Hi Sebastian,
It interfaces with the PC via USB, but all the signal processing is done internally and they just have a proprietary driver and program that interfaces with it. There's no way to use it with any other measurement program that I know of.
What I could potentially do is use a really good DAC and a really good ADC and interface them to the PC via SPDIF. That would at least get the performance up to level where you might be able to see something. I'll look into it and let you know.
That's definitely putting it a lot more eloquently than I did
. We should probably go with that.
My wording was a little strong, and just to be clear, I've been on that end of things several times. Way back when, I had a nice little NAD integrated amp that I was using to power a subwoofer. During a comparison, I was swapping cables, with the amplifier on, and turned pretty much all the way up with music running. The cables had giant gold plated banana plugs on them and while I was moving the cable, it snagged and I dropped the end. Needless to say, as it hit the floor it let off a pretty epic spark and the amplifier went dead. After a good facepalm, I pulled the amplifier apart to find that it was just a fuse on the output that blew, and not the entire output stage. I replaced the fuse, and I was good to go again!
I was indeed thankful that someone had designed in those fuses, but I was also quite aware that what I was doing was incredibly stupid. I did learn my lesson though, and it has never happened again since.
As for the target audience, I think everyone here will be fine building and using The Wire. The low gain and relatively low AC supply rails make it pretty safe overall. I don't turn mine off to swap inputs or headphones, and I've been dragging it around with me as a bare board for the past year, so it doesn't even have the convenience of a good case and proper connectors!
Cheers,
Owen
It interfaces with the PC via USB, but all the signal processing is done internally and they just have a proprietary driver and program that interfaces with it. There's no way to use it with any other measurement program that I know of.
What I could potentially do is use a really good DAC and a really good ADC and interface them to the PC via SPDIF. That would at least get the performance up to level where you might be able to see something. I'll look into it and let you know.
That's definitely putting it a lot more eloquently than I did
. We should probably go with that.My wording was a little strong, and just to be clear, I've been on that end of things several times. Way back when, I had a nice little NAD integrated amp that I was using to power a subwoofer. During a comparison, I was swapping cables, with the amplifier on, and turned pretty much all the way up with music running. The cables had giant gold plated banana plugs on them and while I was moving the cable, it snagged and I dropped the end. Needless to say, as it hit the floor it let off a pretty epic spark and the amplifier went dead. After a good facepalm, I pulled the amplifier apart to find that it was just a fuse on the output that blew, and not the entire output stage. I replaced the fuse, and I was good to go again!
I was indeed thankful that someone had designed in those fuses, but I was also quite aware that what I was doing was incredibly stupid. I did learn my lesson though, and it has never happened again since.
As for the target audience, I think everyone here will be fine building and using The Wire. The low gain and relatively low AC supply rails make it pretty safe overall. I don't turn mine off to swap inputs or headphones, and I've been dragging it around with me as a bare board for the past year, so it doesn't even have the convenience of a good case and proper connectors!
Cheers,
Owen
Thanks for the detailed reply.
I see. I thought those newer AP units had digital AUX i/o, don't know where that misconception came from.
Only if you happen to have those handy. Please no extra effort if it's just me requesting the data.
It's not that I have doubt about the performance. I was just asking for the fun of it.
No offense taken, you merely 'spoke' out what some may think.
Unfortunately I know myself regarding those things. I'm an idiot at times myself, so assurance about safety doesn't hurt.
I'd like to make clear, though, that from my perspective I can confirm that nothing in the design of The Wire poses a particular threat to headphones or other equipment.
You actually managed to make it extra easy to play safe by also offering the matching power supply. This one certainly doesn't brown out the ugly way or suddenly drops one rail while the other stays stiff.
In other words: the way you offer The Wire it should be as safe as a headphone amp gets – and those DIYers who mess with your circuit by deviating from your BOM or recommended application better know what they're doing.
Seriously, it can't hurt to supply a disclaimer with the shipments...
Cheers,
Sebastian.
PS: Is there some curcuit property actually preventing thumping in this application (seeing that OPA1632 and LME49600 operate from about +/-2.5V and LME49990 from +/-5V, thus pretty much right from the start)?
I see. I thought those newer AP units had digital AUX i/o, don't know where that misconception came from.
Only if you happen to have those handy. Please no extra effort if it's just me requesting the data.
It's not that I have doubt about the performance. I was just asking for the fun of it.
No offense taken, you merely 'spoke' out what some may think.
Unfortunately I know myself regarding those things. I'm an idiot at times myself, so assurance about safety doesn't hurt.
I'd like to make clear, though, that from my perspective I can confirm that nothing in the design of The Wire poses a particular threat to headphones or other equipment.
You actually managed to make it extra easy to play safe by also offering the matching power supply. This one certainly doesn't brown out the ugly way or suddenly drops one rail while the other stays stiff.
In other words: the way you offer The Wire it should be as safe as a headphone amp gets – and those DIYers who mess with your circuit by deviating from your BOM or recommended application better know what they're doing.
Seriously, it can't hurt to supply a disclaimer with the shipments...
Cheers,
Sebastian.
PS: Is there some curcuit property actually preventing thumping in this application (seeing that OPA1632 and LME49600 operate from about +/-2.5V and LME49990 from +/-5V, thus pretty much right from the start)?
I have to confess I've been busy with other things and not following this thread. Agdr pointed out the above and wondered if I would be interested. The answer is I possibly would. Obviously I would love to have the O2's performance verified by someone else on an Audio Precision analyzer. That's even more true if they're free of commercial interests.
It's likely that Tyll Hertsens at InnerFidelity will someday test an O2 on his Audio Precision along with a subjective review. But I don't know for sure when that will happen. And, as much respect as I have for Tyll, a variety of commercial headphone manufactures are ultimately responsible for his paycheck. The commercial influence puts him in a somewhat awkward position reviewing an open source not-for-profit design like the O2 (or The Wire) when there are banner ads on InnerFidelity pushing $1000+ headphone amps that probably don't measure as well.
The main question I have: Is there really much interest in having The Wire tested by someone else on professional equipment? From my perspective Owen seems to have done a credible job and, like me, isn't trying to profit from his design which removes the largest typical source of bias and cause for suspicion in measurements. If there is enough interest, a swap might be worthwhile.
At the risk of sounding overly active in this regard: I am particularly interested.
The reason is simply that I know a small couple of people in the professional studio and recording field who are particularly interested in a supreme quality headphone without bells and whistles. These guys demand more than budget equipment can deliver but don't trust the voodoo behind premium manufacturers.
I'd be more than happy to prove to them how enthusiastic people and some serious engineering can easily outperform brand names. The more measurement results to support this the better.
Besides, seeing O2 and The Wire coexist in peace – supporting each other – would make me happy too...
PS: Please not only the SE-SE for comparison, at least the BAL-SE (if not the BAL-BAL, too) should be considered due to it's relatively sophisticated approach.
The reason is simply that I know a small couple of people in the professional studio and recording field who are particularly interested in a supreme quality headphone without bells and whistles. These guys demand more than budget equipment can deliver but don't trust the voodoo behind premium manufacturers.
I'd be more than happy to prove to them how enthusiastic people and some serious engineering can easily outperform brand names. The more measurement results to support this the better.
Besides, seeing O2 and The Wire coexist in peace – supporting each other – would make me happy too...
PS: Please not only the SE-SE for comparison, at least the BAL-SE (if not the BAL-BAL, too) should be considered due to it's relatively sophisticated approach.
Last edited:
I'm not up on the differences between the SE and Balanced Wire. But it's worth pointing out Single Ended designs, all else being equal, usually outperform Balanced designs. Doug Self has written extensively on the topic. Balanced inputs and outputs add more distortion producing active components, have inherently more noise, and contrary to myth, only the less objectionable even distortion harmonics are partially cancelled in a good balanced stage. The more objectionable odd distortion harmonics tend to be additive in balanced stages.
Balanced audio's main benefit is reduction of common mode noise in cables. But in typical home headphone audio, that's rarely a problem even with single ended connections.
That's a long way of saying the SE should test better than the Balanced amp (without me knowing more about the differences).
The balanced Wire isn't simply a bridged amp. It's a topology (partially) leveraging super symmetry (courtesy of Nelson Pass, applied by Texas Instruments) and contains the same number of devices (per output) as the single ended Wire: one opamp and one current buffer (per output). The OPA1632 isn't to be underestimated, performance wise.
In case of the usual three opamp input stage, this would be the case. A typical comparison between single ended and balanced circuitry (like the one Self made) usually simplifies by assuming that all active circuitry (and the number of accompanying passives) is doubled. This does partially apply to the BAL-SE Wire. But the BAL-BAL Wire employs a fully differential opamp which is not significantly more complex than a single opamp with a second output stage (as I'm sure you know), see p. 13, fig. 29.
Again, that applies to the two topologies assuming the same type of components and their number essentially doubled.
Consider the inherently doubled voltage gain of a fully balanced stage. Bringing this down to the same gain as the SE stage improves noise figures by 6dB, assuming the inherent noise in the employed opamps is equal. With the fully balanced Wire only the output buffer is configured as a bridged stage, the input and error correction stage still has the same number of device(s).
In other words, with the noise increased and the gain increased as well, the dynamic range probably doesn't change for the worse if done right.
Haven't heard of that myth.
As far as textbooks (and I've read Self's amplifier book) are concerned, complementary stages usually only cancel even order harmonics (as the distortion mechanisms leading to odd order harmonics aren't symmetric in their nature).
I agree that a change in sound signature could be in effect here, but with a total distortion level (referred to full scale) possibly still lower than in the SE case, as adding up the harmonics (geometrically) still shouldn't eat up the 6dB (linearly) higher dynamic range inherent to the topology.
I don't think any kind of balanced circuitry is for the typical home user.
That being said I am thankful Owen considered the BAL-BAL case in his project. It would be a shame to not acknowledge the outcome now.
On a personal note I might add that I like to make it a point to propose balanced transmission lines throughout, whenever possible, i.e. from the pickup to the ADC and from the DAC to the speaker or headphone transducer. Not because it's required in all cases, but because it's better and not really more expensive (in DIY). But that's another topic...
Again, if accounting for their gain difference and measuring from a balanced source into a balanced input, I'm not so sure.
Properly considering dynamic range is key here, though, as the (absulute) noise voltage could indeed be slightly higher.
Actually, this kind of comparison is exactly what I would be interested in...
But I'm only one guy, though.
Last edited:
@Sek, thanks for filling in some of the details. I know many here on diyAudio are more informed than most, but I still find a majority believe Balanced is automatically Better with little regard for the details. Many are aware Balanced audio offers distortion cancellation but fail to realize it's almost never of any audible benefit as it doesn't apply to the most objectionable distortion.
If the Balanced Wire uses a completely different approach, I agree it would be useful to compare the performance of both versions. If Owen has already done that, I could test whichever version is more popular (I suspect the SE?). If the two versions have not already been measured and compared, Owen is in the best position to do so.
If the Balanced Wire uses a completely different approach, I agree it would be useful to compare the performance of both versions. If Owen has already done that, I could test whichever version is more popular (I suspect the SE?). If the two versions have not already been measured and compared, Owen is in the best position to do so.
Last edited:
nope, depends on the load i guess, but given the recommended dual psu for the BAL-BAL i would expect with double the load would come double the swing and thus 4 x the power. actually the dual psu shouldnt make much difference here
Hi Guys,
PM's have gone out to several people who had PCB only orders for payments. Boards will ship out on Monday!
There are a few important updates:
1 - The GB is completely closed with the exception of BAL-SE PCB only orders. I'm placing the parts orders today, so no more kits.
2 - There was a serious mistake on the SE-SE PCB which means I need to re-order that board. If you look closely at the schematic I posted a while back you'll notice that the supplies for U9 and U11 are backwards. It was a really stupid mistake, but to be honest with 5 different boards it was the only thing wrong so I'm not going to get worked up about it. I'll be ordering a few panels with just SE-SE boards and PSU boards since the PSU oversold by about 25 pieces. I'll have them in a week.
The above means I now have 50 SE-SE boards that will need rework to function. It involves cutting 4 traces with an exacto knife, and soldering 4 jumpers to nearby vias. It's a little tricky, but not a big deal, so if anyone wants one of the messed up boards, I'll liquidate them at $2 each to help re-coupe some of the costs for the new boards. This offer is only open to those who have existing orders for The Wire (power amp or headphone amp).
I want to make it clear that the new boards I'm ordering this weekend will be perfect, so anyone with an existing SE-SE order will get a proper board.
3. I've got all the measurements done and I'm working diligently to get them put up in a wiki. I re-did all the crosstalk measurments and I'll clarify them in the wiki. The ones I posted for the SE-SE weren't that far off, but they're a little inconsistent.
4. I've attached the final updated list with all orders. Please check that your order is correct.
Cheers,
Owen
PM's have gone out to several people who had PCB only orders for payments. Boards will ship out on Monday!
There are a few important updates:
1 - The GB is completely closed with the exception of BAL-SE PCB only orders. I'm placing the parts orders today, so no more kits.
2 - There was a serious mistake on the SE-SE PCB which means I need to re-order that board. If you look closely at the schematic I posted a while back you'll notice that the supplies for U9 and U11 are backwards
. It was a really stupid mistake, but to be honest with 5 different boards it was the only thing wrong so I'm not going to get worked up about it. I'll be ordering a few panels with just SE-SE boards and PSU boards since the PSU oversold by about 25 pieces. I'll have them in a week.The above means I now have 50 SE-SE boards that will need rework to function. It involves cutting 4 traces with an exacto knife, and soldering 4 jumpers to nearby vias. It's a little tricky, but not a big deal, so if anyone wants one of the messed up boards, I'll liquidate them at $2 each to help re-coupe some of the costs for the new boards. This offer is only open to those who have existing orders for The Wire (power amp or headphone amp).
I want to make it clear that the new boards I'm ordering this weekend will be perfect, so anyone with an existing SE-SE order will get a proper board.
3. I've got all the measurements done and I'm working diligently to get them put up in a wiki. I re-did all the crosstalk measurments and I'll clarify them in the wiki. The ones I posted for the SE-SE weren't that far off, but they're a little inconsistent.
4. I've attached the final updated list with all orders. Please check that your order is correct.
Cheers,
Owen
Hi Owen,
many thanks for all your time & efforts you put on this.
My orders are listed correctly in the summarizing .pdf-file.
But if possible, I'd like to add a BAL-SE board. I am happy that you haven't already closed this GB.
Edit: You can also sign me up for one of the erroneous SE-SE boards.
TIA again.
many thanks for all your time & efforts you put on this.
My orders are listed correctly in the summarizing .pdf-file.
But if possible, I'd like to add a BAL-SE board. I am happy that you haven't already closed this GB.
Edit: You can also sign me up for one of the erroneous SE-SE boards.
TIA again.
Last edited: