@xevman
Figured this post is more appropiate here
not surprising they arent hot.... thats disappointing.
I own the UDA38 Pro and although I enjoy the way it sounds its clear that it struggles in some ways, the recent measurements confirmed it.
I became less reluctant to modding it after seeing the measurements (SMD was intimidating, 1 mistake could cost you).
It uses 2 OPA2134 op amps before the OPA1622 which always seemed like its biggest flaw, the OPA2134 doesnt seem too bad on paper and it does sound nice but when comparing it in an O2 headphone amp with the OPA1612 (which the D50 uses) it's obvious the 1612 is on a different level, in fact it is the most transparent op amp available right now I believe. The 16
22 packs some serious power and is little less dry sounding at the cost of being slightly less accurate than it's predecessor , the 1612.
I replaced the 2134 with 1612 in UDA, there was no DC offset and there was sound which was a great start.
After some listening the differences were noticeable, the speed and seperation of the music had improved , particulary in the bass, soundstage is bigger and more realistic. vocals and instruments sound more distant which is less 'fun' but again seems to be result of these op amps ability to better convey a sense of space in the music.
Unfortunately both the 1612 chips are getting quite hot even with nothing playing which indicates instability, there isnt much that can be done to fix this.
However I just discovered after plugging in some sensitive IEMs that one of the OPA1612 seems to be faulty, with a very low, constant hissing.
Although both chips are getting hot I have experienced one faulty chip can effect another in a seperate channel, so there is still some hope.
If after replacing the faulty chip they are still getting hot the next replacement to try will be an LM49720/LM4562. not far off the 1612 in performance and not as sensitive to the application.