode checken
PWR-LED mit Mosfet + only 220 R ? as much current as possible
⇾ implemented for v1.0, https://github.com/orgua/shepherd-targets/tree/main/hardware/shepherd_nRF_FRAM_Target_v1.0
Changes after v1.0 / Bugfixes
marking of module footprint is irritating ⇾ copper crosses more inwards
name LEDs!
C7 is 0604? use same as C2
make longer and all parts on one side? ⇾ NO
⇾ implemented for v1.1, https://github.com/orgua/shepherd-targets/tree/main/hardware/shepherd_nRF_FRAM_Target_v1.1
Hardware-Tests
see readme.md in targets/hardware-folder https://github.com/orgua/shepherd-targets/tree/main/hardware