| pcm1702, K grade or not? - Click HERE for Original Thread |
| PatPet |
Hello,
I'm interested in this DAC. But when I look at the price, I start to consider if it is worthwhile to go for the K-grade chips which costs nearly 60% more. Well, it seems 29USD for each 1702-K is too much. But I start to think about balanced/parallel solution but how will 1702-J compare with 1702-K in these situations?
Guys, I'm stuck between the two options, either balanced/parallel 1702-J or balanced/parallel 1702-K. Which one shall I choose?
Thanks
Pat |
|
|
| Bernhard |
| I have measured lots of PCM56, PCM58, TDA1541 and I do not give a penny for those stamps anymore. |
|
|
| PatPet |
Thank you Benhard,
And which would you prefer, if you have to choose between pcm1702 and pcm56?
I'm deciding which is better for balanced/paralleled application.
The pcm56 is very cheap, around 10USD each, but doesn't rate as good as pcm1702 in dynamic range. But in the datasheet, pcm56 is described as the state of the art chip. |
|
|
| Elso Kwak |
| quote: | Originally posted by PatPet
Thank you Benhard,
And which would you prefer, if you have to choose between pcm1702 and pcm56?
I'm deciding which is better for balanced/paralleled application.
The pcm56 is very cheap, around 10USD each, but doesn't rate as good as pcm1702 in dynamic range. But in the datasheet, pcm56 is described as the state of the art chip. |
Hi patpet, The PCM56 is an antiquated chip. I found the AD1851 better and the AD1865J-K much better. Just my personal experience of course.
:bigeyes: |
|
|
| PatPet |
Hello Kwak
Could you describe the difference in sound with the three chips that you've mentioned? I'm eager to know how the analog chips sound.
Thanks |
|
|
| Bernhard |
Pat,
I have 4 pieces of the 1702 but did not try them yet, performance differs a lot from chip to chip, that is why they are selected.
But manufacturers do not look at the spectrum, only measure total thd :(
Also performance may change after burn in. |
|
|
| PatPet |
Thank you Benhard
I've got the idea that it's unnecessary to buy the "select" graded chips. |
|
|
| Bernhard |
| quote: | Originally posted by PatPet
Thank you Benhard
I've got the idea that it's unnecessary to buy the "select" graded chips. |
Best try for yourself... |
|
|
| jwb |
I used to balk at the cost of the 1704U-K (US$36 each), but after the design was done the difference between the K and the J (US$12 each) seemed insignificant. A DAC that can take advantage of the performance of the 1704 will undoubtedly run into the many hundreds of dollars anyway, so why not spend the extra $48?
On the other hand with the 1702 I can understand that you might not be spending as much money on a 44.1kHz/16-bit design as you would be forced to spend on a 96kHz/24-bit implementation. |
|
|
| PatPet |
| Thank you everyone for your suggestions. I've decided to go for ad1865-k. Shall I start a new thread to discuss about the project or shall I just continue right here? |
|
|
| BrianL |
A friend of mine has measured the linearity of a number of B-B
DACs. (he has such a setup as part of his bench equipment since
he deals with converters a lot). There is a significant improvement
in linearity when measuring the high-grade parts, given randomly
chosen samples of each. Whether this translates to audible
improvement is left as an exercise for you the user. I don't
believe that he's tested any ADI DACs, but I would assume the
results are the same.
The reason for the cost difference between high grade and low
grade is the amount of laser trimming of the die at wafer level
(or post-packaging electronic "trim"). The high grade gets trimmed
to tighter tolerances (and/or the low grade may be untrimmed). |
|
|
| Jocko Homo |
The AD1862, when using the MSB trimpot, stays adjusted for as long as I have had them running. That goes back to around '92 or so. The units that come back for upgrades are still set to the precise point. Hasn't drifted at all.
I agree with Brian...........on the bench.......there is a measureable improvement in the selected grades. Some of the "non-selected" ones are terrible, which I why I always use the "high-priced spread".
Whether you can hear that low............more agreement.
Jocko |
|
|
| Elso Kwak |
Hi,
An AD1865J-K (highest grade) does sound way better than a MSB adjusted AD1865J-N (lower grade) and does not need an adjustment pot......Yet I prefer a TDA1543. Ha-ha it even does not accommodate a pot. Low-level signal is OK, dunno how they did it.
For those interested in the adjustment I attach a small low-pass filter amplifier adjusted for NON-OS DAC to visualise the low level signal on your scope. The adjustment is difficult to perform without the rig, as the signal is drown in noise.
:cool: |
|
|
| PatPet |
Hi
Thank you for all of your responses.
In my plan, the ad1865-k are to be stacked in parallel. Which means, by hooking up at least 4 of them in each channel, I may have better "spec", and probably better sound also. The reason I consider parallel solution is that the conversion errors would be reduced and that I may have greater S/N, THD, DR specifications.
This may be quite expensive. But I would invest more on this DAC as I have much expectation from it.
And for the low-pass filter, I must say it is very useful. But I don't have a scope and I'm afraid I have to leave the MSB unadjusted, may be at a expense in low level signal performance. I may consider that the adjustment becomes less useful in parallel DAC situations. |
|
|
| Bernhard |
| quote: | Originally posted by Elso Kwak
For those interested in the adjustment I attach a small low-pass filter amplifier adjusted for NON-OS DAC to visualise the low level signal on your scope. The adjustment is difficult to perform without the rig, as the signal is drown in noise. |
This is just a joke, isn`t it ? |
|
|
| Elso Kwak |
| quote: | Originally posted by Bernhard
This is just a joke, isn`t it ? |
Bernard,
You are funny too!
Is everybody funny?
;) |
|
|
| Jocko Homo |
Well, if you want to see what that -100 dB test tone on your CBS disc looks like you will need to amplify and filter it.
Jocko |
|
|
| Elso Kwak |
| quote: | Originally posted by Bernhard
This is just a joke, isn`t it ? |
Drowning in a Sea of Noise
Sie will ein Fisch im Wasser sein
Im flaschengrünen, tiefen See.
Sie wil mit Wasser sich besaufen
Und ein paar Blasen blubbern lassen
Was Sie dann will, das ist
Mit Neptun schweigen
Und in Ruhe thun,
Was sie sonst nie thut,
Was sie sonst nicht kann & soll.
For our Anglo sax friends, especially MalichiConstant (I know you are browsing this forum), I have translated this:
She wants to be a fish in the water
in the bottle green, deep lake.
She wants to fuddle herself with water
And bubble up a few bubbles
What she then wants, is
being with Neptune silently
and doing quietly
what she otherwise never does,
What she otherwise not can do and has to.
Elso/ Pisces
Courtesy Nina Hagen
:hbeat: |
|
|
| Elso Kwak |
| quote: | Originally posted by PatPet
Hi
Thank you for all of your responses.
In my plan, the ad1865-k are to be stacked in parallel. Which means, by hooking up at least 4 of them in each channel, I may have better "spec", and probably better sound also. The reason I consider parallel solution is that the conversion errors would be reduced and that I may have greater S/N, THD, DR specifications.
This may be quite expensive. But I would invest more on this DAC as I have much expectation from it.
And for the low-pass filter, I must say it is very useful. But I don't have a scope and I'm afraid I have to leave the MSB unadjusted, may be at a expense in low level signal performance. I may consider that the adjustment becomes less useful in parallel DAC situations. |
Well PatPet, I certainly do not want to hear more silly jokes about piggy backed or stacked DACs after Lars' dry humour with electronic parts handling.:clown:
As I explained the K-grade is the best and really does not need a MSB adjustment as this makes matters only worse.
In my experience and that of Peter Daniel the highs suffer when DACs are paralleled.
:cool: |
|
|
| PatPet |
Thank you Kwak
I understand that ad1865-N-K is already good enough. But I just got the idea from pcm1702 application note, in which paralleling was mentioned. I just wonder if that is untrue to claim that the SNR or the DR improved.:clown:
http://focus.ti.com/lit/ug/sbau029/sbau029.pdf |
|
|
| Elso Kwak |
| quote: | Originally posted by PatPet
Thank you Kwak
I understand that ad1865-N-K is already good enough. But I just got the idea from pcm1702 application note, in which paralleling was mentioned. I just wonder if that is untrue to claim that the SNR or the DR improved.:clown:
http://focus.ti.com/lit/ug/sbau029/sbau029.pdf |
Hi Patpet,
Paralleling certainly improves signal to noise ratio but if this iinfluences sound quality I wonder......Noise is usually not a problem with digital audio.
:confused: |
|
|
| Bernhard |
PatPet,
the situation is, that DAC sound is judged subjectively by most people.
This is the reason why some technically bad performing chips like TDA1543 are praised to heaven.
The human ear is pleased by low and even order harmonics and disgusted by high and odd order harmonics.
Now if a circuit of any kind produces the right harmonics, it is preferred over another circuit that adds nothing to the signal.
If one stacks chips, the lovely harmonics might get lost :bawling:
I recently found that SAA7220C combined with TDA1541 shows dominant even order harmonics and sounds very good :D
Maybe I make a RF relay switch between 7220A and 7220C. |
|
|
| QSerraTico_Tico |
| quote: | Originally posted by Bernhard
PatPet,
the situation is, that DAC sound is judged subjectively by most people.
This is the reason why some technically bad performing chips like TDA1543 are praised to heaven.
The human ear is pleased by low and even order harmonics and disgusted by high and odd order harmonics.
Now if a circuit of any kind produces the right harmonics, it is preferred over another circuit that adds nothing to the signal.
If one stacks chips, the lovely harmonics might get lost :bawling:
I recently found that SAA7220C combined with TDA1541 shows dominant even order harmonics and sounds very good :D
Maybe I make a RF relay switch between 7220A and 7220C. |
Bernard,
Sound quality is subjective by definition. But that is the only result what counts. I do not listen to measurements but to sound. The TDA1541 has so little energy at 6000Hz that it sounds dull and boring to my ears. King Gang Bang will raise in anger now...... The nice thing about DIY is you can build whatever you like.
:rolleyes: |
|
|
| Bernhard |
| quote: | Originally posted by QSerraTico_Tico
1) Sound quality is subjective by definition. But that is the only result what counts.
2) The TDA1541 has so little energy at 6000Hz that it sounds dull and boring to my ears.
|
1) That`s the point.
Some people love Ghetto Blasters.
I prefer clean sound over rich harmonics, the harmonics make every CD sound the same.
2) What does that mean ?
Did you measure that ?
Did your ears tell you ?
Bad implementation of chip ? ? ? :D :D :D
Dr. Guru`s advice:
Try 3D Voodoo ground plane and don`t forget hell`s gate decoupling caps :D :D :D |
|
|
| PatPet |
| I've got quite a number of ad1865n-k in hand now. I think I shall start the project soon. |
|
|
| erozsolt |
Kwak: Do you have a good schematic layout of AD1865-K modified with your clock (Or any good source DIY project, which is a good starting point to a mod)? I would like to use my computer as a source, locked to external clock, and I found your clock is the best for implementing into DIY projects!
Thanks |
|
|
| Elso Kwak |
| quote: | Originally posted by erozsolt
Kwak: Do you have a good schematic layout of AD1865-K modified with your clock (Or any good source DIY project, which is a good starting point to a mod)? I would like to use my computer as a source, locked to external clock, and I found your clock is the best for implementing into DIY projects!
Thanks |
You mean the Asynchronous Reclocker?
Just mail me.:cool: |
|
|
| Peter Daniel |
| quote: | Originally posted by Elso Kwak
In my experience and that of Peter Daniel the highs suffer when DACs are paralleled.
:cool: |
Yes, it was my experience. When paralleling you loose ambience and directness, with less resolution in highs. Sort of similar effect to using regulated supply on a GC ;)
But, in my recent DAC, using TDA1541S2 I actually prefer 2 paralleled DACs, but only when oscillator caps from each chip are connected together. Otherwise I would prefer a single chip.
BTW, both of my DACs (parallel TDA1541 and a single TDA1543) sound pretty much similar. However, when it comes to a purity of the sound, my choice is well implemented TDA1543. I guess the other DAC still needs more work. |
|
|
| erozsolt |
| quote: | Originally posted by Elso Kwak
You mean the Asynchronous Reclocker?
Just mail me.:cool: |
I replied to the AD1865-K project topic. |
|
|
|
|
