ChuckT said:
Hi Chuck,
I will definately try the PSU blackgate cap.
I will have a look at the SMD caps. If they have left them there Iwill try desoldering them. (Have never hassled around with SMD parts so far, but there has to be a first time for everything...).
Are these Rubycons any good?
I will order blackgate NX as digital decoupling caps.
Do you think that it is best to remove MKP bypasses on those caps that will be exchanged for blackgates?
What caps would you recommend next to the regulators?
BTW, diodes... Do you think we should upgrade the diodes in the SPSU? The change of diodes in the linear PSU of the SACD 1000 has brought an incredible sound-improvement.
Regards,
Lucas
Do you think that it is best to remove MKP bypasses on those caps that will be exchanged for blackgates?
I heard from multiple sources that bypassing BG NX's makes them sound worse and destroys their sonics.
BG site recommends using two half-valued NX's paralleled for the least possible destortion (super-E configuration)
That is if you need a value of 1000uF , you parallel together two 500uF caps
Alex
I heard from multiple sources that bypassing BG NX's makes them sound worse and destroys their sonics.
BG site recommends using two half-valued NX's paralleled for the least possible destortion (super-E configuration)
That is if you need a value of 1000uF , you parallel together two 500uF caps
Alex
Hi Lucas,
Yes, do not bypass BG caps, I've tried it.
Also, I would only remove the SMD ceramic cap for the opamp decoupling only if you already have a film cap bypass. If you use BG for those, you can get rid of the film cap also.
I have no idea why 1000uf is used for coupling, unless your preamp has a very low input impendance, I'd say 100uf is more than enough.
I believe the change of diodes in the PS should improve things, but I can not say how much. Also, my stupid PS gets very hot, so maybe I've changed too much. So don't touch a Switch Mode unless you know what you are doing.
Could you tell me if your PS board gets hot also??
Yes, do not bypass BG caps, I've tried it.
Also, I would only remove the SMD ceramic cap for the opamp decoupling only if you already have a film cap bypass. If you use BG for those, you can get rid of the film cap also.
I have no idea why 1000uf is used for coupling, unless your preamp has a very low input impendance, I'd say 100uf is more than enough.
I believe the change of diodes in the PS should improve things, but I can not say how much. Also, my stupid PS gets very hot, so maybe I've changed too much. So don't touch a Switch Mode unless you know what you are doing.
Could you tell me if your PS board gets hot also??
I am interested in modding my 963sa. I have already ordered a replacement AD1986 upsampler (140db snr) and an AD1955 DAC to stack. Does anyone know where to buy the clock M2XO boards used in Hatasa's mods from? It looks like they use Mita TCXOs instead of normal crystals. I hope that they are cheaper than the superclock II offerings! Thanks
I am currently using the 963SA connected via digital to a Tag McLaren AV32R processor with 24/96KHz dacs but I am not entirely sure if it sounds better than Audiolab 8000CD transport that I am comparing it to! (yet)
V
I am currently using the 963SA connected via digital to a Tag McLaren AV32R processor with 24/96KHz dacs but I am not entirely sure if it sounds better than Audiolab 8000CD transport that I am comparing it to! (yet)
V
ChuckT said:
Hi Chuck,
Waht good are these SMD ceramics anyway?
Yes, strange, this high capacitance... My premp has a normal impedance.
I was planning to use BGs for the output caps. A BG seller told me that the FK is somewhat sweeter (and a bit less detailed) than the N or NX.
Now I read the thread:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=19005&perpage=15&pagenumber=4
In this thread some state that BGs are not so ideal as output caps. Peter Daniel prefers Siemens MKV etc...
How good are the Rubycons and Nichicons that Philips employed...?
Yes Chuck, the standard, not modded, SPSU in my 963 gets pretty hot. Enough to make the righter part of the enclosure quite warm.
About modding the SPSU: Does anyone know of general tips or examples how to improve such a SPSU?
Regards,
Lucas.
vulture said:
Hi Vulture,
Where did you order your AD1986? Is it 100% compatible with the AD 1895?
Let us know whether you hear an improvement.
Up till now I am always somewhat sceptical about the upsampling fuss. Often it seems that the balance shifts toward a more pronounced treble, at the expense of tonal integrity, giving the illusion that there is more detail, but in reality I missed some "bottom" in the sound. However, that may be only a matter of cheap implementing of upsampling...
Can you (or anyone else) explain the advantage of stacking DACs?
About the use of the 963 as a drive: It is my impression too that the 963 (in its standard form) is not the very best drive with an external DAC. My Tascam professional CD recorder gets a little more out of my external Music Fidelity DAC than the Philips.
This might however change when we do some PSU and clock-modding...
Regards,
Lucas
Hehehe!!! See what I done today???
I've tried to stack up the AD1955.... It is hard and took me 20-30 mins to finish it. I "simulate" the process of the work in my mind last night. Then I decide to try this morning. Finally, done!
Mmmmm... have the burn-in and trial on the performance there, so the review have to delay a little bit!!!Attachments
vulture said:
dear all,
One could ofcourse buy 3 clocs, there are several option, mine, Elso's, others
I will however come up with a decent clock upgrade for the 963, this is needed because
- The clock architecture of the 963 is a disaster
- The clock distribution takes many gates (yuck....)
- More clocks are present, and for CD / SACD onl 2 need replacement
In addition I plan a linear power supply as a drop in replacement for the very noisy SMPS
A similar supply for SACD1000 will be available soon as well
all the best
Guido Tent said:
Guido,
This is very good news!
I am sure that with your experience on clocks, we can make the 963 much better.
So you think that the other clock (for upsampling) needs replacement as well. This seems highly logical to me and I was surprised that all the expensive modding firms on the web seem to ignore this.
You plan to build an extra linear powersupply. There is enough room in the enclosure. Would you only feed the analogue output circuit with that, or also the the digital circuit?
BTW, the SACD 1000 does have a seperate linear PSU for its analogue circuit. However, some people responsible for the aquisition of parts at Philips seem to have been pennywise. After the exchange of the standard diodes in the SACD 1000 (by the Dutch firm van Medevoort) this player sounded incredibly better. Its previously dull sound is really gorgeous now. (They also did the MKP bycaps, but I think that this diode upgrade was the real thing.)
Please keep us informed during the various stages of your development.
Regards,
Lucas.
About the stacking of the DACs:
I found info on the AD1955:
http://www.analog.com/UploadedFiles/Data_Sheets/422968845AD1955_0.pdf
However, I could not find info on putting them in parallel.
I know that for instance Arcam chooses to put 2 or more Wolfssons in paralell (per channel) in their more expensive models.
Are there no disadvantages either? What about impedances, powerconsumption, etc..?
Did anyone do listening tests on this DAC stacking?
Regards,
Lucas.
I found info on the AD1955:
http://www.analog.com/UploadedFiles/Data_Sheets/422968845AD1955_0.pdf
However, I could not find info on putting them in parallel.
I know that for instance Arcam chooses to put 2 or more Wolfssons in paralell (per channel) in their more expensive models.
Are there no disadvantages either? What about impedances, powerconsumption, etc..?
Did anyone do listening tests on this DAC stacking?
Regards,
Lucas.
There is also some interesting info here:
http://www.analog.com/Analog_Root/productPage/productHome/0,2121,AD1896,00.html
AD clearly states that the 1896 is some 17 dB quieter than the 1895. They also write why 140 dB should be audible in another pdf:
http://www.analog.com/Analog_Root/static/pdf/dataConverters/140db.pdf
At ADs site they mention a price of $ 12.50
Worth a try?
Lucas
http://www.analog.com/Analog_Root/productPage/productHome/0,2121,AD1896,00.html
AD clearly states that the 1896 is some 17 dB quieter than the 1895. They also write why 140 dB should be audible in another pdf:
http://www.analog.com/Analog_Root/static/pdf/dataConverters/140db.pdf
At ADs site they mention a price of $ 12.50
Worth a try?
Lucas
I got the AD1986 directly from the analog devices site, compared to the AD1985 it appears to have better snr (and costs twice as much!). It is pin compatible with the AD1985 with the exception of pin one that controls the long/short group delay, I think the pin is pulled up internally so it should just go as a direct replacement for the AD1985.
I did some more listening tests comparing using the 963sa as a transport and an upsampler (against the audiolab 8000CD). The 963 is definitely less detailed and engaging, I think it should definitely benefit from a better clock circuitry and a power supply. At this stage I am not sure whether to keep it or send it back - the video section seems also worse than a Philips Q50 and a Philips 985 recorder (the latter two both have the previous generation FLI2200 deinterlacers). The picture of the 963 is grainy and has the green stripes on left (the latter due to a dodgy supply line to the deinterlacer chip) Oh well, this is an audio forum so I better not go on and on
I did some more listening tests comparing using the 963sa as a transport and an upsampler (against the audiolab 8000CD). The 963 is definitely less detailed and engaging, I think it should definitely benefit from a better clock circuitry and a power supply. At this stage I am not sure whether to keep it or send it back - the video section seems also worse than a Philips Q50 and a Philips 985 recorder (the latter two both have the previous generation FLI2200 deinterlacers). The picture of the 963 is grainy and has the green stripes on left (the latter due to a dodgy supply line to the deinterlacer chip) Oh well, this is an audio forum so I better not go on and on
TCXO
The TCXO's are from this Japanese firm, the MTX-1625 series:
http://www.mitadenpa.co.jp/tcxolist.htm
The site clearly shows this is a frequency deviation spec. No jitter specs are given but the jitter spec might be good. Jitter specs for can type oscillators usually are better for TCXO's.
Hi nForce,nFORCE said:
The TCXO's are from this Japanese firm, the MTX-1625 series:
http://www.mitadenpa.co.jp/tcxolist.htm
The site clearly shows this is a frequency deviation spec. No jitter specs are given but the jitter spec might be good. Jitter specs for can type oscillators usually are better for TCXO's.
http://www.avsforum.com/avs-vb/showthread.php?s=f597f9af48d0830f592f9bfc794b3b2f&thre
The video guys are now talking 12 months about the 963 and the 1646 messages are mostly very enthousiastic about the video quality.
But that is of course rather boring and not relevant because we love to hear music.
Guido Tent Hurry !!!!
The video guys are now talking 12 months about the 963 and the 1646 messages are mostly very enthousiastic about the video quality.
But that is of course rather boring and not relevant because we love to hear music.
Guido Tent Hurry !!!!
Another Dutch 963 owner mailed me his experiences:
"I too have had my 963 upgraded by van Medevoort, and I was also atonished by their choice on output caps. Each channel with two electrolitics bypassed by silvermica, as Mr. van Medevoort told me personally. He said that on listeningtest this was the best combination...
I removed them all and I have put BGs in place. This improved things a lot. Especially the treble is much better now.
In other words: van Medevoort made a dubious choice.
Furthermore, I exchanged some of the caps on the SPSU by better ones (Oscon). Particularly the bass has gained much of it (and this was very desirable).
Finally, I substituded the two 5532 for rear, center and sub-channels by 2 AD826.
I am considering the Tent-clock, but at this moment I am occupied with other things.
Maybe my experiences can be of any help to you."
"I too have had my 963 upgraded by van Medevoort, and I was also atonished by their choice on output caps. Each channel with two electrolitics bypassed by silvermica, as Mr. van Medevoort told me personally. He said that on listeningtest this was the best combination...
I removed them all and I have put BGs in place. This improved things a lot. Especially the treble is much better now.
In other words: van Medevoort made a dubious choice.
Furthermore, I exchanged some of the caps on the SPSU by better ones (Oscon). Particularly the bass has gained much of it (and this was very desirable).
Finally, I substituded the two 5532 for rear, center and sub-channels by 2 AD826.
I am considering the Tent-clock, but at this moment I am occupied with other things.
Maybe my experiences can be of any help to you."
Re: TCXO
Hi
98% of the off the shelf canned oscillators are no good for audio (my own experience)
Elso, why are TCXO's usually better ?
regards
Elso Kwak said:
Hi nForce,
The TCXO's are from this Japanese firm, the MTX-1625 series:
http://www.mitadenpa.co.jp/tcxolist.htm
The site clearly shows this is a frequency deviation spec. No jitter specs are given but the jitter spec might be good. Jitter specs for can type oscillators usually are better for TCXO's.
Hi
98% of the off the shelf canned oscillators are no good for audio (my own experience)
Elso, why are TCXO's usually better ?
regards