Now, with so many knowable people here... Is UAC-1 able to deliver 24/48, 24/96 or even higher than that?
I bougt Borges@Audiowidgets USB dac+analog board and whatever I trow upon it of what I have it plays it... Could it be that I have a +10y old SB Live! in the box? Now I disabled the SB device - the dac still plays. Could the audio-widget use SB drivers to achieve the bitrate of UAC-2? How would I check upon whats going on?
Brgds
I bougt Borges@Audiowidgets USB dac+analog board and whatever I trow upon it of what I have it plays it... Could it be that I have a +10y old SB Live! in the box? Now I disabled the SB device - the dac still plays. Could the audio-widget use SB drivers to achieve the bitrate of UAC-2? How would I check upon whats going on?
Brgds
UAC1 maxes out at 24/96 with 2 channels in one direction (you can't simultaneously record and play at that rate). The 24 bit part is, by far, the most useful--i.e. 24/44. As it allows using software volume controls with less loss of bit resolution and a lower noise floor.
Many blind tests, including ones you can do yourself with Foobar and the ABX add-on, demonstrate 24/96 nearly always sounds identical to 16/44 when the volume control is after the DAC in the analog domain and the PC volume is set to the maximum. Meyer & Moran, and others, have also demonstrated that 16/44 can be fully transparent under normal conditions as a playback format when the DAC is fed with a bit accurate datastream.
There is some benefit, however, to being able to play 24/88 and 24/96 content (such as sold by HD Audio) as it's often better mastered than its 16/44 counterparts. It's not the higher bit depth or sampling rate that sounds better, it's the audiophile mastering.
EDIT: To answer the second part of your question I'm guessing your operating system is performing the sample rate conversion for you.
Many blind tests, including ones you can do yourself with Foobar and the ABX add-on, demonstrate 24/96 nearly always sounds identical to 16/44 when the volume control is after the DAC in the analog domain and the PC volume is set to the maximum. Meyer & Moran, and others, have also demonstrated that 16/44 can be fully transparent under normal conditions as a playback format when the DAC is fed with a bit accurate datastream.
There is some benefit, however, to being able to play 24/88 and 24/96 content (such as sold by HD Audio) as it's often better mastered than its 16/44 counterparts. It's not the higher bit depth or sampling rate that sounds better, it's the audiophile mastering.
EDIT: To answer the second part of your question I'm guessing your operating system is performing the sample rate conversion for you.
Last edited:
Just built my O2 and its not working properly. When I'm using the low gain setting the left channel volume is much lower than the right. When I switch to the high gain setting it works like a charm. Another weird thing is that the volume level is pretty much the same in both high and low gain with the pot untouched.
Have you applied the gain switch modification by trimming the one leg? From here:
http://www.diyaudio.com/forums/head...headphone-amp-diy-project-14.html#post2774393
The first person who had the issue reported normal gain in the high gain setting but loss of one channel in the low setting...
I think agdr nailed it here.. I saw the same thing with one of my builds.. before I started trimming the switch..
The Alps pot in the O2 is the same one in my $1600 Benchmark DAC1 and lots of other audiophile gear. It tracks to under 1 dB error down below -50 dB. That's a really quiet level (for those wondering, use the calibrated volume control in Foobar to see just how quiet -50 dB is). The tracking results are in the first O2 article.
It's likely the gain switch problem or the gain resistors are wrong.
Last edited:
fair enough, i dont see why that would make for channel imbalance if it doesnt have any imbalance at higher levels though, a mistake in the gain resistors would apply to both would it not?. maybe if the mistaken resistor was only slightly different and thus not as apparent subjectively with higher volume. or did i misunderstand and when H3nk said higher level he meant the higher gain setting but not the lower gain setting?
besides i mentioned it because people like benchmark and many other companies will buy more pots than they need and weed out ones that have imbalance, its not at all uncommon, even with some higher priced pots. measurements on yours doesnt automatically rule his out.
anyway i only mentioned it as a possibility
besides i mentioned it because people like benchmark and many other companies will buy more pots than they need and weed out ones that have imbalance, its not at all uncommon, even with some higher priced pots. measurements on yours doesnt automatically rule his out.
anyway i only mentioned it as a possibility
Last edited:
Does anybody else have a build with the default small B2-080 enclosure? There are five notches for the PCB to go into. I tried putting the board into the second-lowest spot (as opposed to the lowest), and it fits in fine. Actually, the batteries press up against the top of the enclosure, so there's some friction to overcome when sliding in and out, but I reckon that's a good thing since that hopefully means the batteries are secure inside. Also (1) this prevents the PCB isn't sliding around loose, which is relevant since I'm not using any front panel that could secure it and (2) it definitely eases concerns about leads like on the gain switch, coming into contact with the enclosure.
Any drawbacks to doing so? I notice that the PCB looks very slightly bent down in the center, so it's very slightly like a U shape. The curvature is subtle at most. I imagine that the mechanical stress is not desirable, but I'd also guess that very many PCBs in any number of applications out in the wild are regularly put under greater stress with little harm. It's not like it's being regularly flexed back and forth.
Any drawbacks to doing so? I notice that the PCB looks very slightly bent down in the center, so it's very slightly like a U shape. The curvature is subtle at most. I imagine that the mechanical stress is not desirable, but I'd also guess that very many PCBs in any number of applications out in the wild are regularly put under greater stress with little harm. It's not like it's being regularly flexed back and forth.
@mikeaj, The regulators, Q1/Q2, or the power supply capacitors don't hit the top? It's critical any metal on those has enough clearance.
I'm not crazy about the metal battery cases rubbing the enclosure either but it's probably OK. If you decide to use a front panel you would need to rework the supplied one (obviously) and if there's much of a bend to the board that might throw off some of the height placement for the panel openings.
Finally, it's probably not a big deal, but the board "likes" to have a "ground plane" (connected to the input jack ground) under it to help suppress some residual hum from the half-wave power supply. Moving that "plane" further from the board might lessen the noise reduction benefit a bit.
I'm not crazy about the metal battery cases rubbing the enclosure either but it's probably OK. If you decide to use a front panel you would need to rework the supplied one (obviously) and if there's much of a bend to the board that might throw off some of the height placement for the panel openings.
Finally, it's probably not a big deal, but the board "likes" to have a "ground plane" (connected to the input jack ground) under it to help suppress some residual hum from the half-wave power supply. Moving that "plane" further from the board might lessen the noise reduction benefit a bit.
If I wanted to point wire a 1/4" headphone jack to the O2 board, I'm under the impression that 40 AWG wire is best (20 AWG for ground) but I can't remember why. Anybody want to chime in? The reason I ask here is because I also believe I'll be adding about 1/2 Ohm output resistance with that wiring. Is that correct, and can I drop the output resister value because of it?
Thanks. I knew you were going to mention the ground plane.
None of the components like Q1/Q2 or power supply capacitors are touching the top, but I just stuck some electrical tape along the top anyway (doesn't quite seem like a permanent solution, but oh well). I think the battery only touches because the terminals are not quite aligned.
And the battery in question is the Tenergy 200mAh NiMH low self-discharge variant. It was cheaper than the 250mAh version when I was purchasing. The outer casing (not just the wrap) looks plastic to me I think.
do you realise how thin 40awg wiring is? probably about the thickness of a human hair and i dont even know where you would get such wire. 30awg is the smallest easily available (and even that not so much). you would limit the output power of the amp in a large way. this cant be a real recommendation, where did you see it?
@ethanolson & @qusp, 30 AWG wire is about 100 ohms per 1000 feet or 8 mOhm per inch. So 3 inches worth to a headphone jack and back is 0.05 ohms. This won't "limit the output power of the amp in a large way" as qusp suggests. Even with worst case 16 ohm headphones it will only cause a 3% drop in voltage. It's also 1/10th of the resistance already in series with the output of the amp. 30 AWG wire is widely available as it's the most popular gauge for wire wrap use. It's great for repairing damaged (or wrong) traces on PC boards. I agree with qusp that 40 AWG is neither generally available or practical.
What 30 AWG wire will do, if you use a single ground wire for a 3 wire jack, is raise the crosstalk into lower impedance loads. As stated in the O2 Details article, I would suggest removing a few twisted pairs from regular 24 AWG CAT5 etherenet cable and using one pair for each channel (2 ground wires). That won't degrade the crosstalk and the twisted pairs provide inherent shielding and immunity to EMI.
@H3nk glad your amp is working perfectly!
@mikeaj, if the batteries are the plastic ones, sounds like you're in good shape if you want the board higher. I agree it's an elegant way to help keep the batteries in place. You may have noticed my mention that JDS Labs plans to sell front panels and there are the laser cut acrylic ones as well. If you decide to later use one of those, you can just slide it back into the lower slot.
What 30 AWG wire will do, if you use a single ground wire for a 3 wire jack, is raise the crosstalk into lower impedance loads. As stated in the O2 Details article, I would suggest removing a few twisted pairs from regular 24 AWG CAT5 etherenet cable and using one pair for each channel (2 ground wires). That won't degrade the crosstalk and the twisted pairs provide inherent shielding and immunity to EMI.
@H3nk glad your amp is working perfectly!
@mikeaj, if the batteries are the plastic ones, sounds like you're in good shape if you want the board higher. I agree it's an elegant way to help keep the batteries in place. You may have noticed my mention that JDS Labs plans to sell front panels and there are the laser cut acrylic ones as well. If you decide to later use one of those, you can just slide it back into the lower slot.
i was suggesting the 40awg would limit the power, not the 30 and it would seem an odd choice to use, i use 30awg ribbon regularly for internal wiring (only line level, digital, or control signals admittedly and i wouldnt normally use anything less than 26-24 for headphone output, more of a practical matter). the 40awg is only rated for 13ma, though perhaps more for AC, while 30 is rated for a much more friendly 142ma
the point is moot, its the smallest size they list on the american wire gauge and ive never seen it for sale anywhere, even magnet wire doesnt come that thin anywhere ive seen (although i'm sure it exists for lab work perhaps), can you imagine stripping it if it was anything but enamel?
suppose you could still burn it off unless it was ptfe; but you might find the wire itself also burned away lol
the point is moot, its the smallest size they list on the american wire gauge and ive never seen it for sale anywhere, even magnet wire doesnt come that thin anywhere ive seen (although i'm sure it exists for lab work perhaps), can you imagine stripping it if it was anything but enamel?
suppose you could still burn it off unless it was ptfe; but you might find the wire itself also burned away lol
Last edited:
@qusp, the 30 AWG for chassis wiring is rated at 860 mA. And those are continuous RMS currents where heating is a concern which has little to do with a musical signal. The average current delivered to even current hungry headphones playing loud is under 20 mA. See: American Wire Gauge table and AWG Electrical Current Load Limits with skin depth frequencies. And even the 40 AWG would result in only 0.5 ohms of resistance which still won't "limit the output power of the amp in a large way". Sorry, but I'm just trying to be factual.
I can't dig up my source of info on the 40 AWG wire but it left a strong impression on me so I'm inclined to think that one of the well respected gurus (i.e. Pass, Self, and such ilk) had recommended it. Yeah, it's super thin but for chassis wiring it's rated to 90 mA which is fine for my intended build (max of 3.3V, <70mA).