Do you really think they may have used carbon MELF parts in the signal path?
Re. MELF/thin film at the output: the meaning of my comment was that they saved a few pence by using thin film parts in the lower legs of the voltage dividers, where the effects of their possible deficiencies on the signal are attenuated.
As regards my listening impressions of the DacMagic, they are a result of my entire setup, and therefore hardly transferable to other situations.
Regards,
Braca
You know guys, you are discussing a mid-fi product, probably designed for 'bang for the buck' with measured distortion as the determining factor in making improvements. The metal film resistors were definitely added in places where the thick film resistors used throughout the unit might effect the distortion spec. This is good 'conventional' engineering, that most here would do. However, it most probably not enough to make a really sonically successful product.
In other news, now that the subject of Max's 'additive' has been brought up, I might update Max that Jack Bybee has something new out. I have not heard it yet, but it looks similar to a conventional Bybee purifier (high current) but with the equivalent of a dipole antenna attached to each end. Who knows?
Good question, maybe the Q. After all Star Trek used quantum slipstream first.
We all know. He designs like it will be sold for much more. That is why you'll see people using his ampz for subs in million dollar systems and such - they know they are not run of the mill.
Of course if you listen to JC's or Nelson/Wayne's designs they sound significantly better than the mass marketed mid-fi stuff.
There are some simple reasons like that IEC standards that seem to have been influenced by high volume manufacturers call for minimum loudspeaker impedance to be 80% of rated impedance. Or in simply terms a mass market mid-fi audio power amplifier will sound nasty if your loudspeaker impedance drops below 6 ohms. As the AES standard for many years was 50% and many folks would put two loudspeakers on one channel most better gear can drive 2 ohms or even less. As some high end loudspeakers may drop even lower they would sound less than their best on amplifiers that could only hand loads above 6 ohms.
Then there are many other issues such as handling the ever worsening EMI environment.
On a different subject I spoke to JC about building a preamp. One of the issues is the volume control. It seems the future will be relay bank types for high end use to allow remote control and avoid limited production parts.
Obviously relay banks can get better channel matching and more precise steps. One of the issues that pops up is number of steps and step size. As everyone "knows" step size should be logarithmic. Of course cheap controls have about 45 dB of range and then a bit more sharp attenuation below that. Partly because tracking really falls apart at the low levels.
What is of interest is that human hearing (exempting some here by their claims and results) tends to compress the log scale theory above 70 dB ish and below 25 dB ish.
Step size used to bother me on some older attenuators that did 3 dB step size as everyone also "knows" that is the minimum perceptable change. Simple engineering common sense would set the step size below that number. Of course my experience is that 3 dB is not always 3 dB depending on sound pressure level.
So one set of experiments that I will do by my ears only will be on step size. A simple set up with far more taps than needed and a bit of listening to changing the volume sound give me a hint as to what I want/where.
Now by measurement my first step is to select the relays. So I ordered one of each that seemed to be designed for such use from both Mouser and Digikey. If any readers of this nonsense care to suggest particular units and where to get them I am open to suggestions. (Wayne - hint!)
Of some interest except to Scott who doesn't believe there can be discontinuities in the output voltage through a switch, most of the relays use gold plated crossbar switches. Where I think even those fall short is in the armature structure and how close the signal conductors come to the magnetic structure. So that is one of the items I will try to see if it shows up in the measurements.
Another issue is of course thermoelectric potential.
Now JC prefers the TO5 can RF relays and interestingly enough they do avoid much of the armature issues. A minor issue is that where all the other relays being tested are unt $5.00 the TO5 ones are $78.00 and only about 10 units are in stock!
My test relays are 12 volt versions with PC pins. I suspect for the fun of it I will compare Omrons G6 types with gull wings to the pinned version as I have both. Panasonic seems to have more options including one that has a maximum switch current rating of .01 amperes.
Cellphone input ted so expect typos.
There are some simple reasons like that IEC standards that seem to have been influenced by high volume manufacturers call for minimum loudspeaker impedance to be 80% of rated impedance. Or in simply terms a mass market mid-fi audio power amplifier will sound nasty if your loudspeaker impedance drops below 6 ohms. As the AES standard for many years was 50% and many folks would put two loudspeakers on one channel most better gear can drive 2 ohms or even less. As some high end loudspeakers may drop even lower they would sound less than their best on amplifiers that could only hand loads above 6 ohms.
Then there are many other issues such as handling the ever worsening EMI environment.
On a different subject I spoke to JC about building a preamp. One of the issues is the volume control. It seems the future will be relay bank types for high end use to allow remote control and avoid limited production parts.
Obviously relay banks can get better channel matching and more precise steps. One of the issues that pops up is number of steps and step size. As everyone "knows" step size should be logarithmic. Of course cheap controls have about 45 dB of range and then a bit more sharp attenuation below that. Partly because tracking really falls apart at the low levels.
What is of interest is that human hearing (exempting some here by their claims and results) tends to compress the log scale theory above 70 dB ish and below 25 dB ish.
Step size used to bother me on some older attenuators that did 3 dB step size as everyone also "knows" that is the minimum perceptable change. Simple engineering common sense would set the step size below that number. Of course my experience is that 3 dB is not always 3 dB depending on sound pressure level.
So one set of experiments that I will do by my ears only will be on step size. A simple set up with far more taps than needed and a bit of listening to changing the volume sound give me a hint as to what I want/where.
Now by measurement my first step is to select the relays. So I ordered one of each that seemed to be designed for such use from both Mouser and Digikey. If any readers of this nonsense care to suggest particular units and where to get them I am open to suggestions. (Wayne - hint!)
Of some interest except to Scott who doesn't believe there can be discontinuities in the output voltage through a switch, most of the relays use gold plated crossbar switches. Where I think even those fall short is in the armature structure and how close the signal conductors come to the magnetic structure. So that is one of the items I will try to see if it shows up in the measurements.
Another issue is of course thermoelectric potential.
Now JC prefers the TO5 can RF relays and interestingly enough they do avoid much of the armature issues. A minor issue is that where all the other relays being tested are unt $5.00 the TO5 ones are $78.00 and only about 10 units are in stock!
My test relays are 12 volt versions with PC pins. I suspect for the fun of it I will compare Omrons G6 types with gull wings to the pinned version as I have both. Panasonic seems to have more options including one that has a maximum switch current rating of .01 amperes.
Cellphone input ted so expect typos.
Last edited:
If it helps narrow things down a bit ---
I know the switching used in sota ultra low distortion analyzers use relay to do any signal or level switching. small rectangular metal can. Start with the similar style from Panasonic.
[ Update: photo of PA assembly at factory wont upload.... file size too large]
THx-RNMarsh
Last edited:
Hi John, thanks, interesting/curious.
Jack's new gizmo doesn't worry me in the slightest, my form is totally different, (coating, putty, potable, moulded, extrudable, 3D printable etc) and able to be implemented anywhere in systems and without electrical series connection.....and foremost economical/affordable and not unobtainium.
The effect at first order is to 'clarify' the system sound, but that gets boring because it exposes power, cables, system components and transducers, ruthlessly.
The second order effect is to deliberately add known 'dirt' back in....this is where the 'goove', the 'vibe', the peace and the fun happens lol.
Tone control ?....yes, but so is every cable, system component and transducer already in existence, however the system resultant is not perfectly predictable because some materials 'play nice', some don't.
Add to this signal incorporated noise and signal driven system noise dependence and the result is a crapshoot.......filtering system energy inputs and energy outputs changes everything, subjectively at least.
I might even get around to properly attempting to capture system differences soon now that I'm pretty satisfied with what I have now....this has been a long road.
Dan.
- Status
- Not open for further replies.