| till |
hi,
2 questions
Is WIMA MKS-2 100nF ok for the 14 decoupling caps of 1541,
and why is on some schematics pin 13 and 18 decoupled with larger caps (some uF) ? |
|
|
| jean-paul |
Those are the MSB pins. Arcam used tantalum caps for that but I sure would not recommend those unreliable caps.
Wima MKS is OK ( no more than that ) but 220 nF instead of 100 nF is slightly better. Avoid long wires to the caps ( bigger is not better here ). Peter Daniel tried BG NX Hi Q there with good results IIRC. I never tried them on that spot.
Discussions about the value can last long, some will propose 1 uF. Better try it and choose yourself what you like best. Datasheet says 100 nF so that should be enough. Quality of the caps has more priority than the 100 nF/220 nF/ 1 uF question IMO. Polypropylene ( Wima MKP 10, 100 nF or 220 nF ) caps are a wise choice considering cost/quality. |
|
|
| Bricolo |
But the size is also a concern, since you have to place 7 caps per side.
I wanted polypropylene too, but the 1uF ones were way too big.
I'm on the way to buy some wima MKS-2 too
PS: jean-paul, are you sure about the location of the MBS pins? Elso and Thorsten don't agree about that |
|
|
| siu sin man tho |
I test many decoupling caps in the 1541a. I feel tin foil is quite good.
this is my opinion.
thanks
thomas |
|
|
| jean-paul |
Just checked for you and in various schematics of brands that bought their stuff and technology from Philips pins 13 and 18 have the label MSB. For instance Arcam used 0.47 uF on that pins. I used the same value on all pins like the datasheet says.
Of course size is a concern as I tried to make clear. 220 nF MKP will fit, I never thought of using 1 uF. |
|
|
| mdlover |
Hi all,
I have tried Philips MKP (130nF) and Siemens stacked film (470nF) for those decoupling caps. Both sound smooth and natural. Personally prefer Siemens stacked film more (may be the value?)
;)
Will try BlackGate N, Nx soon.
Below pics showing the board using Philips MKP. :)
pics for Siemens stacked film will follow. |
|
|
| jean-paul |
| Nice board ! Wish I got the design around the TDA right. |
|
|
| Bricolo |
The wimas MKP-2 0.1uF are the same size as the MKS-2 1uF
So, I will be able to try, when I'll be rich :D |
|
|
| Bricolo |
| quote: | Originally posted by jean-paul
Just checked for you and in various schematics of brands that bought their stuff and technology from Philips pins 13 and 18 have the label MSB. For instance Arcam used 0.47 uF on that pins. I used the same value on all pins like the datasheet says.
Of course size is a concern as I tried to make clear. 220 nF MKP will fit, I never thought of using 1 uF. |
so Thorsten was wrong
http://www.fortunecity.com/rivendel...gio/Adagio.html |
|
|
| jean-paul |
I think that's an error in the drawing. It seems to me that the creator of the chip knows what pin is MSB. When the same value for all pins is used we won't need to discuss about that ;)
BTW I have an amount of MKP10 100 nF and they were very cheap. No need to be rich for buying MKP10. They're small too. They may not be the best but their price/quality ratio is hard to beat.
Who said Mundorf ? |
|
|
| weissi |
Why not use NPO ceramics???? these are high frequency decoupling points, so ceramic won´t hurt. Not those 4cm MKP´s...
I wasn´t able to hear a difference when I changed the ceramic C´s in my CD50 to SMD polypropylensulfids......
OK, NPO 220nF cost much, but IMHO they should be way better than some of that monster caps....
:angel:
I used these cap´s:
http://www.panasonic.com/industrial...ABD0000CE13.pdf |
|
|
| zygibajt |
>Bricolo
Don't use Wima,they are not that good.
The best cheap caps i found for that purpose (much better than Wima) were small orange Philips polypropylene 1uF/63.They are really small ,about the size of op-amp and they sound good.
my friend prefers small 470n Rifa over Philips ,says they are better,but I haven't tried this yet.Bipolar Black Gate might be better still,and they are not that expensive either.
Bartek |
|
|
| mdlover |
here is the board using Siemens decoupling caps.
now i'm thinking of using BG-N 0.1u or BG-NX 0.47u and wonder how they sound. :)
cheers |
|
|
| Bernhard |
| quote: | Originally posted by jean-paul
I think that's an error in the drawing. It seems to me that the creator of the chip knows what pin is MSB. When the same value for all pins is used we won't need to discuss about that ;)
|
How wise...
I'm wise too.
Because it seems to me that the creator of a pair of chips knows which one is the economy - low performance and which one the high performance.
But lots of DIYers question that.
| quote: | Originally posted by till
as i need ca. 50 of them for this |
4 x 1541 ?
I have tin foil 0,15µ, but they are 15mm diameter :( |
|
|
| Bricolo |
Would BC caps (philips) BC371 be better than WIMAs MKS?
Both are polyester |
|
|
| Bernhard |
IMHO it is very important that the rolled film caps have contact material all over the flat end faces and the wires bonded to them.
So rolled MKP caps will be low inductive like the stacked Siemens MKTs.
But better because they are MKP.
If the wires are just sticked into like old or cheap styroflex, they will be inductive and also the electrical contact will be not so good.
15mm tin foil caps: |
|
|
| Bricolo |
Isn't it better to place the caps this way, rather than to align them?
The path seems shorter with this configuration |
|
|
| Bernhard |
| quote: | Originally posted by Bricolo
Isn't it better to place the caps this way, rather than to align them?
The path seems shorter with this configuration |
No.
Foil caps are mostly long.
And this way the path is prolonged by the lenght of the cap.
Only useful if you want to use very fat 1µ Siemens MKT stacked foil caps.
They are short. |
|
|
| hifi |
boys... use the bottom side of the board to mount half the caps... :)
/micke |
|
|
| Bernhard |
| quote: | Originally posted by hifi
boys... use the bottom side of the board to mount half the caps... :)
/micke |
That saves 5mm in my case. Very good idea. |
|
|
| Bricolo |
| quote: | Originally posted by Bernhard
No.
Foil caps are mostly long.
And this way the path is prolonged by the lenght of the cap.
Only useful if you want to use very fat 1µ Siemens MKT stacked foil caps.
They are short. |
No, there are short caps that will fit (like the one on my drawing)
Wima MKS-2, MKP-2
Siemens MKT
| quote: | Originally posted by hifi
boys... use the bottom side of the board to mount half the caps... :)
/micke |
And how will you solder them? I mean, with radial caps. |
|
|
| Bernhard |
| quote: | Originally posted by Bricolo
No, there are short caps that will fit (like the one on my drawing)
Wima MKS-2, MKP-2
Siemens MKT
And how will you solder them? I mean, with radial caps. |
Wima are not round.
Round is better.
Siemens MKT is MKT.
MKP is better. |
|
|
| Bricolo |
| Do you know 1uF MKPs for less than 0.5€? |
|
|
| weissi |
is this a contest of who makes the finest looking alignment??
Use SMD Part, you can wire it under the socket, you can choose betwen foil or ceramic and inductivity is a very minimum...
Why do you spend so much time on this, no one would use these fat caps that some of you use to decouple the supply pins!
And the 14 decoupling pins aren´t anything other than that!
Correct me if I´m totaly wrong...
Markus |
|
|
| till |
I don´t really understand all of the esoteric caps discussion, its the same like listening to different "resistor sounds". The most difference of sound i experience from different recordings, not from changing caps, resistors or cables. And tuning the value of for example passive I/V resistor of the DAC changes the performance in a measurable way. I doubt there is any difference in sound caused by the brand of the resistors used. I´m sure on this board the brand would be the more important topic of discussion than the value. And the best "performance" of corse would be reached with a part unobtainable not manufactured anymore. The second best and only acceptable would be 20 times more expensive as a standart parts.
50 i need for 2 DACs, balanced. That would be 28 for the decoupling pins and the others for decoupling the different supply pins of all ICs used. I´ll go for 28 * MKS 100nf, and all supply pins also 100nf MKS + standart electrolytic. I don´t want to spend 80€ on fancy caps for this project. |
|
|
| Fin |
| If the choice is between Wima MKS2 220nF and Wima MKP2 100nF.........which will give the better results? |
|
|
| JonHarrison |
Hi,
Can someone explain why Oscons are used for the MSB caps ?
100-220nF and 1-2.2uF for the MSB seem to be typical values so a good polyprop foil cap would seem a sensible choice. However, I have seen people using 10-15uF Oscons.
I believe that you want a cap with a very low leakage current. What benefit does an Oscon give in this application ???
Jon |
|
|
| Archwn |
i think MSB should need more current than the others as it is the base of all wave form.
my friend was trying to use os-con at the other pin and he could hear the difference from the first three MSB
feel free to correct me if i m wrong. |
|
|
| JonHarrison |
Hi,
Yes it is correct that the MSB caps should be larger but why an Oscon 'lytic rather than film / foil ?
Cheers,
Jon |
|
|
| Hartono |
large capacitor will make the loop longer (long lead length = more inductance)
which diminish it's benefit at higher frequency |
|
|
| Thor66 |
Here's one way of doing it. The caps are 220nF PP. The originals were
multilayer cheramics, so changing them seemed sensible. |
|
|
| Hartono |
Hi Thor66,
The caps path to ground is way too long for my liking.
what is that copper box on the right ? clock ?
Cheers,
Hartono |
|
|
| Thor66 |
Yes, it is a rather long path to ground, but only slightly longer than the original pcb-tracks.
It's a quick and easy way of doing it when there's little space underneath the pcb.
A shield between the digital and analog sections is a good thing when using large foil caps.
Together with a standard CDP-clock comletely cast in silicone rubber and better OPs in the output, the result was amazing for a weekends modification-work. So keep on tweaking ! |
|
|
| dcathro |
| Did anyone ever try the Black Gate bgnx 0.1uf caps? |
|
|
| Hartono |
I used to sell 0.47uF NX to my customer for decoupling TDA1541A. They seems to be happy , I don't know about the actual improvement myself in that application. This might be off topic, get double crown version if you can, sounds MUCH better than single crown. Where did you get Blackgate in Melbourne ?? now Jaycar stock them ? ;)
Hartono |
|
|
| Bricolo |
why is no one using ceramics here?
the decoupling for the 1541 has to be:
-low impedance
-low resistance
-no noise from leakage
ceramics fit, don't they? |
|
|
| Hartono |
Hi Bricollo,
why is no one using ceramics here?
seriously ?
Hartono |
|
|
| Bricolo |
yes, everyone is raving about film caps or high quality electrolytics
what about C0G ceramics? |
|
|
| Hartono |
where to buy 0.1 uF :hypno2: :hypno2: :hypno2:
btw, me not using TDA1541A |
|
|
| Bricolo |
| seems that 33nF is about the maximum, let's take 3 |
|
|
| Hartono |
"seems that 33nF is about the maximum, let's take 3
no hole to put them on the PCB :D |
|
|
| Bricolo |
0.33 is the max I found for SMD
for leaded maybe they can be bigger
and you can stack smd ;) |
|
|
| Hartono |
"0.33 is the max I found for SMD
for leaded maybe they can be bigger"
well for starter leaded 0.1uF C0G is expensive, and hard to obtain , at least locally.
"and you can stack smd"
not a good idea, unless done correctly, for example how to clean the flux ?;)
Hartono |
|
|
| Yomanze |
My schematic shows that TDA1541A's supply pins are decoupled with 33u caps bypassed with 22n ceramic caps.
Would it be a good idea to try 82uf Pana FC or 100uf Silmic caps bypassed with 0.082uf metallised polypropylene caps and do the same for SAA7220 decoupling? Have got some lying around :) Or... would it be better to leave the 22n bypass values, of which I have some metallised polyester caps.
I don't have os-cons available. |
|
|
| oshifis |
I bring this topic up. I can confirm that pin 13 is MSB for R channel, and pin 18 is MSB for the L channel (if the TDA1541A is driven with I2S signal). I tried 22uF tantalum capacitors on pins 12, 13, 18 and 19. The result was catastrophal as I described it in this thread on the same subject:
http://www.diyaudio.com/forums/show...540#post1459540
I was also thinking over the operation and found that the MSB is responsible for the positive half period of the analog output signal only. It has no effect on the negative half period. |
|
|
| John007 |
| Does decoupling TDA1541 makes difference in sound? Can i tailore sound changing caps? |
|
|
| oshifis |
| I can not tell that because I modified also other things. I will try some suitable size MKP capacitors. |
|
|
| Archwn |
| 0.1 / 0.22uF c0g/np0 ceramics are alot cheaper now. I wish to try them. :D |
|
|
|
