
Jitter only really matters when it comes to converter clocks, or their sources (like word clock, in a primitive PLL like the ADA8000 seems to use). Unfortunately I realized there are yet a few more variables here we have the actual datastream but also the wordclock signal (with for instance 1 or 256 times the sample frequency), so jitter specs also need to be accompanied by which frequency we're actually talking about, which in practice they hardly seem to do, making a comparison or understanding fuzzy. Still don't get 'em ? Ask here but post the best link you could find, it's usually easier for us to add to information that's already out there than to write it all up from scratch). (Don't get some of the words/expressions ? Use Google. There you go, your very own 48k word clock generator! Parts cost (without supply and case) likely less than a fiver. Carefully re-check all your connections Connect the shell of the BNC connector to ground. Connect the other end of the 75 Ohm resistor to the center conductor of the BNC connector.

Connect one end of the 75 Ohm resistor to pins 11.18 of the 74HC245 Connect pin 13 of the 74HCT4040 to pins 2.9 of the 74HC245 Connect pin 19 (/OE) of the 74HC245 to ground. Easiest way is to wire it to pin 20 of the 74HC245. Connect pin 1 (DIR) of the 74HC245 to VCC. Connect pins 11 thru 18 (B8.B1) of the 74HC245 together in the same way. Either use lots o'little jumpers ore one big blob of solder, just be careful not to overheat the part. Connect pins 2 thru 9 (A1.A8) of the 74HC245 together. Only the '4040 doesn't have enough oomph to drive a wordclock cable, this is where the 74HC245 comes in. This divisor is available at output Q7, pin 13 of the 74HCT4040. We need a 48KHz word clock, this equals 12.288MHz / 256. Now, the 74HCT4040 is a chained divider which divides its input signal by powers of two. Connect the output of your crystal oscillator to pin 10 of the 74HCT4040 (/CP). Connect pin 11 of the 74HCT4040 (MR) to ground. Optimal placement of the caps is likely on top of the ICs (for this step, the oscillator counts as an IC). Put a 100nF capacitor across the supply pins of each IC, with one capacitor leg to the GND pin and one capacitor leg to the VCC pin. Ignore the last pin, you can leave it open. These invariably have four pins: one ground, one output, one VCC/+5V and one NC/OE. Get the datasheet for your particular crystal oscillator. See which placement allows for the shortest connections. Hint: Do a dry run on perfboard before soldering anything. a 5V power source (in a pinch, 3 AA batteries will do). This should be a 75 Ohm impedance model (although a 50 Ohm model will likely work with minimal degradation) a BNC connector, for the word clock cable. a small piece of perfboard (printed circuit board for experiments with holes on a 0.1in grid) Easiest to solder is DIP (such as Digi-Key part SE1733-ND) or metal can. one 12.288MHz 5V crystal oscillator, for 48k word clock.

Does anyone know how easy/difficult these are to make?įifteen minutes of soldering, if you don't care too much about jitter.
