Ken, I also don't know if it is the pixies that live in my CD player that are contributing in some way. If you can present an analysis, then please do so.
Lets, for the sake of discussion, pretend this mechanism exists and is a problem. Is the solution to reduce the noise from error correction, servos, power amps and the neighbours washing machine or is it to isolate and reclock?
I do understand it is difficult to measure jitter in the pico second range, we only measure disc jitter to the closest nano second.
There are other ways to do this. Try to push the limits of the system. Modify the system to test the hypothesis. If you want to test the error correction, add defects and test. Repeat this until it fails. Is it audible? My experience is that it is not, until it fails and then it is very audible.
If you want to test clock jitter then artificiality add the ground noise. You could even try at what frequency/amplitude it fails or becomes audible.
Stabilizing the plater will also have no audible effect. How can it? But that could also be tested.
Lets, for the sake of discussion, pretend this mechanism exists and is a problem. Is the solution to reduce the noise from error correction, servos, power amps and the neighbours washing machine or is it to isolate and reclock?
I do understand it is difficult to measure jitter in the pico second range, we only measure disc jitter to the closest nano second.
There are other ways to do this. Try to push the limits of the system. Modify the system to test the hypothesis. If you want to test the error correction, add defects and test. Repeat this until it fails. Is it audible? My experience is that it is not, until it fails and then it is very audible.
If you want to test clock jitter then artificiality add the ground noise. You could even try at what frequency/amplitude it fails or becomes audible.
Stabilizing the plater will also have no audible effect. How can it? But that could also be tested.
True for all three - and another example how companies cut corners, even in the CD-Pro. In first generation players, the cone was on a spring. It centered the CD automatically, What only mattered was tight clearance between cone and shaft.
Today, the cone is fixed on the platter (or platter and cone are one part) and therefor the cone has to give more clearance so the mechanism can easily eject/insert a CD,
Hi Mark,
The problems can and would be caused by unstable supply voltages if they were allowed to exist. Part of separating the power supplies includes keeping the ground currents out of other circuits. Even the momentary shorts in motors (feed and disc) cause huge current spikes, easily seen using an oscilloscope. Those spikes and ground currents could very easily cause deterministic jitter (now that's bad!). The periods would be related to the cycle of the disc spinning.
-Chris
The effects have been designed out on the mid and better machines. With the cheap ones, I imagine they suffer from all kinds of layout issues and design (schematic) issues. Untested except from casual observations over the years. I have to admit I didn't see much point in examining really cheap designs for errors and trying to chase down what problem was caused by what defect. Take your pick in those.
The problems can and would be caused by unstable supply voltages if they were allowed to exist. Part of separating the power supplies includes keeping the ground currents out of other circuits. Even the momentary shorts in motors (feed and disc) cause huge current spikes, easily seen using an oscilloscope. Those spikes and ground currents could very easily cause deterministic jitter (now that's bad!). The periods would be related to the cycle of the disc spinning.
-Chris
Hi Jay,
(2) I'm not certain what the length of the shaft has to do with anything realistically. Most of my good current machines have the longest shaft length that I can not get for some reason. It would make a difference if the disc was punched off-center, or the table was not true.
(3) We had one heck of a time with the first machines and CDs getting stuck on the shaft. Everything to holes too small, flashing still on the disc (causes scratches in automotive transports that use the soft rollers). Due to poor tolerances with many CDs, the centering cone has been forced to be slightly smaller.
When a CD player skips on some discs, the customer doesn't care about the why. They just want the thing to play. At least with the more expensive machines we could ask the customer why he is feeding cheap discs into an expensive machine. That was a valid observation actually. A concept that people spending lots of cash seem to understand. Try telling that to the bargain hunter with their $30 find on sale. You've lost that battle as soon as they lay their money down. Years of observation.
We are being ripped off in other words.
You should have seen the results of a laser disc coming apart in the player - at speed. It was dangerous to the technician! I have heard of, but never witnessed a DVD coming apart after de-lamination of the two halves. I suspect that the vibration is enough to release the DVD or laserdisc from the clamp. It all comes apart once it contacts the support structures for the clamp, or flapper. Laserdisc failure can bend the top cover outwards from internal contacts. Lots of energy there. With DVDs, not so much.
-Chris
(1) Defects in manufacture, the player has to deal with that.
(2) I'm not certain what the length of the shaft has to do with anything realistically. Most of my good current machines have the longest shaft length that I can not get for some reason. It would make a difference if the disc was punched off-center, or the table was not true.
(3) We had one heck of a time with the first machines and CDs getting stuck on the shaft. Everything to holes too small, flashing still on the disc (causes scratches in automotive transports that use the soft rollers). Due to poor tolerances with many CDs, the centering cone has been forced to be slightly smaller.
When a CD player skips on some discs, the customer doesn't care about the why. They just want the thing to play. At least with the more expensive machines we could ask the customer why he is feeding cheap discs into an expensive machine. That was a valid observation actually. A concept that people spending lots of cash seem to understand. Try telling that to the bargain hunter with their $30 find on sale. You've lost that battle as soon as they lay their money down. Years of observation.
Any approach in particular?
Error reporting doesn't seem to be designed in. You would have to lock another servo to track the internal one and compare that to the eye pattern. Along with the slice level you should be able to determine where the determinations are being made exactly. Then looking at the eye pattern, you could tell what the probability would be of an accurate determination. That is why a clear, wide eye pattern is so critical to getting accurate data off the disc to begin with. If they mess this step up, everything that follows is a problem. Today they seem to be relying on error correction and error concealment to do the job that the transport is tasked with doing.
We are being ripped off in other words.
Well, it isn't uncommon for DVDs to be processed by a different signal path than CD data is. How's that for your assumption? They change the servo gains and time constants between DVD and CD play. It's like using two different machines. The DVD only has greater momentum because its rotating faster. There is a limit to how quickly a mass can be accelerated and controlled. The better DVD players are more heavily built than CD players. However that used to be true of CD players of higher quality too. Another factor is the level of static friction. Once you exceed that you have lost control of the CD or DVD entirely. There are very real limitations on how much force can be applied to any moving part.
You should have seen the results of a laser disc coming apart in the player - at speed. It was dangerous to the technician! I have heard of, but never witnessed a DVD coming apart after de-lamination of the two halves. I suspect that the vibration is enough to release the DVD or laserdisc from the clamp. It all comes apart once it contacts the support structures for the clamp, or flapper. Laserdisc failure can bend the top cover outwards from internal contacts. Lots of energy there. With DVDs, not so much.
-Chris
<sigh>
Mark, I have presented an anlysis. I clearly explained how power supply noise contamination modulates CMOS gate propogation delay. Such modulation directly manifests as jitter. The solution to minimizing jitter induced by this real mechanism is as obvious as it is difficult, which is to minimize supply noise on all components handling clock signals. Such remedy can be after the contamination is done via regulators and filters, but is better to remedy before the contamination via whatever measures that reduce the effect of loads injecting noise on the supplies. Preventing damage in the first place is superior to trying to minimize the damage after the fact.
As far as reclocking being the solution, I have one question for you. How do you use a jitter contaminated master clock signal to clean itself via reclocking?
Ken, you could buffer the data and use one clock for everything before the buffer and a dedicated clock for the DAC, a process that in a way does take place in a CD player (FIFO buffer).
Then again, if the DAC and clock are so susceptible to noise, how do you prevent them from contaminating each other or building a PSU that is noiseless?
I am sorry, for me, your analysis is too thin. I could give you a better one blaming pixies and include photo's. At least give us some numbers. How much noise is produced by which systems and this noise produces how much jitter or as I explained before increase the noise and see what happens. Take this hypothesis further.
Then again, if the DAC and clock are so susceptible to noise, how do you prevent them from contaminating each other or building a PSU that is noiseless?
I am sorry, for me, your analysis is too thin. I could give you a better one blaming pixies and include photo's. At least give us some numbers. How much noise is produced by which systems and this noise produces how much jitter or as I explained before increase the noise and see what happens. Take this hypothesis further.
Mark, that's an interesting take. I rather suspect that you are simply being obtuse about this. You assert that my analysis is too thin without providing any of your own analysis supporting that assertion. At any rate, there's no point in us continuing to knock heads on this. Peace.
No, I am serious. If this is a problem then the engineering costs are going to be very high. Air bearings, glass/marble platter, noiseless motors, compressors, vacuum pumps, the list goes on and on. It would be a shame if in the end the gains are minimal or even zero. Better to spend the money where it will have more effect.
You should not forget that all good designs were already done. Mechanically, it will probably be more or less copying. To me, building something like the BU-1 is the goal.
As far as I was able to research, no problem with patents, because those drives are 30 years old.
Electronically (i.e circuit design) the sky is the limit.
As far as I was able to research, no problem with patents, because those drives are 30 years old.
Electronically (i.e circuit design) the sky is the limit.
Hi Salar,
Yes!
Without getting too silly with the design. Something solid will do everything we need of it.
Hi Mark,
So far Ken has given you solid explanations. I don't know what else he could tell you - or me for that matter. Are we missing the core of your question, or misunderstanding what you are asking?
-Chris
Yes!
Without getting too silly with the design. Something solid will do everything we need of it.
Hi Mark,
That's how things currently work. It's the only way it can work.
Have a look at some good CD players where this has been done to a point where it isn't a major problem.
So far Ken has given you solid explanations. I don't know what else he could tell you - or me for that matter. Are we missing the core of your question, or misunderstanding what you are asking?
-Chris
Are the servos audible or not? If they are. How, why, to what degree and can/has it be measured? Normal engineering questions.
Is the engineering solution to improve the drive or to improve the DAC. Lets say we build a player using Ken's design. How will we compare it to an existing design or know there is a real reduction of jitter.
Is the engineering solution to improve the drive or to improve the DAC. Lets say we build a player using Ken's design. How will we compare it to an existing design or know there is a real reduction of jitter.
Last edited:
Hmm, been watching a while... Now, get us a mechanism compatible with the one used by Tvicol Audio in ShigaClone MKII. I do hope you are not trying to better any long gone old CD player of the past - are you? The ShigaClone MKII is a masterpiece and I would love to see a mechanism that is up to it - the ones used today leaves a bit for wishing. The actual one used today is SFP101N-16P.
Regards
Regards
Last edited:
Hi Mark,
CD technology has come a long way in reducing these issues, and they thought long and hard about it before the first one was produced. As for measuring this effect on the audio path (including digital data), you are chasing something that in good machines have been made into a non-significant problem. However, introduce bad engineering practices and you will probably have a problem.
So what is the heart of the matter for you? Are you trying to prove these things can not affect CD player performance in a round about way, or are you truly interested in how possibly servo power output could affect a modern CD player?
-Chris
In what way do you mean? As soon as you say the word "audible", you may as well forget trying to quantify it.
CD technology has come a long way in reducing these issues, and they thought long and hard about it before the first one was produced. As for measuring this effect on the audio path (including digital data), you are chasing something that in good machines have been made into a non-significant problem. However, introduce bad engineering practices and you will probably have a problem.
So what is the heart of the matter for you? Are you trying to prove these things can not affect CD player performance in a round about way, or are you truly interested in how possibly servo power output could affect a modern CD player?
-Chris
As I wrote, make a new SFP101N-16P and shut up believing and soft flavouring. Every bit of it can be bettered but it won't happen if you folks are batteling over non issues.
Servos sounds - can you hear it - yes if you put your ear to it - sh*t - do you usually stand in that position when listening to music? If not - no problem after a few feet away it is totally silent to your ear. So what? Still makes sounds? Sigh...
What I am not happy about with the SFP101N-16P is that it has too soft metals and to thin rods. For long time endurability it also have weak transport durability due to harsh metal against soft plastic.
But the transport itself is fantastic. I hope sincerely that whatever is created is compatible with the SFP101N-16P or many would be viewing the fact - buy the Shigaclone MKIII board. It's all about the money these days...
Regards
Servos sounds - can you hear it - yes if you put your ear to it - sh*t - do you usually stand in that position when listening to music? If not - no problem after a few feet away it is totally silent to your ear. So what? Still makes sounds? Sigh...
What I am not happy about with the SFP101N-16P is that it has too soft metals and to thin rods. For long time endurability it also have weak transport durability due to harsh metal against soft plastic.
But the transport itself is fantastic. I hope sincerely that whatever is created is compatible with the SFP101N-16P or many would be viewing the fact - buy the Shigaclone MKIII board. It's all about the money these days...
Regards
Last edited:
Hi Mark,
A stable platter as defined in the industry is harmful to getting the data off the disc. It will not allow the disc motor to track speed changes as required. Plain and simple. Of course, given a perfect disc where speed changes are not required at all, more rotating mass will not harm that process. It's hell on the motor when reading the TOC or jumping tracks though.
Hi Turbon,
The OMS-7 transport was the best. Period. The SFP101N-16P was okay, but not as good. The OMS-7 allowed mechanical adjustments in each plane, then became an intrinsic block of metal.
You can't beat that.
-Chris
A stable platter as defined in the industry is harmful to getting the data off the disc. It will not allow the disc motor to track speed changes as required. Plain and simple. Of course, given a perfect disc where speed changes are not required at all, more rotating mass will not harm that process. It's hell on the motor when reading the TOC or jumping tracks though.
Hi Turbon,
The OMS-7 transport was the best. Period. The SFP101N-16P was okay, but not as good. The OMS-7 allowed mechanical adjustments in each plane, then became an intrinsic block of metal.
You can't beat that.-Chris
Hi Mark,
A stable platter as defined in the industry is harmful to getting the data off the disc. It will not allow the disc motor to track speed changes as required. Plain and simple. Of course, given a perfect disc where speed changes are not required at all, more rotating mass will not harm that process. It's hell on the motor when reading the TOC or jumping tracks though.
Hi Turbon,
The OMS-7 transport was the best. Period. The SFP101N-16P was okay, but not as good. The OMS-7 allowed mechanical adjustments in each plane, then became an intrinsic block of metal.
-Chris
But TEAC (whoever the parent company is) made a living off their stable platter VRDS mechs (and also providing relatively unstable cheap VRDS versions too, ahem.. ) without playback issues in any sense.
I have a TASCAM CD701 studio player myself. One audio engineer told me it was the only CD player he could find for cues on stage with bands back in the day - all others would skip due to external noise and vibration, whilst the CD701 could be placed on top of a bass amp turned up full whack and still play fine.. Plus the CD701 copes perfectly with scrub in forward and reverse, pitch change, precise cueing to the split second (all of this without the optional RAM buffer board) suggesting that fine control of rotational speed is a non-issue with it.
With the VRDS's own motor and bearing added to the Sony mech, there are no bearing wear or servo control issues I'm aware of in this multi-thousand dollar studio machine.
A similar thing was done using similar/same Sony rail mech and the same KSS151A laser head in both the Denon DP-S1 and Yamaha GT CD1, the former using a weighty puck which covers the whole of the CD. Both machines use their own propreitary substantial motor assemblies :
http://www.thevintageknob.org/denon-DP-S1.html
http://www.thevintageknob.org/yamaha-GT-CD1.html
Last edited: