Build thread for ES9023 + JG Buffer boards (betatest) - Page 13 - diyAudio
Go Back   Home > Forums > Source & Line > Digital Line Level

Digital Line Level DACs, Digital Crossovers, Equalizers, etc.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 5th June 2013, 04:14 PM   #121
Mull3t is offline Mull3t  United States
diyAudio Member
 
Join Date: Apr 2012
Location: Gamehendge
Yeah, the more I think about it I think I'll go with a barrier terminal to split the power supply to multiple components. On another note, this decoupling thing interests me.

If I were to go the DV+ route, I would be hooking this up to the Acko Isolator/Re-clocker board which has 5v in that goes to a ADP150 Reg then to a ferrite bead. For decoupling, does the FB need to come before the Reg or does it not matter? By the way, what is the purpose of the decoupling in this application?

I thought of a few more questions regarding C702 and C202. I'm considering using these caps as alternatives to C701 and C201. What are the benefits of the Wima caps? I'm thinking of using these caps... C702 -- 505-MKS2.1/63/5 and C202 -- 505-MKS21.0/63/5

Last edited by Mull3t; 5th June 2013 at 04:24 PM.
  Reply With Quote
Old 6th June 2013, 10:02 AM   #122
diyAudio Member
 
Join Date: Aug 2010
The regs are quite slow (e.g. compared to the signals/rise time of XO) and work just "like a trace" at HF in both directions. Therefore it makes sense to decouple the different parts of a mixed signal circuit by using inductors. Ferrite beads (FB) also include a resistance that reduces the risk of peaks (added damping to LC -> LCR). They also have a (quite small) DC resistance. Thus, as far as I have learned, it makes sense to place the FB in front of the reg (and in front of the local reservoir cap) if the reg only supplies one IC (XO, DAC,...). This way you "clean" the incomming DC, prevent crosstalk of high frequency noise to other parts and keep the DC resistance low. Whenever you supply several ICs with one reg (The flipflops in my DAC) it makes sense to decouple each single IC with a FB after the reg.

The Wimas are used to reflect an alternative case. To be honest I didn't have the time to do own experiments and used Elcaps (Pana FM or oscons) until now. Jean-Paul used BC components/Vishay solid electrolyte MnO2 caps with good results as it seems and those would also fit here The Wimas might be a bit small (1F), but I didn't try.
  Reply With Quote
Old 6th June 2013, 11:41 AM   #123
qusp is offline qusp  Australia
diyAudio Member
 
qusp's Avatar
 
Join Date: Oct 2009
Location: Brisbane, Australia
does nothing for common mode though

the majority of radiated noise and ripple will be common mode too
  Reply With Quote
Old 6th June 2013, 11:58 AM   #124
diyAudio Member
 
Join Date: Feb 2009
Location: Brisbane, Australia
I think at least part of the point of the series L is to reduce the charging current in the power supply lines to reduce the amount that those traces radiate, by slowing them down so that they are still able to recharge the local caps for the reg but not allow the short current pulses in the PSU circuit that would otherwise create noise. Perhaps not so much about filtering noise that is picked up in the circuit from it's environment.
  Reply With Quote
Old 6th June 2013, 12:13 PM   #125
diyAudio Member
 
Join Date: Aug 2010
I am using common mode chokes in the power supply. Haven't seen them widely used on PCBs though (for Audio gear we are talking about here).

Small SMD types would be available e.g. from murata.
  Reply With Quote
Old 6th June 2013, 12:29 PM   #126
diyAudio Member
 
Join Date: Feb 2009
Location: Brisbane, Australia
Don't get me wrong, you're preaching to the converted about common mode chokes here. I'm just saying there is other reasons to have a series ferite bead.
  Reply With Quote
Old 6th June 2013, 01:03 PM   #127
diyAudio Member
 
Join Date: Aug 2010
Quote:
Originally Posted by hochopeper View Post
Don't get me wrong, you're preaching to the converted about common mode chokes here. I'm just saying there is other reasons to have a series ferite bead.
Oh, didn't know that there is obviously a controversal discussion going on Maybe one should just take both, common and differential mode noise into acount

Maybe qusp can guide to some practical solutions...
  Reply With Quote
Old 6th June 2013, 01:58 PM   #128
qusp is offline qusp  Australia
diyAudio Member
 
qusp's Avatar
 
Join Date: Oct 2009
Location: Brisbane, Australia
Quote:
Originally Posted by hochopeper View Post
I think at least part of the point of the series L is to reduce the charging current in the power supply lines to reduce the amount that those traces radiate, by slowing them down so that they are still able to recharge the local caps for the reg but not allow the short current pulses in the PSU circuit that would otherwise create noise. Perhaps not so much about filtering noise that is picked up in the circuit from it's environment.
the amount of attenuation is not so amazing really

Quote:
Originally Posted by curryman View Post
I am using common mode chokes in the power supply. Haven't seen them widely used on PCBs though (for Audio gear we are talking about here).

Small SMD types would be available e.g. from murata.
there is one on every low noise power supply I make and liberally scattered through my DAC and ALL over Ackos DACs, always has been since the first version a few years ago; every reg on his board has one. they are all over Ians fifo stuff too. I think maybe you arent looking hard enough

on Ackos regs, the prereg (linear) has CM chokes on it, then the final reg input is SMD CM choke, followed by matched low value resistors, regulator, followed by CRCLCC on every reg; there is ~20-25 CM chokes in a stereo AKD12/18P build. in my build each has its own secondary or battery and every consumer has its own regulator.

but even those (CM chokes) can be somewhat limited in their attenuation by themselves and a bit scattergun unless you know the source and load impedance, as well as exactly what frequency bands you are targeting. the composite filters require matched caps, matched inductors (or CM chokes) matched resistors etc and youre flying blind unless you have real information to act on. just choosing to target the known busy GSM and wifi bands means the Q on the filter isnt going to be able to be that steep, or deep and the response is moved around by everything on the board and every change can have flow on.

there are no real rules of thumb, general practical solutions, its a case by case thing and sims wont get you there either. black magic

depends on what level of noise you are targeting, with the 9023 you have a much easier task than 9012/8

Last edited by qusp; 6th June 2013 at 02:08 PM.
  Reply With Quote
Old 7th June 2013, 12:11 AM   #129
diyAudio Member
 
Join Date: Feb 2009
Location: Brisbane, Australia
Quote:
Originally Posted by qusp View Post
the amount of attenuation is not so amazing really
Have you got some measurements on RF radiation vs regulator to decoupling cap di/dt? The stuff I'm talking about is probably only important on fast digital stuff, flip flops, clock drivers etc.

Last edited by hochopeper; 7th June 2013 at 12:19 AM.
  Reply With Quote
Old 7th June 2013, 05:18 AM   #130
Mull3t is offline Mull3t  United States
diyAudio Member
 
Join Date: Apr 2012
Location: Gamehendge
Quote:
Originally Posted by curryman View Post
The regs are quite slow (e.g. compared to the signals/rise time of XO) and work just "like a trace" at HF in both directions. Therefore it makes sense to decouple the different parts of a mixed signal circuit by using inductors. Ferrite beads (FB) also include a resistance that reduces the risk of peaks (added damping to LC -> LCR). They also have a (quite small) DC resistance. Thus, as far as I have learned, it makes sense to place the FB in front of the reg (and in front of the local reservoir cap) if the reg only supplies one IC (XO, DAC,...). This way you "clean" the incomming DC, prevent crosstalk of high frequency noise to other parts and keep the DC resistance low. Whenever you supply several ICs with one reg (The flipflops in my DAC) it makes sense to decouple each single IC with a FB after the reg.
Ok, makes sense. Glad I could spark up some debate. I know we're getting into the realm of another board, but in the spirit of DIY could I stack a FB on top of another cap, which is right before the Reg? Something like this might do the trick... 963-BK0603TS601-T. However, as mentioned it might have to be in series and not parallel.

Then again, splitting the supply before hand is more practical.

Quote:
The Wimas are used to reflect an alternative case. To be honest I didn't have the time to do own experiments and used Elcaps (Pana FM or oscons) until now. Jean-Paul used BC components/Vishay solid electrolyte MnO2 caps with good results as it seems and those would also fit here The Wimas might be a bit small (1F), but I didn't try.
It sounds like this would be quite the experiment. Did I read your schematic right? It does say 100nF for C702 and 1uF for C202. For now, it does seem Elcaps are the way to go.

Last edited by Mull3t; 7th June 2013 at 05:21 AM.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Fastron Inductor Group Buy for JG Filter Buffer pchw Group Buys 176 8th February 2013 09:19 PM
ES9023 DAC build heartwinter Vendor's Bazaar 8 7th July 2012 09:27 AM
Preamp / buffer suggestions for TA2020 / TA2024 boards lossfound Class D 11 29th January 2012 10:35 AM
ES9023 and buffer before 10K pot. Neutrality Digital Line Level 1 12th December 2011 07:00 PM
DIY Op Amp Buffer Boards aziltz Swap Meet 20 18th October 2011 12:35 AM


New To Site? Need Help?

All times are GMT. The time now is 10:23 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2