View Full Version : Pinging those *really* familiar with SB2 board

2007-09-25, 19:32
I've removed the 11.2896 crystal. Now there's two empty solder pads: a left pad and a right pad if holding the board with the i/o jacks facing up. The crystal is directly below a 74HCU04, and the crystal has a trace that inputs into this chip (one of several traces...and I can't tell which is the input from the clock). I am set to replace this removed crystal with my DAC's master clock. Which solder pad--the left or right one--should I solder this mclk input to? It's the pad that feeds the 74HCU04. Datasheet is not helping me in this instance. thank you!

2007-09-26, 09:36
I don't have the board in front of me but you want to feed your clock into the pad that leads to the 74HCU04. You will probably also need to pick up a ground close by. From what I recall there is an option nearby.

BTW, an alternative is to cut the trace between the 74HCU04 and the XILINK pin 42 and feed the clock directly into pin 42. This worked for me, and may give a cleaner signal. I know that it did not work for someone else so it may be clock dependent.

Also, is your clock 3.3V? If it is 5V, you will need a 100R resistor across the clock signal and ground.

2007-09-26, 10:00
"...you want to feed your clock into the pad that leads to the 74HCU04. You will probably also need to pick up a ground close by."

Thanks Jitterbug. Yes, this is what I'm attempting to do but it appears there are several traces from the clock to the 74HC (even though there are just 2 leads on the bottom of this clock). I thought I'd get lucky and see just one trace to the top left pin (input #1) of the 74HC but this is not the case as far as I can tell.

I took a look at the 74HC datasheet *hoping* that the name of chip (or at least some marker) would be on the diagram so I can determine with certainty the orientation of the chip on the SB2 board. But this is not the case! I am stumped. I have a 3.3V signal and prefer to input to the 74HC if I can find a way.

I committed a cardinal sin and forgot to document the orientation of the cyrstal on the SB2 board...before removing it...because the 11.89 is marked on the top, and if I remembered the orientation of the clock on the SB2 board I could look up the datasheet to see which side is the output...that would be another way to find out which pad to solder the new mkclk wire to.

2007-09-26, 11:03
Oh, I see what you're asking. Your want the pad that leads to pin 1 on the HCU. Pin 2 then feeds the clock to the xilinx.

2007-09-27, 10:16
Well, after examining the board closely using a magnifier (i.e. no schematic), the best I can tell is as follows. If holding the board with the i/o jacks facing up, the HCU04 is located to the right and slightly above the Xilink. With reference to the HCU04 chip, I'll name the first pin on the chip on the side closest to the Xilink as "pin 1". Following this convention, the 3rd pin has a trace to the left solder pad of the removed 11.28 cyrstal while the 4th pin has a trace to the right side. If I have the orientation of this chip correct then the HCU04 datasheet suggests:

- 3rd pin would be 2A (data input) with trace to left solder pad of removed crystal
- 4th pin would be 2Y (data output) with trace to the right solder pad

So I plan to solder the new mclk input wire coming from my dac to the left solder pad of the removed crystal...unless anyone knows that I have this backwards?

2007-09-27, 10:57
Your numbering on the HCU is correct, however you want the pad that connects to pin 1 (this also connects to pin 4 so that may be a source of confusion.) Pin 3 wont help you.

The attached partial schematic should clarify that.

2007-09-27, 13:39
I used a Tentlabs XO in my SB3.

Edit: sorry I just read that you want to use the DACs masterclock slaved to the SB2.

2007-09-27, 20:30
Jitterbug, I am very grateful! The schematic clears up everything. The right pad (with trace to cap C15) goes to pin 1 and is therefore the pad I'll solder the mclk input cable to. Thank you!

Tingtong, I followed your thread some months back and it inspired me to shift directions away from an USB implementations to SB ethernet. I have built a synch reclocking circuit following the schematic and design posted by John Swenson. I hope to have it up and running this weekend.

Big thanks to you guys for assisting!