Summary of how I made my Arcam Delta Black Box (v.1) work with the PMD100:
The final PMD100 board has the following circuits:
Here a picture of a working prototype (with an 26C32 in place of the 75ALS176B and input switching via 74LCV1G125 as shown in #80):
- changed R309 to 360 Ohm in order to correct the -6V supply to -5V supply (this should be done in any case)
- removed the input board
- wired in the PMD100 PCB as shown in post #59 (added a 390 Ohm resistor for the DATA_R line)
- connected the mute relay to the Attiny (which is also on the PMD100 board) as shown in #101
The final PMD100 board has the following circuits:
- glue logic as of #77
- coax input as of #41 but with U10 deleted and instead pin2@75ALS176B controlled
- secondary 75 Ohm resistor deleted as of #78
- the Attiny handles soft mute of the PMD100
- the Attiny mutes the output relay of the Arcam via an opto coupler which is connected as D1 of #80
- the Attiny controls the input selection by pulling the enable lines of the LVC125 / ALS176
- the front panel "Phase" button of the Arcam is used to switch inputs
Here a picture of a working prototype (with an 26C32 in place of the 75ALS176B and input switching via 74LCV1G125 as shown in #80):
Attachments
Last edited:
After playing a bit with the modded DAC I noticed that at very(!) low volumes I hear some kind of distortion which sounds similar to bad dithering.
What I tried so far:
I attach a recording (extremely raised in volume) of the noise with recordings from 3 other CD players for comparison: the 3 recordings sound fine, only the TDA1541A in the modded Arcam with PDM100 sounds noisy. Even the 14it CD-Player sounds fine! It's a DDD recording by DG from 1985.
Any ideas? Or is that normal? I remember this kind of low level noise very well: it sounds exactly like my old Sony Discman D-100 (with PCM55)...
What I tried so far:
- set dithering of PDM100 off
- disconnected my DEM reclock circuit und used standard 470pF film cap as shown in data sheet
- changed TDA1541A for TDA1541 non-A (with DEM pins unconnected)
I attach a recording (extremely raised in volume) of the noise with recordings from 3 other CD players for comparison: the 3 recordings sound fine, only the TDA1541A in the modded Arcam with PDM100 sounds noisy. Even the 14it CD-Player sounds fine! It's a DDD recording by DG from 1985.
Any ideas? Or is that normal? I remember this kind of low level noise very well: it sounds exactly like my old Sony Discman D-100 (with PCM55)...
Last edited:
Related or not trying to fix DACs by listening without measuring is waste of time. Taking a few measurements is very easy.
How can I measure that noise at all...?
Dither on the PMD100 is disabled.
Possibly! But I am not sure how to check. Do you have advice for me?
I thought the i2s timing in my scope shots earlier in this thread was correct - so LSB shouldn't be dropped?
Also strange that it sounds smooth on the 14bit CD-player...!
Can I simulate the behaviour by deleting the LSB from the ripped track and verify that it's the same sort of noise?
Last edited:
I'm fairly sure that it is measurable. And why not just make some measurements and see the results. Contrary to what some of here claim what you can hear you can also measure.
You can e.g. measure THD+N at -1dBFS and -60dBFS and compare those to TDA1541 datasheet values. JTEST would show how well your SPDIF works.
Last edited:
I enabled dither on the PMD100 and tried some test tracks: a 500Hz fade-out revealed that from around -70dB the sine starts to buzz (as my soundcard is at repair I can only playback test files or CDs for now). I found an online review stating that "listening to the fade of the dithered 500Hz tone from -60dB to -120dB on the CBS CD-1 test disc revealed a relatively pure tonal quality, which acquired a buzz of what sounded like mainly even-order harmonics as it approached, then went below -90dB" . So it seems my DAC has the same buzz but only louder. I had it recapped but kept all of the film caps - should I renew them as well?
Update: I tested a longer delay for the DATA lines (by feeding the inverted XTI to their FFs) so that the falling edge of BCK happens very shortly after the rising edge of DATA. But I couldn't provoke a different behaviour (i.e. skipping LSB or MSB). So I guess the timing is fine as it was in #58...?
Tested a couple of DEM circuits, see attached scope shots, taken from the TDA1541A pins 16/17. I think the Grundig circuit sounds best, but my problem described in #108 did not change at all. I guess it comes from bad caps at the bit decoupling pins, will change them and report...
