CS3318 PCB Layout - Page 12 - diyAudio
Go Back   Home > Forums > Source & Line > Analog Line Level

Analog Line Level Preamplifiers , Passive Pre-amps, Crossovers, 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 11th February 2014, 09:13 AM   #111
maxw is offline maxw  United Kingdom
diyAudio Member
 
maxw's Avatar
 
Join Date: Jun 2004
Location: London
Linuxworks or Dr_EM, could you post RMAA test results for your CS3318 builds?
  Reply With Quote
Old 11th February 2014, 08:07 PM   #112
Dr_EM is offline Dr_EM  United Kingdom
diyAudio Member
 
Dr_EM's Avatar
 
Join Date: Oct 2006
Location: Swindon
Interesting discussion. I still plan to use a rotary myself, I've used a cheap one with a hardware de-bounce into the microcontroller and it's flawless as fast as you could turn it, so an optical should only be better! Could even make use of the 1/4dB resolution of this volume controller if I get one of the higher resolution EM14 types linked earlier.

I have performed some listening tests. I built a simple headphone buffer using an NE5532 per channel, so 2 op-amps in parallel, 56R current share resistors. This is still a very low noise op-amp and it worked well here. In testing as a buffer I could not hear any noise at any setting, a signal at full output would have been deafening but still no hiss. Setting the CS3318 to max attenuation, 96dB, and giving it full output from my soundcard the signal is still clearly audible although very quiet, so this shows it must be well above noise.

I then tried adding a 20dB gain stage after the CS3318 and before the headphone buffers. Using this some noise can be heard at full setting (input connected to ground with 100R) but it's just benign hiss. If I were to apply a signal at this setting it would destroy the headphones almost immediately, but again, winding the level down no noise is heard and music comes through clearly (probably putting a bit too much trust in the code at this point!).

So anyhow, it seems promising. 2 small issues; a clicking sound when making adjustments at the very top end. I strongly suspect this happens exclusively when GAIN is applied as I believe it is DC offset from the internal op-amp being shifted around. It occurs with the headphone stage AC coupled or DC coupled and of course the input is AC coupled so it can only really be an internal gain stage causing it. Certainly it wouldn't be audible with a signal passing at this level. To absolutely rule out the potential that it's digital noise passing into the analogue outputs I will be setting the Master 1 mask on all channels so that the exact same data can be passed to the Cirrus but without any volume changes. I don't expect to hear any clicks in this setup, even with gain fixed at +20dB, but we will see.

The other small issue is a 'shimmering' when making volume adjustments. I believe its a combination of being able to make very rapid adjustments with that single turn pot along with the slow SPI which takes longer to send data than the zero cross timeout period! Still want to get that SPI up to speed, pretty sure it's a code issue.

I can attempt an RMAA. My soundcard is 112dB SNR in loopback theoretically, but I measured 'only' 106dB with a 1m cable, so it might be very sensitive to noise pickup. In theory we shouldn't see any worsening of the quality by adding the volume controller in the loopback but I will try it out. I may be able to make use of an AP alternatively, in theory it's no better as it's an old System 1, but in practice I think the result is more accurate.
  Reply With Quote
Old 11th February 2014, 09:14 PM   #113
Dr_EM is offline Dr_EM  United Kingdom
diyAudio Member
 
Dr_EM's Avatar
 
Join Date: Oct 2006
Location: Swindon
I gave it a shot anyhow. I think it only really shows that the CS3318 board surpasses my measurement setup (which I only really expected to be good enough for power amps). Still good to see some numbers though!

A slightly better result should be obtained with the CS3318 all boxed up and shielded and using shorter, better shielded cabling. The drop from my previously posted 106dB SNR comes from now using 2 sets of 3.5mm jack-jack leads and short unshielded sections connecting to and from the Cirrus PCB. The test lead measurement posted here should be your reference point as it includes both leads. Cirrus is set to unity gain, default at power on.

EDIT: Ignore the '...' response, I put the cable right next to the bench PSU on that one by accident!

Attached Images
File Type: png CS3318 THD+N.png (39.8 KB, 149 views)
File Type: png CS3318 Noise level.png (40.2 KB, 148 views)
File Type: png CS3318 SNR.png (39.8 KB, 141 views)
File Type: png CS3318 IMD.png (40.9 KB, 142 views)
File Type: png CS3318 Numbers.png (73.1 KB, 124 views)
  Reply With Quote
Old 12th February 2014, 04:14 PM   #114
diyAudio Member
 
linuxworks's Avatar
 
Join Date: Jul 2008
Location: santa clara, CA
posting pic of my latest build (using relay atten and arduino; not the cs3318, though) - but I am still using the motor pot instead of a rotary:

Click the image to open in full size.

instead of lcd, I went with old school nixies. just for grins and giggles and to have something a bit different.

internally, its 2 pcf8574 i2c port expanders that I send 2 bytes to. the nixies are sitting on BCD encoder chips so I have to send just 4 4-bit nibbles to them to get them to show numbers. its very fast - much faster than updating an lcd.

not completely cs3318 related, but maybe it will give people ideas on alternative builds instead of always picking the same kind of input and output devices

(for those that are going to this weekend's bay area head-fi meet, I'll have this arduino-controlled beta22 phones amp with me).
__________________
My Photostream:http://www.flickr.com/photos/linux-works/
  Reply With Quote
Old 12th February 2014, 05:06 PM   #115
diyAudio Member
 
linuxworks's Avatar
 
Join Date: Jul 2008
Location: santa clara, CA
to see how the pot works, via twist and via IR remote, I have some video clips:

Index of /nixies

one file is using a handheld IR remote and the motor pot eventually showing the absolute value. I split out the faster and slower elements so that the nixies show the current volume value, while the motor scoots along as fast as he can, but he does that in the background. eventually he gets there and then rests and the value shown matches the display. when the user twists the knob, the new volume level is acted on immediately, the nixie display is updated immediately and the motor pot is asked to move to the new position (that is, if the vol change request came from IR). if the vol change came FROM the pot, the motor isn't going to fight you

you have 2 ways to change volume: so-called in-band and out-of-band. in-band might be the front knob and oob might be the IR. if you change the knob, the display updates and the IR remote does not need to know or care. otoh, if you use the IR remote, the display and vol engine update immediately and the motor pot is moved to show the new value. that way, if you go from the remote to the front panel, there is no sudden surprises, with one control being out of synch wrt the other.

if you use a pot, you need to consider making it a motor pot.

if you use a rotary, there is no absolute position indicator and so it skips this whole problem and there is nothing to update but a bargraph or something on some display.

as far as usability, see how fast the pot responds when I twist it and how fast the display updates and stays in sync. you can sweep from 127 to 0 db pretty reliably with a fast twist and half db accuracy in steps. it speaks well of the pot (and motor pot) solution.
__________________
My Photostream:http://www.flickr.com/photos/linux-works/
  Reply With Quote
Old 22nd February 2014, 08:12 PM   #116
Dr_EM is offline Dr_EM  United Kingdom
diyAudio Member
 
Dr_EM's Avatar
 
Join Date: Oct 2006
Location: Swindon
Those nixie tubes are a nice touch! The motorised pot seems quite effective in this setup. Myself, I now have my nice Bourns encoder and plan to get a graphic OLED soon. The graphic OLED will allow me to use a font twice as large as on a 2x16 line alphanumeric for just the volume readout, other details can be in a smaller font.

I just ran a test with the volume mask in place, so that I could deliver the same digital data without the volume actually changing. I set the level to a fixed +10dB and am using the +20dB gain headphone stage. With the mask in place, no amount of turning the control could cause clicks in the headphones. For me this proves that the clicks I can sometimes hear at very high settings are due to shifting of DC offsets inside the CS3318 and not from digital noise coupling to the analogue circuitry.

I have already demonstrated to myself that the board handles SPI at much higher speed without any overshoot etc. and so that issue of SPI speed is too just a software bug. In all, it appears this board works very nicely!

I have 2 extra PCBs sitting here ready to go if there are any takers, as well as 2 CS3318 chips. PM me, I can get more boards made as well but of course it will be a few weeks
  Reply With Quote
Old 22nd February 2014, 08:13 PM   #117
diyAudio Member
 
Join Date: Dec 2009
How much will a board cost + shipping to Denmark?
  Reply With Quote
Old 22nd February 2014, 08:19 PM   #118
Dr_EM is offline Dr_EM  United Kingdom
diyAudio Member
 
Dr_EM's Avatar
 
Join Date: Oct 2006
Location: Swindon
Looks like about 15/18 Euro
  Reply With Quote
Old 14th April 2014, 09:33 AM   #119
chiily is offline chiily  United Kingdom
diyAudio Member
 
Join Date: Mar 2003
Location: Malvern
Quote:
Originally Posted by Dr_EM View Post
Looks like about 15/18 Euro

Apologies, I've not read the whole thread..

Is this PCB a group buy, or will it become one? I'd be very interested in buying..

Thanks

Sent from my GT-I9305 using Tapatalk
__________________
H&Ks
Garf.
  Reply With Quote
Old 14th April 2014, 09:08 PM   #120
Dr_EM is offline Dr_EM  United Kingdom
diyAudio Member
 
Dr_EM's Avatar
 
Join Date: Oct 2006
Location: Swindon
Hi,

For now it's just the 2 spare boards I have around up for sale but perhaps a group buy is possible? There did seem to be some interest and I can confirm the design works well.
  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
Eagle vs. Sprint-Layout for PCB design/layout hollowman Parts 11 12th January 2014 09:01 PM
cs3318 and arduino linuxworks Analog Line Level 72 26th March 2013 11:56 PM
Digital Volume Contol - CS3318 - MAX5440 Ridin '24'z Digital Source 5 8th June 2011 01:36 PM
Who can do PCB layout d3imlay Solid State 3 23rd November 2007 11:55 PM
help for a PCB layout pencoat Parts 0 4th October 2006 01:52 AM


New To Site? Need Help?

All times are GMT. The time now is 12:37 AM.


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