Michael Stetner / glass_udrive - History

Dev branches merged into master branch. Optimizing the board routing and
incorporating GFL comments:
===========================
1- top right of AD8224 (pins 12 and 13): don't run a trace under the corner of the chip, as per the data sheet.
2- power bypassing: capacitor placement does not agree with datasheet recommendation (pg 23 § "Power supplies" ¶ 3 sentence 2)
3- via placement for bypass capacitors should be as close as possible to the 0.1uF caps
4- power planes: you could easily use copper pours instead of traces for the copper planes here, I think you could move the few traces on the internal layers around a pour.
5- gain resistor should be as close to pin 2 and 3 as possible, with equal capacitances. I think they could stand to be closer. Maybe rotating the AD8224 90 degrees clockwise would allow for tighter routing of high-impendace low voltage signals, minimizing the chance of noise sources coupling into the system. Be wary during construction of the +-15V hoving right over the elctrode traces. As the wires move w.r.t. the high impedance traces, the capacitance will vary with the motion inducing current flow (and therefore voltage).
6- C6 may block the connector. I would see how much clearance you have and if the bend angle you're hoping for is achievable. (you can look at my boards as an example of what bend angle is possible)
============================

minor routing and dimension and docu layer changes.

solved the order of CAPCOM and Bridge balance nets and routing as per
GFL suggestions.

extended drive length to give more space at the bottom for dental cement
anchoring. PCB = 9.26mm X 19.91 mm

1-Incorporating Michael Stetner's comments:
======================================================
1. Move vias carrying output signals away from the corners of the AD8224. According to the datasheet (bottom of p. 22), vias should not be placed under the corners of the chip.
2. If you have enough space, add a 10 uF bypass capacitor on each supply, as suggested by the datasheet. These can be far from the AD8224.
3. Add a through hole for GND. This hole will be used during surgery to solder the ground wire that you put under the bird's skull. You already have one GND through hole, but that will be in use by the motor.
4. Remove the vias for +15V and -15V that are underneath the AD8224. Just run those two of the traces on the top side of the board.
5. Change capacitance of C1-C2 to 0.1 uF, per the datasheet
6. Remove the trace between the reference pads of the AD8224. This creates a loop in your ground, and loops pick up electromagnetic interference.
7. Add an orientation marker for the AD8224 on the silkscreen
8. On the bill of materials, fix the device name for the AD8224 (it says AD8424).
9. Increase the width of the traces for Ix, ICOM, +15V, and -15V. A wider trace will have a lower resistance, which is especially important for these traces that will be carrying the most current.
======================================================
2-Redesigned and re-routed to match the available componenets' package sizes.
3-Added the headplate structure with thru-holes
4- fixed the dimension layer

the modified version w/o output channel 2 and metal electrode input.

Revision AD8224 redesigned to use instrumentation amplifier with a
reference electrode. This version has 2 input and 2 INA output channels
one for glass electrode and another for a metal electrode.