When using tones and silly amounts of bass boost I've got the clipping leds to flash and distortion is immediately audible. The input is SPDIF from CD player.
I should have put a smiley there, but I was just kidding on the AES/SPDIF thing. 
Excuse my autism. The digital input level voltage difference is not related to clipping.
Haven't tested it personally but I read that the digital input of the DCX does not factor in Crest factor. A previous post stated something like this:
Some digital audio signal may be at 0dB, but still have 15dB of dynamic headroom.
It appears that is not the case with the DCX: 0dB on the input is with no headroom to spare (immediate clipping when reached).
Be sure to check this with test appropriate test files, I have not confirmed this.
If it does check out, -15dB on the input should set the right level to allow normal overhead. But do check it.
And in any case, the output clipping leds will be the ones to watch, with digital signal it seems the input clipping indicators aren't reliable.
This week I will be stripping output boards of two DCXs. Will post pictures + signal/noise measurements. Last time I went from -90dB to almost -109dB (DAC spec) iirc, a good start.
Excuse my autism. The digital input level voltage difference is not related to clipping.Haven't tested it personally but I read that the digital input of the DCX does not factor in Crest factor. A previous post stated something like this:
Some digital audio signal may be at 0dB, but still have 15dB of dynamic headroom.
It appears that is not the case with the DCX: 0dB on the input is with no headroom to spare (immediate clipping when reached).
Be sure to check this with test appropriate test files, I have not confirmed this.
If it does check out, -15dB on the input should set the right level to allow normal overhead. But do check it.
And in any case, the output clipping leds will be the ones to watch, with digital signal it seems the input clipping indicators aren't reliable.
This week I will be stripping output boards of two DCXs. Will post pictures + signal/noise measurements. Last time I went from -90dB to almost -109dB (DAC spec) iirc, a good start.
I did test the DCX for clipping and red lights. I don't remember exactly what I found, but IIRC red comes on just before clipping. But Scott seems to be driving past that, and hearing the distortion.
You could dial back the gain on those channels a few dB to see if that keeps the red lights off. Make up for the gain change at the amp.
If you like, I can make and post some test files for you.
You could dial back the gain on those channels a few dB to see if that keeps the red lights off. Make up for the gain change at the amp.
If you like, I can make and post some test files for you.
That was just an example for Robbintip that when the clip led illuminates it distorts, it was silly boost at 20Hz so the distortion is apparent. Normally the led never comes on in the bass channels. I can't hear the distortion in the HP when it flickers and it's only momentarily, just puzzled as to why it should come on at all with no gain/boost/eq. Presumably the input is upscaled to 24/96?
Very helpful Pano.
I had occasions where clipping occured so shortly that the VU meter registered a little bit late, VU seems to have some time-weighting to it. After clip occured, VU leds were still on when Clip was already off. This was with e.g. forgot to mute mic input on PC when inserting jack plug or (PC) ground-loop related high signal.
Ground-loop? Upsampling artefacts maybe? What bitrate do you output to DCX?
I had occasions where clipping occured so shortly that the VU meter registered a little bit late, VU seems to have some time-weighting to it. After clip occured, VU leds were still on when Clip was already off. This was with e.g. forgot to mute mic input on PC when inserting jack plug or (PC) ground-loop related high signal.
Ground-loop? Upsampling artefacts maybe? What bitrate do you output to DCX?
Sorry, I just missed your previous reply.
----
When I had the device running at work, it would sometimes do that when some heavy machinery (car wash) switched on/off.
When boosting heavily it's also needed to create headroom for calculation at the input sometimes
And if your bass channels don't light up at all normally, there is some resolution to be gained
Waveform of the input could in theory also trigger Clip led; if a square wave is long enough?
----
When I had the device running at work, it would sometimes do that when some heavy machinery (car wash) switched on/off.
When boosting heavily it's also needed to create headroom for calculation at the input sometimes
And if your bass channels don't light up at all normally, there is some resolution to be gained
Waveform of the input could in theory also trigger Clip led; if a square wave is long enough?
While commercial S/PDIF digital signal (over coax line) does indeed have lower voltage than AES (over XLR), that doesn’t translate to lower ANALOG signal levels encoded within. The DCX2496 doesn’t have a commercial S/PDIF input anyway, so the quoted statement above is misleading and useless.
S/PDIF over BNC or RCA is “commercial” whereas AES/EBU over XLR is “professional”.
strange noises - need ideas
Hello,
my DCX is fitted with both the clock and active output mods from Pilgham Audio. The upgraded device served me well for years now, but lately started acting strange.
The device is configured to output LLMMHH. when playing 10KHz on the right side, two out of the 6 outputs, left low and left mid start to output cracking sound. both high outputs are clean.
when applying the input to the left side I don't experience cracking on any output. I use the analog inputs for debugging, but the cracking sounds the same when digital input is used.
I did my best to trace the issue, and I'm quite certain that I see the signal on the oscilloscope directly on the corresponding DAC chips - left most and middle one (thick thumbs and fine spacing between the legs do not make it easy).
on the scope the noise looks mostly as downwards glitches, but it is hard to characterize.
Ideas are welcomed.
the thing that puzzles my most is the asymmetry - they all share the same voltages and clocks.
Shmulik
Hello,
my DCX is fitted with both the clock and active output mods from Pilgham Audio. The upgraded device served me well for years now, but lately started acting strange.
The device is configured to output LLMMHH. when playing 10KHz on the right side, two out of the 6 outputs, left low and left mid start to output cracking sound. both high outputs are clean.
when applying the input to the left side I don't experience cracking on any output. I use the analog inputs for debugging, but the cracking sounds the same when digital input is used.
I did my best to trace the issue, and I'm quite certain that I see the signal on the oscilloscope directly on the corresponding DAC chips - left most and middle one (thick thumbs and fine spacing between the legs do not make it easy).
on the scope the noise looks mostly as downwards glitches, but it is hard to characterize.
Ideas are welcomed.
the thing that puzzles my most is the asymmetry - they all share the same voltages and clocks.
Shmulik
Some fun with a rare digital volume control. It works like a dream still.
BMC-2
20171218 220141 - YouTube
BMC-2
20171218 220141 - YouTube
I know that design quite well ;-)
Are you saying the cracking is on the DAC outputs going to the output board?
One recurring issue here is that the flat cable(s) become unseated. Try to disconnect and re-seat the flat cables.
Does the remote level control work correctly?
Jan
This is the I2S mod
Remove TR1 and R15
Isolate/lift pins 12 and 14
Jumper pins 2 and 3 and 5
Jumper pins 4 and 23
Jumper pins 27 and 28
Jumper pins 19 and 22
Isolate pin 20 and tie to pin 23
Connect I2S leads from XMOS:
ILRCK = pin 12
ISCLK = pin 13
SDIN = pin 14
Do you know if it is master or slave I2S and if therr is reclocking? Reclocking could be worse than spdif jitter.
I am actually feeding spdif and clock from the dac and switched the spdif receiver in the dcx to non reclocking.
the only synchronous non reclocking mode of CS8420 is mode 5 and mode 4
I did mode 4: SPDIF input synchronous with dac clock
Lift pin 27 from pcb and connect to +5V via a 2.2k resistor
remove the 24mhz christal and feed the same clock that the DAC uses (take the clock from inside the DAC and feed it to one pin of the christal
Done now you have no reclocking and jitter should be as low as the DAC jitter
Unfortunatelly no hardware mode for I2S input without reclocking
Otherwise we should bypass the CS8420 but the DSP has a master I2S input therefore you need a DAC with slave I2S output. This is quite rare because all DACs hare source of the clock therefore are I2S master
It needs a chip that makes I2S master to I2s master connectivity
anyone can suggest one?
I did mode 4: SPDIF input synchronous with dac clock
Lift pin 27 from pcb and connect to +5V via a 2.2k resistor
remove the 24mhz christal and feed the same clock that the DAC uses (take the clock from inside the DAC and feed it to one pin of the christal
Done now you have no reclocking and jitter should be as low as the DAC jitter
Unfortunatelly no hardware mode for I2S input without reclocking
Otherwise we should bypass the CS8420 but the DSP has a master I2S input therefore you need a DAC with slave I2S output. This is quite rare because all DACs hare source of the clock therefore are I2S master
It needs a chip that makes I2S master to I2s master connectivity
anyone can suggest one?