Yes I know that JLSounds has its own clock.
But in Jesper’s case with an 11/22 Mhz external clock I don’t think its a good idea to take things apart and use the onboard 22/24 clock if even avaible.
Hans
JL Sounds vIII uses 45/49MHz clocks. To get 22/24MHz it uses a frequency divider in the CPLD.
The earlier version of JL Sounds used 22/24MHz clocks.
You must prefer showing close-in phase noise. Why should anyone care about what you prefer?
Hans,
The electrolytic cap in series with the dac output is already lossy at RF. Did you ever measure one at high frequencies? The dissipation factor goes up quite a bit. The 3-conductor cable transmission line is nonuniform because the ground conductor (not a full sheild) is only connected at the source end. The cable length and direction (because of non-uniformity) has some effect on the sound. Since nobody else has the same cable or transformer, I suggested to experiment with whatever is on hand. When the sound is pretty close to right, FR measurements usually come out just about right too. Easy to tell if the filter is substantially off. Also, if I am listening for, say, for example, the sound of IMD on vocal harmonies, FR has very little to do with whether the vocal harmonies are garbled together to make a new sound human voices cannot make, or else the dac sounds like two distinct, mostly undistorted voices producing a harmonious beat note (which is not new frequency). In such cases, judging dac performance is not always highly dependent on perfect FR. Depends.
Moreover, I think worrying too much about such minutia as TL termination of an line level audio cable is not the most important thing right now. If instead of using the opamp filter, the passive filter gets rid of opamp-related blur, reveals more low level musical details, and sounds more like real music, then its probably better than other alternatives, at least for some initial purposes.
Last edited:
I wouldn’t call a cap in series a sort of termination.
You still have an open and reflecting node at the end of your interlink
A test with putting a 100R in series with 2n7 over the transformer is a 5 minutes job, it can’t be easier.
It made me suspicious when different cable length gave different results.
Hans
You still have an open and reflecting node at the end of your interlink
A test with putting a 100R in series with 2n7 over the transformer is a 5 minutes job, it can’t be easier.
It made me suspicious when different cable length gave different results.
Hans
A different cable length should make a difference since at audio frequencies it looks like a capacitor when terminated into a higher impedance.
Also, a cap in series can act as series termination providing the cap looks mostly resistive at several MHz.
Besides, putting a resistor and cap across the cable at the line amp end will require building two female to male RCA adapters, each with a 100R in series with 2n7 to ground. Also, the wiring from RC connector to the transformer is on the bottom of the PCB. The PCB is in a case. I would have to disassemble the thing to modify it. Definitely don't want to do that.
Also, a cap in series can act as series termination providing the cap looks mostly resistive at several MHz.
Besides, putting a resistor and cap across the cable at the line amp end will require building two female to male RCA adapters, each with a 100R in series with 2n7 to ground. Also, the wiring from RC connector to the transformer is on the bottom of the PCB. The PCB is in a case. I would have to disassemble the thing to modify it. Definitely don't want to do that.
Last edited:
I have a similar arrangement to Mark whereby I am currently (on one of my RTZ DACs) using passive filtering with series electrolytic/foil and into Lundahl LL1588 so I could give this a try easily enough later on today.
However, the "line" length on my setup though is very short, with the transformers at the location of the DAC pcb with only 4" or so of wiring between.
So maybe not a scientific test but worth investigating even so? I have to say though I am really liking the sound with the passive filtering. It's not perfect though, something has been diminished about the midrange and transients that the active filter had but then again it has gained different properties which are very pleasant. I'm sorry I am not very good as explaining what I hear but ultimately I would like to achieve my preferred characteristics of both the passive and the active modes. To that end my next experiment is separate and better power supplies for the active filter. Since I have two DACs I can use them to leapfrog changes and listen for hours during my working days.
However, the "line" length on my setup though is very short, with the transformers at the location of the DAC pcb with only 4" or so of wiring between.
So maybe not a scientific test but worth investigating even so? I have to say though I am really liking the sound with the passive filtering. It's not perfect though, something has been diminished about the midrange and transients that the active filter had but then again it has gained different properties which are very pleasant. I'm sorry I am not very good as explaining what I hear but ultimately I would like to achieve my preferred characteristics of both the passive and the active modes. To that end my next experiment is separate and better power supplies for the active filter. Since I have two DACs I can use them to leapfrog changes and listen for hours during my working days.
This would be all you need, see attachment.
No disassembly or whatever.
But its up to you, I don’t want to push, its just because you wanted to keep no stone unturned
From previous postings I thought to have understood that you used a transformer instead of Marcel’s reconstruction filter and that your line amp had also a transformer as IPS.
Now I get the impression there is only one transformer.
In that case I will have the overview diagram changed accordingly.
Hans
Attachments
There are a pair of transformers at the input to the line amp. They are essentially part of the line amp.
DAC Analog output from dac board -> DC blocking caps -> non-uniform directional cable of some length as needed -> RCA input connector on line amp -> transformer on line amp PCB -> amplifier active electronics -> ...
Also, I am okay with ordering some of those three way RCA connectors from Amazon. Could take a few day, have to see.
Can I ask the relevance of the information that the guy who dropped by is a 'Pro hi end audio designer'? It's just that in my experience there's not necessarily a correlation between 'hi end' audio products and great music reproduction.
If you are taking the balanced outputs then the large blocking caps may not be necessary going into the transformer. Maybe this will improve the SQ
Tried that with Andrea's balance dac board outputs. To some extent, the same problem as doing it with opamps. There is some loss of detail. Again, perhaps that's because of element matching issues where low level musical sounds are not exactly equal and opposite between dac output phases. Maybe that's where .01% resistors could make some difference, don't know.
As an aside, could be that using a small film cap followed by a FET buffer would sound better than a large cap and no buffer. Smaller, better caps do tend to have less SQ problems than physically large caps. We have tried that idea here too. We found that all FET buffers we tried have a sound, and it isn't necessarily better than the sound of a good combination of a burned-in electrolytic and a small film cap in parallel. That said, we haven't given up on the FET idea. May just needs more work.