prof_abrasive
Forum Replies Created
-
AuthorPosts
-
Fixed here too. Fantastic! Thanks very much @marqs. Did you end up doubling the internal rate or was it something else?
Thanks also for providing JIC files, I couldn’t get it to accept the image off any of the three SD cards I tried 😐
@marqs it goes way, way up.
Still, with an added RC pole around 20kHz, that might net another 6dB.
I tried putting a cap in parallel with the output. Conveniently, you can attach caps to ground via the AV1 OUT jack with the switch in the out position.
From the schematics the output impedance is about 1040Ω, so a 6.8nF cap puts the cutoff frequency at 22.5kHz, which shouldn’t have a noticeable effect on audio quality. This helps a lot, but it’s still fairly noisy. Increasing the cap would help a bit, but at some point it’s going to start cutting in to the audible range.
Ultimately the filtering problem will never be adequately solved with an RC filter; the difference between the passband (20kHz) and the noise/stopband (50kHz) is too small to get good attenuation. An active filter is required, which means op-amps, power, etc. Messy.
I’ve only got a VA7 – it’s my faithful (and well-thrashed) dev unit.
I wondered if the DAC output amp bypass caps had gone dry, but tacking in another 10µF in parallel on one channel didn’t seem to make a difference.
The PNP and 2x NPNs? That’s on the VA7 and VA13 also. I think that’s a startup muting circuit. Shouldn’t affect sound in my reading.
The PCM1710U datasheet seems to suggest a 3rd order active lowpass on the output. The VA13 schematics show a single pole at 140kHz instead!
I think even by ’90s standards these Saturns are pretty bad! And ultrasonic content can still cause audible problems in analog singal chains (intermodulation).
The datasheet for the ADC used in the OSSC basically says that you shouldn’t need an AA filter unless there is a lot of out-of-band noise. It doesn’t specify what counts as a lot!
My VA7 Saturn uses as a DAC a Burr-Brown PCM1710U. Looking at the VA13 schematics they’d switched to a PCM1717E. Both are designated IC22. Both seem to have the same skimpy low-pass filter. Perhaps the (newer) PCM1717E has better integrated filtering?
Looking at the audio output from the Saturn itself, I see a good 250mVp-p of noise on both channels. It looks like it’s all north of 50kHz.
There’s no antialiasing filter in the OSSC, so those inaudible frequencies get folded back into the audible band by the sampling process in the ADC.
I wonder why some Saturns do this and not others?
I get the same behaviour with my VA7 (Japanese s/n A67014671). Booting without a drive the noise is constant in the CD Multiplayer.
edit: using an RGC sync on luma cable
-
AuthorPosts
