P1FI DAC, I2S, random audio clicks ?(SOLVED READ)

Raspberry Pi related support

Re: PI3 0.4 beta, I2S, random audio clicks ?

Postby TommyBoy » 26 Apr 2016, 23:38

Image

picture of the fixed PIFI 1.00
User avatar
TommyBoy
 
Posts: 111
Joined: 29 Mar 2016, 14:04
Location: Copenhagen Denmark

Re: PI3 0.4 beta, I2S, random audio clicks ?(SOLVED READ)

Postby rastus » 27 Apr 2016, 04:00

I was reading about PiFi DAC+ boards, and I read that the black boards with the silver RCA connectors are cheap copies (cheaper versions) of the PiFi. The original boards are blue with gold RCA's. This applies to the DAC/DAC+ versions and the DIGI/DIGI+ versions.

Special designed power source system by Geekroo, isolating DAC power source part from the Raspberry Pi power source system, keeps the current noise away, gives you crystal clear sound. Please check out the following resisters and capacitors that we used on this PiCobber - we choose the best quality components that are usually used on professional HiFi devices and that's where your good audio quality comes from:
- VISHAY resister (made in Czech)
- ELNA RBP2 capacitor (made in Japan)
- WIMA MKS2 capacitor (made in Germany)
- ERO KP1830 capacitor (made in Germany)
- Nichicon FG series capacitor (made in Japan)

Cheap Copy DAC+:
1. Cheap RCA connectors, affect the sound quality, easy to oxidized
2. Cheap non-RoHS PCB
3. Low cost capacitor, not audio electectrolitic capacitor
4. Low cost isolated power
5. 45 degree conner, high distortion rate

Geekroo Original High Quality DAC+:
1. High quality gold-clad RCA connectors,high sound quality
2. Whole board gold-clad process, anti-oxidation, high reliability
3. Nichicon MUSE fine gold(FG) MUSE HiFi audio electectrolitic capacitor
4. High quality Mornsun isolated power, UL, CE and RoHs certified. Output voltage stability, ripple noise small than 50mV(other bigger than 100mV).
5. Round corner layout, impedance is consistent, small reflection, low dissipation, signal is more smooth


Image
User avatar
rastus
 
Posts: 353
Joined: 21 Aug 2015, 10:29

Re: PI3 0.4 beta, I2S, random audio clicks ?(SOLVED READ)

Postby TommyBoy » 27 Apr 2016, 08:27

yes i know about the original and the copy :-)
however : you can NOT isolate anything here, the data signals need a way to the DAC,
if you isolate/brake ground, there will be a difference potential on the two halves,
this difference will be added to the data signals !!
big mistake in all designs you show with this switchmode supply.

however : it is a very good idea to use a local low noise linear regulator for the DAC, just as they did.
I am going to perform more EMI measurements today,
I expect to open the short in-out (the switchmode converter bypass) and insert a resistor, like 1 ohm
and then a better larger capacitor, this way the dac board can only draw a nice constant noise free dc current from the PI board,
this also affect the current in ground, then this will be exactly the same, and then the ferite bead can be mounted again.

PS: there is nothing wrong with any of the caps on the cheap copy dac

PPS: I am design engineer and work prof with signal integrity, EMI, hispeed and RF designs and microcontrollers.
User avatar
TommyBoy
 
Posts: 111
Joined: 29 Mar 2016, 14:04
Location: Copenhagen Denmark

Re: PI3 0.4 beta, I2S, random audio clicks ?(SOLVED READ)

Postby rastus » 27 Apr 2016, 08:58

I am happy to read your response, and look forward to reading about your further experimentation, thank you.

Maybe of interest: Hans Beekhuyzen will be testing some PSU's, should be online next month...
hansbeekhuyzen-t3526.html

I've been following this thread too...
Digi + - Linear Power and DAC Suggestions
https://support.hifiberry.com/hc/en-us/ ... uggestions
User avatar
rastus
 
Posts: 353
Joined: 21 Aug 2015, 10:29

Re: P1FI DAC, I2S, random audio clicks ?(SOLVED READ)

Postby TommyBoy » 27 Apr 2016, 10:07

yes i keep you updated
meanwhile tech reading:

http://www.ti.com/lit/an/szza009/szza009.pdf

page 10, bullet 2

SNIP :
"The ferrite bead is used only on +V, not on ground."

(fundamental design error also on original design, got copied, the copy use higher inductance value = the split ground problem get worse
and signals are jammed to a clearly hearable level)
User avatar
TommyBoy
 
Posts: 111
Joined: 29 Mar 2016, 14:04
Location: Copenhagen Denmark

Re: P1FI DAC, I2S, random audio clicks ?(SOLVED READ)

Postby rastus » 27 Apr 2016, 10:14

Thank you TommyBoy, reading now!

Edit-1: "The ferrite bead is used only on +V, not on ground. If a through-hole ferrite bead is used, it is mounted with the exposed lead connected to +V."

Edit-2: So they put it on the ground?!

Edit-3: "...the ground from PI is broken hi-frequency wise with a ferite bead called R8 on pifi 1.00 this means the ripple currents from the switchmode converter, generate voltages peaks over the ferite, and since this is ground split, it is perfectly the same as adding this 1V noise directly to clock data latch signals !!"
User avatar
rastus
 
Posts: 353
Joined: 21 Aug 2015, 10:29

Re: P1FI DAC, I2S, random audio clicks ?(SOLVED READ)

Postby TommyBoy » 27 Apr 2016, 14:06

YES, I am so sorry, you got it right, this is exactly what they did !!
I dont have any of the "original" P1FI boards, so I can not tell you if it is same place the ferite is located,
please have a look on yours, or post hi res closeups of this detail area.

WHY any of those designs use a switchmode.. yes you heard it..
A SWITCHMODE !! to isolate ?? to remove noise ?? 5V in, makes 5V out with tons of noise on both sides.
is just so far out. a switchmode newer ever remove noise, they ALWAYS add noise.
anyone own any of the pictured units ?
I dare you : remove the mornsun, short its in-out, and short the ferite, and listen..
NOW you get nice clean audio.
User avatar
TommyBoy
 
Posts: 111
Joined: 29 Mar 2016, 14:04
Location: Copenhagen Denmark

Re: P1FI DAC, I2S, random audio clicks ?(SOLVED READ)

Postby rastus » 27 Apr 2016, 14:16

I own the PiFi DAC+ v2.0, blue PCB pictured above.... I will take pictures tomorrow, it is very late here now. I am too tired.

"However, if there is a lot of noise on this point, the capacitor could couple that noise back onto the +V line."

This is what you mean, yes?
Attachments
Untitled.jpg
Untitled.jpg (51.89 KiB) Viewed 5710 times
User avatar
rastus
 
Posts: 353
Joined: 21 Aug 2015, 10:29

Re: P1FI DAC, I2S, random audio clicks ?(SOLVED READ)

Postby TommyBoy » 27 Apr 2016, 19:57

Image

here are all my error fixes and improvements together:
M1 switchmode module REMOVE, under it, add 1 ohm resistor, in-out
M1 output pins, ADD 220uF 10V electrolyte (local 5V input to the 3.3V regulator is now super clean)
R8 FERITE BEAD, ground problem, remove, ADD 0 ohm WIRE instead
R21 for IR-remote , remove no need to feed clean voltage to stuff we dont need.
B1 ferite bead smd in series with 3.3v regulator, remove, add 0 ohm instead, the regulator perform better when it can "see" the output.
C1 smd cap input 5V side to ground plane !! REMOVE IT, no need to feed input side noise directly to our nice audio ground.
C20 tantal cap, remove, useless part, all they give is bad MTBF
C9 10uF cap left side, remove, add 220uF 10V instead.

more info : the 1 ohm resistor in series with input, drop is 37mV so we know the whole board draw 37mA
User avatar
TommyBoy
 
Posts: 111
Joined: 29 Mar 2016, 14:04
Location: Copenhagen Denmark

Re: P1FI DAC, I2S, random audio clicks ?(SOLVED READ)

Postby TommyBoy » 27 Apr 2016, 20:25

Image
all noise on 5V input, go directly to ground plane of DAC board, AND all noise from DC/DC module in-out difference, is added directly to digital input signals due to R8 ferite bead !! VERY BAD DESIGN.. but easy to repair.

Image
any noise from 5V input, can not go into the ground plane of DAC board.
User avatar
TommyBoy
 
Posts: 111
Joined: 29 Mar 2016, 14:04
Location: Copenhagen Denmark

support RuneAudio Donate with PayPal

PreviousNext

Return to Raspberry Pi

Who is online

Users browsing this forum: Bing [Bot] and 8 guests