view pirelli/chg-circuit @ 408:14302e075f37 default tip

hr-bits: further conditionalize SID-1-diff
author Mychaela Falconia <falcon@freecalypso.org>
date Mon, 22 Jul 2024 10:06:38 +0000
parents b99663e66660
children
line wrap: on
line source

VBUS from USB connector on L1: microvia at (5679,1177).  On L2 a transition is
made to two directly adjacent inner vias at (5689,1157) and (5709,1186).  On L5
another transition is made to a new pair of inner vias at (5565,1089) and
(5569,1125).  On L7 yet another transition to what appears to be a pair of
microvias to L8, guessing center coordinates at (5546,1077) and (5546,1108).
On L8 this VBUS supply passes through a series component marked "H" before going
to the rest of the circuit.  After passing through this "H" series component,
the fat power trace goes to a polygon on L8 whose bottom y coord is 1041, top y
coord 834.  This polygon contains pads for 3 caps (other sides remain to be
traced) and a microvia at (5592,1034).  The left x coord boundary of the polygon
is at 5515 where the cap is, or 5474 to the left and up from the cap.  The right
x coord boundary is at 5627.

The big polygon identified in the previous paragraph feeds VBUS power to a 6-pin
component that might be the power P-MOSFET for charging.  The 4 pins that could
constitute the drain output of this big MOSFET land on another polygon which in
turns feeds to a 2-pin component: it might be the Schottky diode.

Out of the 3 caps connected to what seems to be VCHG (VBUS passed through the
"H" component), the lower little seems to go to GND, but where does the upper
little go?  Apparent microvia from L8 to L7 at (5709,940), on L7 there seems to
be a trace going back to L8 at (5663,858) and an inner via at (5675,965).  It
looks like ICTL net!

Now let's trace the other side of the left bigger cap from MOSFET VCHG input.
L8 pad center is at (5469,950), on L7 that spot is solid GND fill - consistent
with the natural expectation that this cap goes to GND.

Let's trace the presumed ICTL net and see if it really is what it seems to be.
Inner via at (5675,965) - found it on L2.  Microvia to L1 at (2923,1163), on L1
it goes to Iota ball D6 - it is indeed ICTL!

Now let's trace the branch from VCHG (VBUS after "H") polygon that starts with
L8-L7 microvia at (5592,1034).  On L7 it goes to an inner via at (5618,1019) -
but no connections are seen on any other layer!

Let's try tracing VBUS connection to CP2102.  There appears to be an L8-L7
microvia at (5349,1392).  On L7 this spot falls on a fat power trace made out
of a polygon.  This polygon has two visible via holes at (5390,1271) and
(5631,1250).  It also has spots that may contain hidden microvias back to L8:
candidate spots (5459,1177), (5460,1059), (5635,1226).  Looking on L8, we see
that the L7 via hole at (5631,1250) is likely a microvia back to L8, connecting
to the VBUS-after-"H" fat trace, whereas (5390,1271) is probably an inner via.
The other 3 spots found from L7 are either false leads or contain microvias to
the same power trace on L8, as that's where the coordinates fall.

Looking for (5390,1271) inner via on other layers: found it on L2, connects to
L2-L1 microvia at (5371,1298).  Also on L4: a thick power trace to another inner
via at (4031,1549).  Let's go back to L2-L1 microvia at (5371,1298): on L1 it
is a test point and nothing more.  Now let's look for the inner via at
(4031,1549): on L2 it connects to L2-L1 microvia at (4000,1563).  On L1 this
microvia falls into the pad of some 2-pad component to GND, and also goes to
another microvia at (4011,1435).  On the populated pic, the 2-pad component
appears to be a diode, with the cathode connected to the power net (VCHG?) and
with the anode to GND.

Continuing the trace from L1-L2 microvia at (4011,1435): on L2 there is a power
trace to another microvia at (3003,1363).  On L1 it goes to Iota ball A5 - it
is VCHG.

The via hole on L7 at (5631,1250) appears to be an inner via after all: let's
look for it on other layers.  Couldn't find any connections.

Let's try tracing PCHG (ball B5) from Iota: it goes through a 2-pad component
(probably the precharge current limiting resistor) centered at (2862,1443) and
then to a microvia at (2849,1378).  On the populated pic, the 2-pad component
is indeed a black resistor, marked "470".  Tracing the microvia from L1 at
(2849,1378): it goes to an inner via at (2865,1350).  Found it on L4: goes to
another inner via at (5548,1198).  Looking for the new via: found it on L5, goes
to another via at (5593,717).  Looking for the new via at (5593,717): found it
on L7, microvia to L8 at (5514,741).  On L8 the microvia lies in the polygon
that connects the drain output of the P-MOSFET to the anode of the Schottky
diode, matching TI's reference charging circuit.

Now let's look for the main charging current path after the Schottky diode.  On
L8 the cathode pad is a rectange, center at (5470,621), no visible microvia
hole.  On L7 in this spot there is a polygon separate from GND pour, and in
this polygon 4 inner via holes are visible, arranged in a neat 2x2 pattern.
The upper left of these 4 vias is at (5514,538).  On L1 there is a row of 4
microvias from (5485,513) to (5542,511) - let's see if they fall into the
polygon on L2 that connects to the 4 inner vias for the charging current - they
do!  The polygon on L1 with these 4 stitching microvias is where the charging
current comes to L1 - and lo and behold, on L1 we have a series resistor marked
"R200" whose other side goes directly to the positive rail on the physically
nearby battery connector!

On L6 there is another connection to the 4-inner-vias charging current path,
possibly for the purpose of bringing this net to Iota VCCS - but the trace gets
lost in the grind-down damage.