[Sean H]

Quote:

Originally Posted by Ori Because of this structure, the first opamp is the only one configured as a low-pass filter. The middle opamp is a summing amplifier. All these stages run a very low gain, because the signal(s) out of either DAC chip are line levels. When you remove the opamps, the passive feedback components (which make the opamp circuit) form a path for the audio signals to the outputs. However, two things happen. The gain drops a tiny bit, which is not a big deal, but the output impedance goes up. This could affect some setups. More important, removing the two opamps affects the frequency response of the circuit, which may no longer be what it was meant to be.

Ori - Thanks for the info. So basically other than the impedance alteration and the possibility of a frequency response alteration, there's no risk of damage to the circuit or to your amp/cans/speakers if one removes the two XLR opamps but leaving in the summing opamp?

 

[HumanMedia]

Quote:

Originally Posted by Jon L So you only changed the connectors and kept the cable itself the same? Which RCA connector did you have before, and can you give us a link to the BNC connectors you used? Thanks.

No, I got a new premade cable which was proper 75ohm coax with proper 75ohm BNCs. It was only $10 but has reasonable quality components. I will replace it with a shorter one with better components from BluejeansCable.com, which will still only be $12.50 (for a 2 footer). Maybe down the track I will try a VHAudio Pulsar with Furutech BNC, but I dont know how much improvement that will give (for 10x the price).

There were only three main contenders for proper 75 ohm chassis/bulkhead connectors. I will post details when I get home tonight. Note that there are many such connectors out there but the majority are 50ohm (for computer networking) and not appropriate. Make sure you get isolated ones, which isolate ground from the chassis, usually with a plastic sleeve (the Canare ones require a separate sleeve part to be ordered to be used with it, but the others come with the isolation sleeve).

From memory the manufacturers who supply them are: Canare, Amphenol, Neutrik. I grabbed some Neutriks because I could get them locally for $1 each and are still great quality (and have isolaters)

Also you will want to get the chassis connectors with solder on rear interfaces. Which raises another issue - what do you use to connect the back of the chassis BNC to the DAC PCB? To maintain the impedance as closely as possible I bought a small length of the same 75 ohm coax that my interconnect was made of, and pulled out the centre conductor and braided shield and separated them into very small (1cm) length cables and soldered these from the PCB to the back of the BNC. I even kept a very small length of dialectric on the centre conductor to maintain its characteristics, despite that the shield was no longer wrapped around it coaxially for this last small connection.

 

[Zowbombs]

I have to chime in here. I've owned the Zhaolu for around 3 months now, and I do feel it is all it's hyped up to be.
One thing though, I can't hear a difference between optical and coax inputs, nor did I notice any difference when I swapped opamps around. How noticable are these improvements supposed to be?

 

[Ori]

Quote:

Originally Posted by Sean H Ori - Thanks for the info. So basically other than the impedance alteration and the possibility of a frequency response alteration, there's no risk of damage to the circuit or to your amp/cans/speakers if one removes the two XLR opamps but leaving in the summing opamp?

That's correct. Just to be safe, I'd still measure the offset on the outputs, just in case there is a major assymetry in the circuit.

 

[HumanMedia]

Oh and thanks to yo2tup2 who probably was the first with the 75 ohm BNC tip.

Note that there is a way of getting close to 75 ohm connections and staying with RCA connections. You will need an 75 ohm coax interconnect cable with either Eichmann bullets or WBT Nextgens at each end. Then on your chassis you will need to replace RCA with a WBT Nextgen RCA.

Note that these are much closer to 75ohm, far better than any other RCA (including the Canare RCAs - which are definitely NOT 75 ohm), but still not as good as real 75ohm BNC. The effort in replacing the chassis RCA to WBT Nextgens is the same as installing proper BNCs, so I cant see any reason for going this route with RCA's.

 

[HumanMedia]

And if you are ordering or making a 75ohm coax interconnect with 75ohm connections make sure to keep it less than 0.5m in length or greater than 1.5m. So if you get a short interconnect from BlueJeans you will need to add a note to your order to make the interconnect 1.5ft instead of 2ft.

Why not (length > 0.5m) AND (length < 1.5m)?

I believe this is the length that is optimal for picking up wavelengths of Radio Frequency transmissions (radio, TV?) which for this application is just interference.

 

[HumanMedia]

Quote:

Originally Posted by Zowbombs I have to chime in here. I've owned the Zhaolu for around 3 months now, and I do feel it is all it's hyped up to be. One thing though, I can't hear a difference between optical and coax inputs, nor did I notice any difference when I swapped opamps around. How noticable are these improvements supposed to be?

I found the difference between optical and RCA coax to be '6 of one, half a dozen of the other'. On some material optical sounded better, on some material coax sounded better. Upgrading the coax from a Monster Video cable to a lowly BlueJeans RCA coax nudged the RCA Coax ahead of the optical (for me). Upgrading to proper 75ohm BNC Coax interconnect and connectors throoughout was a quite noticeable step up, and I am informed that upgrading the $12.95 Bluejeans BNC coax interconnect to something better will improve thinsg further.

Opamps.
Some sound similar (2107 and 2604) and some sound quite different (AD823 DY2000). Note that to evaluate the sound of one type of opamp I populate all three slots with that flavor and leave it at least a week. Opamps seem to need 100+ hours of burn in. This was reinforced again recently when I started trialling 3 x 2107 which started quite 'milky' then took almost 2 weeks to blossom and clear up. Others like the DY2000 start congested and rough in the treble then tame down a expand in the midrange.

After getting to know one flavor you can then try mixing combinations.

 

[Lad27]

Quote:

Originally Posted by HumanMedia Opamps seem to need 100+ hours of burn in. This was reinforced again recently when I started trialling 3 x 2107 which started quite 'milky' then took almost 2 weeks to blossom and clear up. Others like the DY2000 start congested and rough in the treble then tame down a expand in the midrange.

There is nothing mechanical moving in inside of an opamp. It is a semiconductor based device and at given and stabilized operational temperature and properly stabilized power supply voltage it will operate from hour 1 and after 100 hours later with the same electrical performance.
Perhaps that the difference you hear after 100 hours are contributed to your ears being conditioned to the new sound?

 

[Filburt]

Quote:

Originally Posted by Lad27 There is nothing mechanical moving in inside of an opamp. It is a semiconductor based device and at given and stabilized operational temperature and properly stabilized power supply voltage it will operate from hour 1 and after 100 hours later with the same electrical performance. Perhaps that the difference you hear after 100 hours are contributed to your ears being conditioned to the new sound?

Yep. They're specifically designed not to drift over time like this, especially over a negligible period like 100 hours. In semiconductor terms, that's _nothing_. The OPA2107 is a precision op-amp designed to be used in applications where the instrument may fail to function properly if it were behaving like this. I can just imagine what sort of flack TI would get from people complaining about having to calibrate their instruments over and over again due to such rapid drift. I don't know what you guys are doing to your op-amps, but mine aren't drifting like that over 100 hours...measurably or subjectively.

 

[HumanMedia]

Quote:

Originally Posted by Lad27 There is nothing mechanical moving in inside of an opamp. It is a semiconductor based device and at given and stabilized operational temperature and properly stabilized power supply voltage it will operate from hour 1 and after 100 hours later with the same electrical performance. Perhaps that the difference you hear after 100 hours are contributed to your ears being conditioned to the new sound?

Just like cables eh?

Yet they change over time also. Im sure gross electrical characteristics are relatively static but I have found repeatedly that cables and opamps change in more subtle ways like spatiality, which Im sure is due to some very tiny timing issues which stabilise over time (perhaps due to a mechanical stress settling?*), and is one thing our ears and brains are very accurate in picking up.

Try having someone walk through a pitch black room in front of you. You will 'hear' the reflective blind spot of them moving in front of you through incredibly small timing and doppler changes in reflected sound. Try sticking a microphone up and looking at the recorded sound waves as data - you probably wont see a thing in the gross measurements, but it doesnt mean its not there.


There is a large 'brain getting used to it' component to 'burn-in', but when I burn-in stuff on the DAC, I do it in the background, and listen to music via the CDP analog outs not through the DAC route. I occasionaly switch over to see how its going and the steps in change are more grossly noticeable and there is extremely little 'getting used to it' involved.




*Is this how cryo treatments settle mechanical stresses, which in turn affect low level electrical stability and the resultant noise levels and coherency of the sound?

 

[Lad27]

HumanMedia,

Fully agree that our brains and ears are very sensitive indeed.

As for cables - you'd agree we can easily fill another 20 pages of this thread with arguments and counter arguments in no time. I do not personally believe in cable burn-in.
Back to opamps. In this particular case (analog amplifier) there is no timing in internal circuit involved at all.
All I can say that perhaps some people can hear a change which cannot be measured even with high precion instruments. You may detect a slight hint of scepticism here, but I'm afraid my 25+ years of experience in test engineering field kicks in a little.

 

[HumanMedia]

Quote:

Originally Posted by Lad27 HumanMedia, Fully agree that our brains and ears are very sensitive indeed. As for cables - you'd agree we can easily fill another 20 pages of this thread with arguments and counter arguments in no time. I do not personally believe in cable burn-in. Back to opamps. In this particular case (analog amplifier) there is no timing in internal circuit involved at all. All I can say that perhaps some people can hear a change which cannot be measured even with high precion instruments. You may detect a slight hint of scepticism here, but I'm afraid my 25+ years of experience in test engineering field kicks in a little.

I dont want to get off track here, nor challenge anyones engineering, scientific knowledge or beliefs, nor crack open the old DBT chestnut. I just want to say that I think high precision instruments CAN pick up the characterictsics of what people hear, but it requires looking at the data in very specific ways. Stuff like the example above can be detected by instruments if one feeds the data into signal processing software setup to look for exactly those sort of changes. And not just in one channels signal but how it relates to the other channels signal. Incredibly small measureable changes that might be below the limit of human perception in one channel can become grossly obvious when the listener is given a second channel where the change is proportionally linked between the channels. Yet a gross interpretation of just the data may not disclose anything obvious even when two channels are looked at, without more sophisticated and specific analysis of the data (and in this case, between both channels). Thats just one example, and if the analysts dont know exactly what to look for in the data, and what analysis to apply, then the more simplistic high level interpretations of the data may not indicate anything, even though it is there.

Anyway, take my opinions or leave em. Before I played around with lots of opamps I didnt think they could burn-in either. And I went through 6 years of denial that power cords could make any difference, but I can now demonstrate in 30 seconds with two cords that they do.

Anyway moving along with another tweak from the lunatic fringe...

 

[HumanMedia]

Isolation feet for the Zhaolu.

The 1.3 ships with foam feet. Ive tried a series of different feet on the Zhaolu including cheap isolation cones, generic rubber feet from the hardware store, sorbothan feet, herbie baby booties, sponge rubber things etc etc.

Playing different types of music, one can get a 'sense' of what the different feet can do, but audio memory is malleable. And getting a sense takes some time. Here is what I use to differentiate the effects very quickly.

Ray Brown. The double bassist. Plays lots of bass scales up and down. Just playing a single track I can get a good idea of how different feet work with the Zhaolu 1.3, as certain notes are accentuated with different feet.

Cones tend to generally dampen the bass, the sorbothane and spongly rubber tend to create a lower bass bloat. The best of the lot were the generic rubber feet from the hardware store and even better were the Herbies baby booties (USD$2 for 4 and $10 for 4 respectively). These seemed to avoid the bass bloat, yet still warmed up the bass without any accentuauting any particular bass notes.

Im getting deja vu typing this as I mentioned it before in the massive 'DAC' from china' thread. Another dampening trick mentioned in the other thread is to drill a hole through the middle of the unit (between the pcbs inside of course) and use a metal screw with rubber washers to tighten both the upper and lower chassis. I havent done my Ray brown test with and without the screw - but it does makes the chassis far less clangy and resonant.

 

[Zenja]

I took out the two side opamps and couldn't really detect an immediate change in sound, but I suck at that sort of thing. I tried going back and forth a whole bunch of times but it was too noisy in the room to detect changes. I was mostly just listening for frequency response changes, as Ori said there might be, and I didn't detect any. I'm listening through the built in descrete amp.

I'll leave the opamps out for a while and see if I can find anything out. One definite change is that I have to turn the volume pot up a lot more for equivalent volume, which is a big positive for me. Normally I can only listen at like 10% volume.

Also, I noted that the input caps on the descrete amp are Bennic polypropelene ones. Considering they're not just crappy electrolytic caps, does it make sense to leave the tiny caps on the underside of the board? I still don't really trust that putting those small caps in parallel with these polypropelene ones will improve sound. I guess I should try to find some high quality polypropelene caps and just put those in place of the pairs that are in there right now.

 

[uzziah]

i'm sorry but i cannot imagine that putting different rubber feet on your DAC will "dampen the bass". seems pretty ridiculous to me. there's no moving parts. the freakin rubber feet are just so it doesn't slide off your lido deck

or scratch your marble countertop. imho, people project changes they expect from an external item they add to the equation: "oh, your dac is to "sharp", well, sand down the edges". or, "the sound is too soft", well, put it on a concrete floor.

all joking aside i find this sort of talk off the deep end