Some answers:
This regulator is SPECIFICALLY designed for the folded cascode gain stage. There is NO delta I, to speak of.
Low noise is everything. Simple circuits are important too. Zeners don't work well. A true Zener would be under 4V so you would have to put 4 in series. A low noise LOW Gm jfet will have limited voltage gain into its load resistor. Only LOW Gm jfets should be used for the current source.
It would be possible to resistively degenerate the low Gm jfets, but to little avail and more complexity.
The final design runs 24 hours/day, there is no significant temp change once the unit has been turned on for awhile.
The relatively high output Z (15ohms) is trivial when the load is constant.
The low frequency gain is 10,000 for the total phono stage, so even low frequencies must be contained in amplitude, and from being generated freely.
This regulator is SPECIFICALLY designed for the folded cascode gain stage. There is NO delta I, to speak of.
Low noise is everything. Simple circuits are important too. Zeners don't work well. A true Zener would be under 4V so you would have to put 4 in series. A low noise LOW Gm jfet will have limited voltage gain into its load resistor. Only LOW Gm jfets should be used for the current source.
It would be possible to resistively degenerate the low Gm jfets, but to little avail and more complexity.
The final design runs 24 hours/day, there is no significant temp change once the unit has been turned on for awhile.
The relatively high output Z (15ohms) is trivial when the load is constant.
The low frequency gain is 10,000 for the total phono stage, so even low frequencies must be contained in amplitude, and from being generated freely.
How is jitter recorded? You record a lot of 0s and 1a with a lot of data packed round it please explain how jitter can be recorded in a digital system..... The only jitter that is critical is that at the point of conversion from digital to analogue....
Last edited:
What rubbish and why so patronising you don't know what layouts I do, we do layouts for all aspects of electronics, I am doing a design with 6oz copper, so there was no reason to have a dig or the superior attitude...
The whole size is irrelevant you don't get 2oz copper plating down the hole you get about 25microns plating on the barrel, or are you thinking you get 2oz down the barrel of the hole as well.
Reading your reply again to my comment is where your misunderstanding is, the plating I was talking about was down the PTH barrel, not board copper weight, again I suggest you READ what I wrote, we are talking about the thickness you get when plating a through hole, it is NOT the same as the laminate copper wieght
Last edited:
Richard, I was thinking about this. If you have a good DAC, with decent reclocking, the jitter in the DAC input bitstream sort of becomes irrelevant. So would a realistic comparison not be playing both the CD and the HD download over the same DAC and listen/look for any perceived differences? If there are any, it's quite unlikely its due to jitter, isn't it?
Jan
This has been the subject of many a long debate over the years on numerous threads, including claims that ripping CDs with different power supplies makes a difference, SSD vs HDD playback makes a difference etc. The main conclusion I have gleamed from the respected answers is that jitter is only critical at the point of conversion, it is how well the final part of the chain handles the jitter that is important... now as to the level of jitter audible!!!
Interestingly I subjectively perceive that playback via my HD is better than via a CD player, the trouble was the test was sighted by me trying a selection of tracks both from the streamer and putting a CD in the machine, so I presume exception bias played a big part But via a HDD is a lot easier.
Interestingly I subjectively perceive that playback via my HD is better than via a CD player, the trouble was the test was sighted by me trying a selection of tracks both from the streamer and putting a CD in the machine, so I presume exception bias played a big part
But via a HDD is a lot easier.
Last edited:
This has been the subject of many a long debate over the years on numerous threads, including claims that ripping CDs with different power supplies makes a difference, SSD vs HDD playback makes a difference etc. The main conclusion I have gleamed from the respected answers is that jitter is only critical at the point of conversion, it is how well the final part of the chain handles the jitter that is important... now as to the level of jitter audible!!!
Interestingly I subjectively perceive that playback via my HD is better than via a CD player, the trouble was the test was sighted by me trying a selection of tracks both from the streamer and putting a CD in the machine, so I presume exception bias played a big part
Yes indeed, I went back and read the refs Richard posted a link to. The 2nd one, Bob Katz, after discussion the various ways a signal can pick up jitter during processing and transferring, he says that the in the final step it really doesn't matter because it's the DAC that determines the jitter during final conversion. This is a nice confirmation.
The first ref is completely wrong. The decision whether a bump/pit on a CD(R) is a one or zero depends on the pit/bump length. IIRC, the ratio of a zero to one bump/pit lenght is 1:3 to provide for some robustness in length variation. So the point he makes about the 'sharpness' of the bump/pit edges and relation to jitter is complete nonsense, he does not know what he is talking about.
Jan
What rubbish and why so patronising you don't know what layouts I do, we do layouts for all aspects of electronics, I am doing a design with 6oz copper, so there was no reason to have a dig or the superior attitude...
The whole size is irrelevant you don't get 2oz copper plating down the hole you get about 25microns plating on the barrel, or are you thinking you get 2oz down the barrel of the hole as well.
Reading your reply again to my comment is where your misunderstanding is, the plating I was talking about was down the PTH barrel, not board copper weight, again I suggest you READ what I wrote, we are talking about the thickness you get when plating a through hole, it is NOT the same as the laminate copper wieght
Wow
What attitude. You mentioned that the plating was thicker then the traces which is not the case in the sample used for the distortion tests.
You have also described often the complex boards you layout. Here the one under consideration is extremely basic and simple. Not what I would waste a talented designer on.
BTY I thought we weren't communicating from the start when you seemed to get the idea that I supported those folks who hear differences from solder type.
ES
Looked it up and my understanding was not correct: it is the change from pit to land and vv that determines a ONE, while an absense of such changes in a specific time frame denotes a ZERO. But at any rate, the edge sharpness has no relation to jitter developed later in the decoding and clocking process.
Jan
Jan
Yes, no excess noise, and low inductance, in a position where a little inductance wouldn't hurt anything (the more extreme case being a JE-990).
What's the gm of a J175? I selected a J271 for the right Idss to get the correct output voltage, but I never thought of either of those devises as low gm. Maybe a 2N5461?
edit: Never mind the last question, I see the original was a J203. That makes sense.
Let me get this clear Ed. You want me to do some work proving that Bob Katz, who agrees with me, ;-) is wrong? (Go ahead, think that one through).
Edit: I no longer have the AP, I demoted myself to a dScope series III...
Jan
Do you guys actually listen to what you are saying? Archived data with error checking is either all right or wrong. You can not "record" jitter, it adds no information to the signal. Designing a final playback chain to minimize it at your end is the best you can do.
EDIT - I see others have already said this, the jist of my point is that "recording" jitter violates first principles of information theory so it is an extraordinary claim.
Last edited:
It is only opinion without data. No matter how clear you perceive it to be. Actually it is these kinds of tests that either confirm expectations or show something else ugly is really there. You'll never know for sure unless you actually try it.
For my next trick I may just show why high PSRR is nonsense!
Two things have been bothering me with the 'simple MC' circuit with load resistors. Firstly without an input blocking cap how much do you have to worry about keeping the input gate at 0V? This is more a how anal with parts matching question vs letting some silicon do it automagically.
Secondly as we are talking about ultra-low noise low output MCs here (not forgetting the vendetta was one of about 3 phono stages that could handle the MC2000 without a transformer) why is the input not balanced? As output from an MC goes down hum pickup becomes a biggy.
Secondly as we are talking about ultra-low noise low output MCs here (not forgetting the vendetta was one of about 3 phono stages that could handle the MC2000 without a transformer) why is the input not balanced? As output from an MC goes down hum pickup becomes a biggy.
Installed base. The vast majority of audiophiles do not have their tonearms and cables set up for balanced. I use balanced exclusively for phono, but am fully aware that I can't plug my stuff into other people's systems and vice versa.
Since there's essentially no gate current, the input is automatically at zero volts DC because of the load resistor.
Since there's essentially no gate current, the input is automatically at zero volts DC because of the load resistor.
Sorry I was blunt in my reply, I read your intention wrongly....
I said that the width of a via when converted to a equivalent trace width using the circumference is often wider than the trace. When you said the holes were not solder filled I presumed via's as other holes have a pin and solder in them usually.
For there to be distortion due to a PTH changing layer is strange and if so there must be some definable reason as to why it is so.
- Status
- Not open for further replies.