Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

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

Ground Pours on a DAC PCB.
Ground Pours on a DAC PCB.
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 25th December 2018, 06:51 AM   #21
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
Quote:
Originally Posted by deusex3 View Post
This is the must for the RF circuits only
[Markw4=rayma;]Agreed, one ground plane can be fine.[/Markw4]
There are may be occur additional interferences from DGND to AGND in this case
Basically, this is an RF circuit. The USB interface inside the PCM2912A uses an internal clock of 96 MHz and will generate lots of harmonics of 96 MHz. The built-in ADC and DACs are sigma-delta converters with 64 times oversampling. Sigma-delta converters are sensitive to interference at all frequencies except exact multiples of their clocks.

With a single-ground-plane design, one would typically complete each supply-decoupling cap-ground loop on the top layer and connect it to the ground plane at only one place. That way the current loops are kept local as much as possible. When there is a common supply for interfering and sensitive circuits, add ferrite beads in the supply connections to keep them separated.

In this case ( http://www.ti.com/lit/ds/symlink/pcm2912a.pdf pages 15 and 24), there is a USB supply that supplies lots of internal 3.3 V regulators for the various circuits, so these regulators handle the separation. These internal regulators have external decoupling.

Last edited by MarcelvdG; 25th December 2018 at 06:57 AM.
  Reply With Quote
Old 25th December 2018, 08:59 AM   #22
Thorp is offline Thorp
diyAudio Member
 
Join Date: Jan 2013
Quote:
Originally Posted by deusex3 View Post
I have mean what it is more relevant for RF , for 50 MHz this is much less important

Is it new design technique la "star ground" by Thorp ?
Actually it is a star ground la AKM. Draw some colored lines on it, too.
Attached Images
File Type: png akm.png (253.2 KB, 97 views)
  Reply With Quote
Old 25th December 2018, 07:26 PM   #23
deusex3 is offline deusex3  Russian Federation
diyAudio Member
 
Join Date: Dec 2011
Location: Russia
Quote:
Originally Posted by Thorp View Post
Actually it is a star ground la AKM. Draw some colored lines on it, too.
You may see the difference regarding your board .
Quote:
Basically, this is an RF circuit.
At RF i have mean >200MHz
Quote:
The USB interface inside the PCM2912A uses an internal clock of 96 MHz and will generate lots of harmonics of 96 MHz.
How that 96MHz correlating with actual working frequency of the Sigma-Delta (which should be max 256*48=12.288MHz)?
Quote:
With a single-ground-plane design, one would typically complete each supply-decoupling cap-ground loop on the top layer and connect it to the ground plane at only one place.That way the current loops are kept local as much as possible.
In this realization "supply-decoupling" should be very good to prevent undesirable returning current flow (to the input's capacitors)
Attached Images
File Type: png akm.png (67.3 KB, 85 views)

Last edited by deusex3; 25th December 2018 at 07:37 PM.
  Reply With Quote
Old 25th December 2018, 09:12 PM   #24
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
Quote:
Originally Posted by deusex3 View Post
At RF i have mean >200MHz
RF stands for radio frequency. There are many radio transmitters working well below 200 MHz. Besides the harmonics generated by the USB interface will extend well above 200 MHz.

Quote:
Originally Posted by deusex3 View Post
How that 96MHz correlating with actual working frequency of the Sigma-Delta (which should be max 256*48=12.288MHz)?
Actually the sigma-delta is 64 times oversampled, so the clock frequency is at most 3.072 MHz.

When it is 3.072 MHz, the spectrum of the digital sigma-delta modulate repeats every 3.072 MHz. The 31st image will then have a signal band centred around 95.232 MHz. Around 96 MHz, there will be strong out-of-band quantization noise, the 31st image of the out-of-band quantization noise around 96 MHz - 95.232 MHz = 768 kHz.

96 MHz interference on the reference or the clock of the DAC will frequency-convert this noise to low frequencies. A DAC by definition multiplies its digital input signal with its reference, and as I'm sure you know, multiplying mixers are good frequency converters.

The 62nd image may be worse than the 31st. Sigma-delta modulates often have a strong idle tone around odd multiples of half the clock rate and 62.5 times 3.072 MHz is 192 MHz, precisely the second harmonic of the USB clock. As long as the idle tone is converted exactly to 0 Hz no-one will hear it, but it will be frequency-modulated by the input signal.

Quote:
Originally Posted by deusex3 View Post
In this realization "supply-decoupling" should be very good to prevent undesirable returning current flow (to the input's capacitors)
Indeed it should, but I'd be mainly worried about the signals that cross the border between digital and analogue and anything that can affect the reference voltage.

Last edited by MarcelvdG; 25th December 2018 at 09:23 PM.
  Reply With Quote
Old 25th December 2018, 09:26 PM   #25
Markw4 is offline Markw4  United States
diyAudio Member
 
Join Date: Jul 2016
Location: California
Radio frequency - Wikipedia
"Radio frequency (RF) refers to an oscillation rate of an alternating electric current or voltage or of a magnetic, electric or electromagnetic field or mechanical system in the frequency range from around twenty thousand times per second (20 kHz) to around three hundred billion times per second (300 GHz). This is roughly between the upper limit of audio frequencies and the lower limit of infrared frequencies."

Last edited by Markw4; 25th December 2018 at 09:28 PM.
  Reply With Quote
Old 25th December 2018, 09:32 PM   #26
rayma is offline rayma  United States
diyAudio Member
 
Join Date: Apr 2011
Quote:
Originally Posted by MarcelvdG View Post
RF stands for radio frequency. There are many radio transmitters working well below 200 MHz.
Even the RF range from 3Hz to 30kHz is used for submarine communication.
Our AM broadcast band begins at 535kHz, though there is also much use below that.
  Reply With Quote
Old 25th December 2018, 09:43 PM   #27
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
Don't you have long wave AM broadcasting in the USA? Down here medium wave starts at 522 kHz and long wave at 144 kHz, if I'm not mistaken. There are no Dutch national stations in either band anymore, but there are still some British, French and local Dutch stations.
  Reply With Quote
Old 8th February 2019, 04:16 AM   #28
deusex3 is offline deusex3  Russian Federation
diyAudio Member
 
Join Date: Dec 2011
Location: Russia
Quote:
Originally Posted by MarcelvdG View Post
With a single-ground-plane design, one would typically complete each supply-decoupling cap-ground loop on the top layer and connect it to the ground plane at only one place. That way the current loops are kept local as much as possible.
Do you have a notion about sufficiency of "supply-decoupling cap-ground loop" caps values ? Is it enough 0.1uF ? Or maybe 2200uf?
Quote:
Originally Posted by MarcelvdG View Post
When there is a common supply for interfering and sensitive circuits, add ferrite beads in the supply connections to keep them separated.
The same , do you have notion about sufficiency of values of ferrite beads to separate clean AVDD from noisy DVDD? Is it enough 10uH Or it's needed 1000uH Or it's needed couple of them with different values?
  Reply With Quote
Old 8th February 2019, 05:20 AM   #29
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
Good question... It depends on what frequencies you are afraid of.

For a sigma-delta-modulator-based DAC that would typically be frequencies well above the audio band, especially near odd multiples of half the sigma-delta clock rate, because those can mix out-of-band quantization noise into the audio band. A BLM18RK102 (high-loss SMD ferrite bead with an inductance of about 6 uH) with a 10 uF X5R capacitor should then be more than sufficient. The resonance of this combination is at the upper edge of the audio band, but gets damped reasonably well by the DC resistance of the bead (Q ~= 1).

Last edited by MarcelvdG; 8th February 2019 at 05:40 AM. Reason: Incompleteness
  Reply With Quote

Reply


Ground Pours on a DAC PCB.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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Is it safe to do this? antenna ground tied to RCA input ground? realflow100 Analogue Source 2 11th December 2018 09:56 AM
Copper Pours Ripcord Construction Tips 13 15th February 2014 11:36 PM
Trimming polygon pours in Eagle, help needed SunRa Software Tools 2 27th March 2011 04:07 PM
Parallel LM3886, input ground, power ground, safety resistor and oscillations. markiemrboo Chip Amps 5 9th August 2007 03:50 PM
does connecting mains ground with circuit ground create a ground loop? jarthel Everything Else 0 25th June 2003 12:55 PM


New To Site? Need Help?

All times are GMT. The time now is 04:46 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki