20 Feb 2006 svo   » (Master)

Here's a PCB of my power supply, this time I got really bored by straight lines and it shows.

Lessons I learned with power supply

1) It seems that a relay rated at 1A must withstand power on currents. But huge bulk capacitors at the input of circuit are effectively a shortcircuit at start. This surges the relay and fries its contacts together. At least a small reed relay can't survive that. Mine has become welded forever. A good solution would be to either use a hexfet power mosfet instead, or just put a NTC thermistor in line with relay to limit the starting current.

2) When programming a microcontroller that requires HV for programming and using its main voltage from one source and +12V from another, never ever let both grounds become disconnected when one of the supplies is on.

After I tested the power supply and soldered down the DAC to the board, I found out that the receiver doesn't work in SPI mode anymore. I haven't figured the root of the problem yet, but here's something I learned by reading its datasheet n'th time. This may not be interesting for everyone, but there are not too many CS8416 projects around and my log here shows among the first hits on google so maybe this will help someone. One very important scoop from the datasheet follows:

CS8416 Datasheet Excerpts regarding hardware/software mode selection

Excerpt one

SDOUT, Pin 26: Serial Audio Output Data (Output) - Audio data serial output pin. This pin must be pulled low to DGND through a 47 kOhm resistor to place the part in Hardware Mode.

Excerpt two

For each mode, every start-up option select pin (except for TX, which has an internal pull-down) MUST have an external pull-up or pull-down resistor as there are no internal pull-up or pull-down resistors for these startup conditions (set after reset).

I hope this is my problem. I missed this part at first. When SDOUT was floating, it could as well be assumed as hi-state and the device worked as I expected it to. When I connected its SDOUT to a finite impedance input, it has likely become effectively pulled down: pull down indicates hardware mode.

This is something I could never expect for a data out line of I2S bus interface, that has no other functions!

3:12AM Update

Yes, it were the pullups that were the problem. I enabled them (had them planted in before, just forgot about the solder jumper on the reverse. I now have a working SPDIF receiver! So happy! I guess I feel like people who assembled their radios in early 1940's. Joy!

Latest blog entries     Older blog entries

Share this page