After reading this blog post series, I realized that schematics for this model of iPad are available if you search around for them. This helps me, but doesn't help someone else trying to reverse-engineer a digitizer pinout for a different product, so here is some information that may be helpful:
- There are a few types types of pin: ground, transmit, receive, and driven shield pins (which my device doesn't have.)
- 5mm is a common electrode spacing (i.e. spacing between adjacent transmit lines or between adjacent receive lines). The mXT1664T3 datasheet also lists 3.8mm, 6mm, and 6.5mm. You may be able to estimate how many X and Y lines there will be based on this.
- The X spacing and the Y spacing will likely be the same.
- The ribbon cable may have connections to one or both ends of the receive lines. (On this device, there are two connections for each receive line (one for each end?), except for
MT_PANEL_IN<29> which only comes to one pin.).
- The ribbon cable should be laid out in a way that minimizes mutual capacitance between transmit lines and receive lines -- transmit lines will be grouped together, receive lines will be grouped together, and there will likely be ground or driven shield pins between transmit and receive.
- Transmit lines will have relatively high voltage signals on them (e.g. 18V on this iPad), whereas receive lines will have small signals (millivolts).
- Adjacent pins on the connector will usually map to adjacent lines on the display. (Though not necessarily monotonically increasing. On mine, one section of cable counts from 0 to 29 then back down to 24.)
To minimize capacitance from transmit to receive lines within the ribbon cable, my device has wide ground traces separating transmit lines from receive lines. These are faintly visible on the cable itself as slight variations in thickness. This may give clues as to the groups of pins.
In the photo below, I've highlighted the transmit pins in red, the ground pins in green, and the receive pins in blue.
Identifying ground pins
Find another ground on the device, and use a continuity tester to identify which pins are connected to ground.
None of the pins on my digitizer ribbon cable connected to any of the case screws on the digitizer, but on the logic board side I was able to identify two ground pins. The two ground pins on the digitizer had about 320Ω between them.
Identifying transmit pins
These pins are driven by the touchscreen controller. On this device, they're driven with an 0-18V square wave, with a few dozen cycles to one transmit pin before moving on to the next.
To identify where the line connected to a particular pin is, you can tape a wire to the screen, parallel to the transmit lines. This wire should capacitively pick up some of the signal from the transmit lines, which should be visible on an oscilloscope. The strongest coupling will happen with the closest line.
Identifying receive pins
On my device, which has two pins per receive line, the resistance from one pin to the other through the digitizer is around 21kΩ. On the logic board, both pins for one line are shorted together.
You may be able to tape a wire to the screen parallel to the receive lines and drive it with a signal generator. This should capacitively couple to the receive lines closest to the wire, which you can then see with an oscilloscope.
You may also be able to watch the receive lines with an oscilloscope and use a finger or strip of foil or something to change the mutual capacitance between the transmit lines and the receive line in question. By moving the finger/foil around, you might be able to observe a change in received signal.