The Audio Video Jack Board is single-sided with wire jumpers and some components (mostly through-hole capacitors and connectors) on the top side. All chips are surface mounted on the bottom side, along with the printed circuit traces and most resistors. This type of construction is cheap and simple but it can mean reaching signals on the bottom side requires either wiring to them on the bottom, or drilling the board which could be difficult. Fortunately all the signals needed are reachable on top side jumpers or comonent leads.

Looking from front to back, the output board runs the length of the back panel, except for the far left side where the power supply board lives. Immediately behind the drive tray you can find 4 jumpers which carry:

  1. DACCLK -- Master Clock, 384 Fs (384 times the sampling frequency)
  2. LRCLK0 -- Word (Left/Right) Clock
  3. BCK0 -- Bit Clock
  4. ADAT2 -- Digital Audio Data, after the AC-3 switch (which, on the DV-414 is missing, along with the onboard AC-3 decoder)
In the photo, some of the jumpers are bent up so the probes had something to grab onto. GNDD is the digital ground. Because this is a single-sided board, it appears pretty often, where it must jump other traces on the bottom side.

Note that these jumpers are simply convenient test locations. It's possible that cleaner signals can be obtained nearer the IC901, 904, 906 Clock Generator chips, which are on the DVDM (green) controller board.

The digital audio outputs are on the far right of the output board. The same signal goes to the coaxial digital output (electrical) and optical, each passing in series through their own pair of 74HCT04 inverter gates before reaching their output connections. (The outputs are therefore non-inverting relative to their input.) The coax digital out is further buffered in that the second inverter stage is actually three inverter gates connected in parallel for better drive.

Tracing out the bottom, the S/PDIF signal appears at one position of a missing transformer. That hole, towards the front of the chassis, has a pointer "S" for signal. On the same (right) side of the missing coil, but towards the back, is a signal ground marked "G":

However, the S/PDIF signal at the missing tranformer holes seems to run about 1 V p-p. Instead we should be able to tap the 5 V version of the signal at the left pin of coil L602, which is fed directly from the '04 inverter. Or it may be desireable to pick up the separately buffered copy of the signal that goes to the Toslink output (or whichever digital output ends up not used).

Next: Waveforms on these jumpers