From Shitty X-DAC V3 To Something

[XTRProf]

May 10, 2008 5:55:58 GMT @sandyk said:

XTRProf
Is the added 60uF of capacitance at each supply rail of the I.C. , AFTER the dropping resistor ?
In that case they will only have a minor effect, as at 20HZ, their capacitive reactance is around 130 ohms !
Alex

Yup! Finally, they will be gone as I'm looking forward to discrete amplification there to improve further on things. Ya, tubes also possible. Oops, I better shut my mouth.

[Deleted]

Ya, tubes also possible. Oops, I better shut my mouth.

Don't just shut your mouth, rinse it out with hot soapy water too !

[XTRProf]

May 10, 2008 7:51:39 GMT @sandyk said:

Ya, tubes also possible. Oops, I better shut my mouth.

Don't just shut your mouth, rinse it out with hot soapy water too !

Yes, sir! Salute ..................

[leo]

The Vcom pins are for the internal bias for the internal analogue stages for the dac chip, using good quality LOW LEAKAGE caps is highly recommended.

[leo]

= 0.0366 Vac based on 4 significant figuresThe Time Constant t = 1 / fBased on 100 Hz, t = 1/ 100= 0.01 sC required = (I / Vac) (T)Therefore, C required = (0.096 / 0.0366) (0.01)= 0.026229508 F= 26.2295 uf based on 4 deciomal placesBut for 60 uf, that will be = 60 / 26.2295= 2.2875 timesThe ripple at the local OPA PS = 0.0366 / 2.2875= 0.016 VacRipple % at local OPA PSs is about = (0.016 / 16.82) (100)= 0.09512 % to 5 decimal places

Looks like a pretty paperweight Miguel

[insomniac]

Heh, heh ......... I still remember one of my lecturers telling the cohorts why have an engine when you can have the donkey to do the job.

Because engines don't leave huge turd's on the road

[leo]

May 10, 2008 18:44:09 GMT mrarroyo said:

May 10, 2008 13:40:20 GMT leo said:

... Looks like a pretty paperweight Miguel

You of little faith!

I'm only joking Miguel, I'm sure it sounds very nice

[Deleted]

... Looks like a pretty paperweight Miguel



You of little faith!

Miguel
Many ladies of the house would love it as an ornament ?
It would make a very nice , unobtrusive night light too !
Shame about the running costs though. ;D

[XTRProf]

May 10, 2008 13:39:08 GMT leo said:

The Vcom pins are for the internal bias for the internal analogue stages for the dac chip, using good quality LOW LEAKAGE caps is highly recommended.

Thanks very much Leo. You are the correct digital man to approach all the time! Great!

Just one more question. How come the caps are connected in series at the Vcom legs? Wouldn't that block dc from going in for dc power? AC will have no problem obviously.



Cheers!

[XTRProf]

May 10, 2008 12:26:20 GMT mrarroyo said:

The enclosed picture is for Mr. Tube aka Alex! ;D

Oh, I love tu ............ Oops again! Sorry sir, please don't throw me into the guard house cell!

[XTRProf]

May 10, 2008 13:41:44 GMT insomniac said:

Heh, heh ......... I still remember one of my lecturers telling the cohorts why have an engine when you can have the donkey to do the job.

Because engines don't leave huge turd's on the road

;D

[Deleted]

Heh, heh ......... I still remember one of my lecturers telling the cohorts why have an engine when you can have the donkey to do the job.

Hey , Singapore Stallion , years ago I had a workmate nicknamed Donk. The signwriter painted a donkey with a big dick,but his head, on his toolbox. The girls from "Service Difficulties" thought it was hilarious ! Are you sure you aren't the donkey in this thread?



That's payback for the tube jibes ! ;D
SandyK

[PinkFloyd]

Looks like some kind of warped female sex toy to me.... what amp is it Miguel? TMS?

[Deleted]

Looks like some kind of warped female sex toy to me.... what amp is it Miguel? TMS?

More likely PMS !

[XTRProf]

Good morning World and Rockgrotto,

I'm back! As it's still quite dark to do some good soldering work, I quess I visit here first and do some more maths and facts finding missionary work. Ya, I got the last 4 EC BP caps and will solder them in when the light is out. I will report back how it sound later with a picture.


FACT 6: Whatever applied here is supported by the application notes of the OPAs for optimum operation. Nothing new. Just check out the spec sheets of AD8066 and LM4562.


FACT 7: Why passive or discrete instead of active stages, like having a passive pre, although they do have their downside like reduced bass? It's because:

Device             Ein (Typical Equivalent Input Noise) dbu

OP27                  -127
OP37                  -124
5532                  -122
TL072                 -116
2N4403                -135
2SA1316               -144
3R3                   -146
1k                    -123

Reference: Rossiter, Paul L., "A Head Amp for Very Low Impedance Moving Coil Cartridges," AudioXpress, Table 2, p26, Sept 2006.


FACT 8: Let's go to the dropping resistors network of 2.5 ohms each rail and calculate the maximum current possible through them.


P = VI

As those resistors are rated at 10W and the voltage drop across each of them is about 0.24 VDC, rearranging

I = 10 / 0.24
= 41.6667 A based on 4 decimal places

So which will current limit as compared with a Vreg? You think, ok, as this is fact finding .................



FACT 9: Let's revisit the 12,000 uf smoothing caps and work out the impedance at say 50 Hz and 1 MHz


Reactance (Impedance when ignoring R and Inductance) X = 1 / 2 pi f C


So at 50 Hz,

X = 1 / (2 x 3.142 x 50 x 12,000 ^ -6)
= 0.26 ohms

At 1 Mhz,

X = 1 / (2 x 3.142 x 10 ^ 6 x 12,000 ^ -6)
= 1.3261192 ^ -5
= 0.00001326 ohms based on 4 significant figures

Of course, this is in theory and it will not go to that due to internal resonance of the e-cap at high frequency. But the theoretical maths is there and I don't have the graph for this 12,000 uf to comment further. That's why we need smaller e-cap bypass to assist in this type of monster.



That's all folks for now as I need to go to the market as daylight is already out. The soldering part will be in the afternoon then. For the next fact finding mission, I will cover the resistance, inductance and capacitance parts that could be present in the PS line for your informed judgements and not fantasies and hearsays.




Note: I will like to highlight here there is no right or wrong. I'm not saying the gurus here are wrong or right. It's just different views or beliefs from another perspective. So flame throwers, please don't come here to spoil the fun on this silly little thing called Hifi. Just use these facts for your beliefs and enjoy the fun. Don't take the fun out of it.

[XTRProf]

May 10, 2008 23:28:06 GMT @sandyk said:

Heh, heh ......... I still remember one of my lecturers telling the cohorts why have an engine when you can have the donkey to do the job.

Hey , Singapore Stallion , years ago I had a workmate nicknamed Donk. The signwriter painted a donkey with a big dick,but his head, on his toolbox. The girls from "Service Difficulties" thought it was hilarious ! Are you sure you aren't the donkey in this thread?



That's payback for the tube jibes ! ;D
SandyK

Well, I don't mind riding the donkey as long as it does an excellent job! Riding a donkey is a challenge, you know ............. I will let the facts do the talking. So don't leave home without it as in American Express. Heh, heh, heh ...........

[Deleted]

XTRProf
You have been laboring the point that your big 12,000uF capacitors have reduced ripple by a factor of approximately 46 times. Well, to put that into some kind of context, even a reduction of 100 times, is only a 40dB improvement. A voltage regulator such as the LM317T with a 10uF capacitor bypassing it's adjust terminal, typically gives a ripple reduction of 80db ! That's a helluva lot better ! So just how does this supply of yours fit in with your statement ? "....with such a low ripple rejection the PS should technically on paper be quieter and more musical sounding."
FACT is, Your FACTS, are mainly conclusions you have jumped to ! I suggest that you change your FACT to CONCLUSION.

SandyK

[leo]

And then stick a JLH on the output and its improved even more! those BG's can then be chucked in the parts box

[Deleted]

those BG's can then be chucked in the parts box

Leo
Did you mean Parts box, or Recycle Bin ?
Perhaps the Prof could resell them on ebay to some other sucker ?
Alex

[leo]

May 11, 2008 10:48:25 GMT @sandyk said:

those BG's can then be chucked in the parts box

Leo
Did you mean Parts box, or Recycle Bin ?
Perhaps the Prof could resell them on ebay to some other sucker ?
Alex

Non ROHS so you can't even recycle them
He could sell them on ebay stating they have been run in for a few thousand hours, should only need say another 2 year burn in time

[XTRProf]

May 11, 2008 3:42:36 GMT @sandyk said:

XTRProf
You have been laboring the point that your big 12,000uF capacitors have reduced ripple by a factor of approximately 46 times.

Nope, I'm not labouring for that as it's just one of the parameters on this fact finding mission. There are other considerations as clearly writtten in my factual points. You are since you only mention that here and nothing else .........

May 11, 2008 3:42:36 GMT @sandyk said:

Well, to put that into some kind of context, even a reduction of 100 times, is only a 40dB improvement. A voltage regulator such as the LM317T with a 10uF capacitor bypassing it's adjust terminal, typically gives a ripple reduction of 80db ! That's a helluva lot better ! So just how does this supply of yours fit in with your statement ? "....with such a low ripple rejection the PS should technically on paper be quieter and more musical sounding."

Alex, Alex guru ............., you are again pointing to only one parameter and nothing else. Musicality comes from a holistic way. Please check again what I had written so far ..........

Anyway, if you want the maths and real interpretation of the datasheets, here they are:


The data is for 120Hz for the LM317. If based on 120 Hz for my case the cap for 10% ripple will be = (5 x 0.096) / (1.414 x 13) (120)
= 217.6042 uf to 4 decimal places

So 12,000 / 217.6042 = 55.1460 times

and the ripple percentage for each of my 12,000 uf smoothing cap = 10% / 55.1460
= 0.1813% based on 4 significance figures

Ripple dB = 20 log (Vac/Vdc)
= 20 log 0.001813
= -54.8302 dB to 4 decimal points!

Anyway, your -80 db is with a power supply cap of 10 uf and without it's about -60 db. I have 60 uf local power supply cap at the OPAs. I let you figure out what will happened with the the 60 uf onboard in a technical minded way. Again, this don't tell everything that only Zen can tell. Why manufacturers have many prototypes before the final product launch?

Like I had said before, there are other parameters to musicality as well. LM317 Imax is only 3.4 A based on K, T, S packages. Mine is almost without restriction with a 2.5 ohms dropping resistor all the way to the max I of my 225VA transformer ie about 10 A. Anyway, line and load regulation is not that important in audio amps, within said 10%, unlike a dac or others at said voltages of 5 VDC and 3.3 VDC. I will elaborate later in my next fact finding mission.

Quieter? I will just say the LM317 has a noise output of 0.003% from 10 Hz < f < 10 khz.

So in terms of dB = 20 log 0.00003
= -90.4576 dB to 4 decimal places

So go and figure that out guru, sir! I will again elaborate on my next fact finding mission on this so as not to dilute from my mission statement and that's to put the figures out and let all decide for themselves here. There is just no right or wrong but your beliefs

May 11, 2008 3:42:36 GMT @sandyk said:

FACT is, Your FACTS, are mainly conclusions you have jumped to ! I suggest that you change your FACT to CONCLUSION.

SandyK

OMG, a supposedly technical person arguing with technical maths. Well, I hope the deans of enginering faculties of the University of Melbourne and New South Wales see this ..........

Anyway, those calculations are not from my thougths or whatever. It's from noted researchers like Ohm, Kirchoff, etc. So they are not my conclusions, as you supposedly said so here, but I'm using what I know taught to me, and not from my research, to show the figures here. If you don't agree please approach people like Ohms and others in their graves and I'm sure they will turn in their graves. As they are they are still facts and not conclusions as they are based on technical maths and when properly calculated correctly.



So long folks as I need to eat and have some relaxation now! Bye .....................

[Deleted]

Anyway, your -80 db is with a power supply cap of 10 uf and without it's about -60 db. I have 60 uf local power supply cap at the OPAs. I let you figure out what will happened with the the 60 uf onboard in a technical minded way

XTRProf
Yet once again, you have appeared to misunderstood the data sheets. The 10uF capacitor bypassing the adjust terminal has NOTHING to do with the 60uF at the opamp. Just about every circuit you will see using these adjustable VRs has this capacitor fitted. It basically smooths an internal reference for the VR.

LM317 Imax is only 3.4 A based on K, T, S packages. Mine is almost without restriction with a 2.5 ohms dropping resistor all the way to the max I of my 225VA transformer ie about 10 A.

That may be so into a bloody short circuit !
But as you want 15V rails to the opamps, that drops YOUR maximum available current to not much more than 1Amp.

I admire the way you are using the formulae to work things out. However, your overall methodology has great big gaping holes in it, as evidenced by the quote above.

I can find better things to do than wade through all your "FACTS" to explain everything that you are doing wrong!

Surely you have noticed that other more technical members are ignoring this thread ?
SandyK

Chung

I apologise for the tone of my last couple of sentences, however the sarcasm of your previous posting was the reason. If you wish to continue this thread without any further pointing out of what I consider are errors, and not learn about some of the mistakes you are making along the way , I am quite prepared to take no further part in it. However, you may find yourself the only one posting in this thread. Perhaps you prefer that?

[XTRProf]

May 14, 2008 21:39:08 GMT @sandyk said:

Anyway, your -80 db is with a power supply cap of 10 uf and without it's about -60 db. I have 60 uf local power supply cap at the OPAs. I let you figure out what will happened with the the 60 uf onboard in a technical minded way

XTRProf
Yet once again, you have appeared to misunderstood the data sheets. The 10uF capacitor bypassing the adjust terminal has NOTHING to do with the 60uF at the opamp. Just about every circuit you will see using these adjustable VRs has this capacitor fitted. It basically smooths an internal reference for the VR.

LM317 Imax is only 3.4 A based on K, T, S packages. Mine is almost without restriction with a 2.5 ohms dropping resistor all the way to the max I of my 225VA transformer ie about 10 A.

That may be so into a bloody short circuit !
But as you want 15V rails to the opamps, that drops YOUR maximum available current to not much more than 1Amp.

I admire the way you are using the formulae to work things out. However, your overall methodology has great big gaping holes in it, as evidenced by the quote above.

I can find better things to do than wade through all your "FACTS" to explain everything that you are doing wrong!

Surely you have noticed that other more technical members are ignoring this thread ?
SandyK

Chung

I apologise for the tone of my last couple of sentences, however the sarcasm of your previous posting was the reason. If you wish to continue this thread without any further pointing out of what I consider are errors, and not learn about some of the mistakes you are making along the way , I am quite prepared to take no further part in it. However, you may find yourself the only one posting in this thread. Perhaps you prefer that?

Alex,

No need apologies and flame suits at least in my threads! As usual, there will always be people who will be interested or uninterested whether they are correct or not. There are those who for face values will just ignore things even when they are correct. There are lots here, btw! Sometimes the truth can be painful for some to swallow. Also, not forgetting, there are those in coalition or collusion. I don't believe in that or work that way and will accept, without conditions, what are righter and logical just like those few you have pointed in previous posting. There are also those standing by the sidelines who like to enjoy the emotions or learn something from it as they are introverts and wanted to avoid controversies. Also, there are those who doesn't know where to start or have the knowledge to contribute meaningfully to what were written. The people presenting here are mostly the "G8 senior citizens" of Rockgrotto, me included, btw. No wonder the forum here is so little traffic which I'm trying my best to boost up through what I had written! As I had said earlier, I let the people decide for what I had written. They are not my conclusions but figures to share for all here to decide for themselves what are righter for their applications as they are projected open ended. You are welcome to write what you feel about things just like I will write what I know to share with all. I'm opened to all things as this is what forums are all about. Isn't forum about sharing of information and debating in a scholistic manner? Why have forums when there is nothing to share or debate in a logical manner and not for face values. Face value forums are not real forums, btw! I'm from a half-full philosophy rather then a half-empty one.

So people here judged not for me or Alex, but for yourself. I rest my case and go back on topic!

[XTRProf]

REPORTING PART NINE




Gone is the mortar board! Ok folks, I'm back on track to post more of what I had done. Here goes :




The full monty EC and BG Super E-Caps PSs at the OPAs!






SQ


As expected a better refinement of the half job Super E-Caps as reported earlier. Now the sound seems more focused like a camera coming into alignment with extreme 3D effect. I will give it a 0-5% improvement range depending on music genres played.



Out of curiousity I replaced the Jamicon 2,200 uf 35V e-cap meant for the 5V dac digital side and the oscillator with an Elna Audio 6,800 uf 63V E-Cap here. The idea is to reduce the ripple seen by the 7805 Vreg so that they have lesser work (correction) to do as in a Solomon Reed error corrector doing lesser correction and sounding better:





SQ



Nope, not my cup of tea for first impression as the cap was not run in yet. The effect is like when I had replaced the analog PS side with the 12,000 uf e-caps with a bigger volume but grainier sound. I will need some more super e-caps to captured more of the extreme 3D magic when I have the time. All-in-all, it's still better than when I replaced the analog PS Jamicon caps with the 12,000uf when there were no Super E-caps around. If based on this finding, it seems the Jamicon E-Caps are not so bad after all as if I put it back at the 5V PS meant for digital, I will again have the extreme 3D very musical sound again.



That's all for now as I now put on my mortar board again for the fact finding mission coming next.

[XTRProf]

???After a short break, the numpty paperweight mortar board is back in another paperweight mission:




FACT 10: Why Super E-Caps as local PS? Here goes:



The data is for 120Hz for the LM317. If based on 120 Hz for my case the cap for 10% ripple will be = (5 x 0.096) / (1.414 x 13) (120)
= 217.6042 uf to 4 decimal places

So 12,000 / 217.6042 = 55.1460 times

and the ripple percentage for each of my 12,000 uf smoothing cap = 10% / 55.1460
= 0.1813% based on 4 significance figures

Ripple dB = 20 log (Vac/Vdc)
= 20 log 0.001813
= -54.8302 dB to 4 decimal points!

Based on LM317, -80 db is achievable with a power supply cap of 10 uf and without it's about -65 db. But I do have 60 uf local power supply cap at the OPAs. Even if this is not that accurate, line and load regulation is not that important in analog audio amps, within said 10%, unlike a dac or others at said voltages of 5 VDC and 3.3 VDC. On top of this, there is a almost a free flow of current (no current limiting) all the way to the transformer Imax of about 10A primary if based on the dropping resistor of 2.5 ohms I'm using. It obviously will be better without that resistor. I don't like it as well but no choice.


FACT 11: Why Super E-Caps as local PS for quieter noise figures? Here goes:



LM317 specs has a noise output of 0.003% from 10 Hz < f < 10 khz.

So in terms of dB = 20 log 0.00003
= -90.4576 dB to 4 decimal places

Based from the BG website here,

at about 10kHz input, the noise and distortion is = -174 db at about 10mV
at about 500mV = -120 db more or less stabilisng here all the way to higher inputs.



FACT 12: Where do we used Vregs?


Ok, I'm back! Ya, where am I? Oh, ok, on track again .............

Regulated power supplies are required where the voltage levels must be tightly controlled for proper operation of the circuit.... This is particularly important in many applications where small deviations from the design voltage level can affect the operation of the equipment being supported by such power supply.Check this out!


For example, say a dac chip at 5 VDC PS input, a line and load variation of say 10% will give = [5 - (5x0.1)] to 5V
= 4.5 to 5V VDC assuming positive regulation and varying mind us. So the dac chip can give more jitters due to the varying voltages. Sometimes, it may not work properly as well depending on how critical is critical for this chip to function properly. At said, 3.3V, this will be worst. That's why at much lower voltages we must have regulators and for dac chip, this will also give lower jitters! Unregulated will not provide that stability at such low voltages and have good jitter with unregulated PS. I will stick with either the LF7805 and LF33 on the digital side until I can stick superregulators here! No choice, really. Ya, thinking of it, I must have a superclock as well to lower jitters further just like regulators can do.

On the other hand, for analog amps, this is not so critical as first they are always operating at a higher voltage and varying voltages of say 10% will not give bad sound or not working properly as in digital cases.



That's it for now from numpty paperweigh mortar boarder. Cheers! .............................. OMG, it's TGIF. I had forgotten that. Heh, heh, heh .....................