Yesterday I brought my DAC to work to
adjust the output of the DAC once more using a scope this time.
We made a nice setup with an
Audio Precision analyzer to generate a 0dB/1Khz sinewave test tone to feed the coaxial digital input of the DAC. An oscilloscope was connected to the analog output of the Valab so we could see the sinewave on the scope and adjust the DAC.
Well, what we saw on the scope is not what you would like to see: A kind of rippled sinus which got a bit better on lower frequencies but it really became problematic above 4-5 Khz. On 10 Khz there was hardly a sinus recognisable anymore. If you would see this without listening to the DAC you would think it was broken...
Could it be something with the scope? We connected a Benchmark DAC-1, Apogee Rosetta 200 and Prismsound Dream DA-2 but they all gave a perfect sinus.
Could it be a problem with my DAC? I took my other, non modded Valab and connected it: A perfect sinus... What the heck!?
Did my heavily modded Valab suddenly die? Quickly back to our listeningroom to listen. No problem at all. 'Playing the stars from the sky' as we say in dutch... What the...!?
We had no clue until I thought of our SSOM: the bypassed filtersection.
I took my non-modded stock Valab and did a quick SSOM. Back to the scope with it. Ahh, there you go: the same disaster above 5Khz as I saw with my DAC. After doing a few more tests we concluded that the distortion is caused by nothing else than harmonics.
Not that i'm scared of harmonics, on the contrary: I love them. But what I saw here is pure distortion, and that is what we are currently listening to. The highs are not what they should be
at all, but I have to admit that they not unpleasant to listen to.
Knowing this I decided to start experimenting with the output filter stage soon. I might try a ferritebead with a nice silver mica cap connected to the starground to see if the results get better without losing that nice open sound we got after removing the filterstage.
More to come...