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.

Valve DAC from Linear Audio volume 13
Valve DAC from Linear Audio volume 13
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 27th June 2017, 12:35 PM   #11
rfbrw is offline rfbrw
diyAudio Member
 
Join Date: Oct 2001
Location: .
Can I get rid of the SRC4392 ?
  Reply With Quote
Old 27th June 2017, 08:25 PM   #12
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
Yes, but only if you are prepared to drastically redesign the DAC, at least the digital stuff in the FPGA and possibly the crystal oscillator.

If you have no objections against asynchronous sample rate conversion as such, you could throw out my prefilter and interpolation chain and try to use the freed FPGA resources for making your own asynchronous sample rate converter. If it fits, you can then use the exact same hardware I used, but with a different FPGA configuration file and without the SRC4392.

If you don't want asynchronous sample rate conversion at all, you would have to change the crystal frequency to a multiple of the usual audio sample rates, for example 28.224 MHz. The interpolation chain would then have to be redesigned (FPGA configuration file change). Depending on the incoming sample rate, you would have to switch to a different interpolation factor. There is no need to interpolate all the way up to 28.224 MHz, as the signal transfer function of the sigma-delta modulator and the response of the analogue reconstruction filter also suppress images. (In fact in the original design I interpolate up to 3 MHz with 27 MHz clock, and then use a zeroeth-order hold. You can probably go lower than 1/9 of the clock frequency.)

Without asynchronous sample rate conversion, if the DAC has to synchronize to an incoming clock, you have to make the crystal oscillator tuneable and add a PLL. As mentioned in the article, I found that this cannot be done with just a reactance valve. The crystal oscillator need not be tuneable if the DAC's clock is used as the master clock.

Last edited by MarcelvdG; 27th June 2017 at 08:29 PM.
  Reply With Quote
Old 29th June 2017, 06:16 PM   #13
rfbrw is offline rfbrw
diyAudio Member
 
Join Date: Oct 2001
Location: .
Thanks for the reply. Very useful. I was looking to shrink the pcb not that the loss of the SRC4392 would make much difference. Air wiring the valve stuff might be the way to go.
  Reply With Quote
Old 29th June 2017, 09:06 PM   #14
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
There are some things you could move from the main PCB to a simple one- or two-layer PCB or a piece of perfboard or whatever with little or no risk of performance loss:

-Power supply
-10 uF output capacitors and mute relays (but not the 8.2 nF polystyrene capacitors at the anodes of the E88CC valves!)
-Anode resistors and potmeters (R5 - R14 - R25 - RV1 - R26 - R27 - R16)
-Relay control transistors
-Neon lamp drivers (neon indicator lamp drivers, that is); you can also remove six out of seven of them if you are not interested in the surprise mode, or replace them with LEDs with series resistors

If you want to change the layout of the actual core of the DAC, make sure you have good shielding between the digital and the analogue circuitry, between the sigma-delta modulate and the clock, and between the sigma-delta modulate and the reference. The sigma-delta modulate must also have a good return path to the digital circuitry, that is, the loop from digital to the sdin pin of Figure 9 and back via C2 - U4 - C38 - C5 or C24 - ground or supply must have a small area. In fact I did this wrong initially, before I added C5 and C24 to locally decouple the 74AHCT74 to the ground and -300 V planes. Besides, you will then probably have to experimentally change C6, L2 and L4 in Figure 13 because of different parasitic capacitances.
  Reply With Quote
Old 30th June 2017, 10:02 PM   #15
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
Actually if you want to make a cheap variant with as few SMDs as possible and are willing to accept a higher noise floor, you could do what I did on the first prototype:

-Take a single-sided copper-clad board, drill holes in it for the valve holders and mount the valve circuits and the 74AHCT74 flip-flops on it. Use the copper as a -300 V plane. Make sure the 74AHCT74s are close to the -300 V plane, but well isolated from it. Also make sure you don't accidentally short the -300 V to ground, especially not through your body.
See the attached photos. For mounting the 74AHCT74 ICs (among other things), I had some very small single-sided PCBs that I could glue to the -300 V plane and then solder ICs in DIL package on.

-Use perfboard without copper islands for the power supply (as seen in the foreground of the second photo)

-Use a Trenz carrier board for the FPGA module, see https://shop.trenz-electronic.de/en/...-TE0300/TE0630. This will require some changes in the user constraint file of the FPGA code because some of the outputs I use are not on the headers of the carrier board, but I can help you with that.

-Use adapter boards and perfboard for the SRC4392 and DIX4192, such as QFP48-56 / LQFP48-56 / TQFP48-56 SMD Adapter Board (0.5mm and 0.65mm pitch), embedded-channel.de

For the reconstruction filter you could use the single-ended variant and maybe use potcores that are one size smaller than the ones I used. I chose rather large ones to be absolutely sure they wouldn't dominate the distortion, but with one size smaller they still shouldn't dominate.
Attached Images
File Type: jpg verkleind.JPG (130.3 KB, 609 views)
File Type: jpg P6090042_klein.jpg (118.5 KB, 601 views)

Last edited by MarcelvdG; 30th June 2017 at 10:06 PM.
  Reply With Quote
Old 30th June 2017, 11:19 PM   #16
rfbrw is offline rfbrw
diyAudio Member
 
Join Date: Oct 2001
Location: .
I would want to stick with a 4-layer PCB and SMD but experiment with the mounting of the valves. It needs to be of a certain size in order to stack with something else.
  Reply With Quote
Old 1st July 2017, 04:27 PM   #17
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
OK, then forget about post 15 and just take into account what I wrote in post 14.
  Reply With Quote
Old 3rd July 2017, 05:08 AM   #18
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
By the way, I estimate that the size of the main PCB can be reduced to about 250 mm by 280 mm when all the not so critical circuitry is removed from it, some axial parts are replaced with radial ones and the digital circuitry is shifted up a bit. This would reduce the Eurocircuits price to

euro 281.02 for 1 PCB
euro 225.96 per PCB when you order 2 of them
euro 179.84 per PCB when you order 3 of them
  Reply With Quote
Old 14th July 2017, 04:31 AM   #19
MarcelvdG is offline MarcelvdG  Netherlands
diyAudio Member
 
Join Date: Mar 2003
Location: Haarlem, the Netherlands
Just out of curiosity I've plotted some intersections of the basin of attraction of the seventh-order chaotic sigma-delta (with clippers switched off). Its basin of attraction is a seven-dimensional figure defined by all combinations of integrator start values that do not result in instability.

These plots have no practical use as far as I know, but I knew from Lars Risbo's PhD thesis that they can be very pretty. My plots are not nearly as pretty as Risbo's, though.
Attached Images
File Type: png SD7_start1en3.png (33.3 KB, 485 views)
File Type: png SD7_start4en6.png (23.1 KB, 470 views)
File Type: png SD7_start5en6.png (27.7 KB, 466 views)
File Type: png SD7_start6en7.png (21.2 KB, 71 views)
  Reply With Quote
Old 19th October 2017, 04:42 PM   #20
vizion is offline vizion  United States
diyAudio Member
 
Join Date: Apr 2007
Hi, is the implementation using Delta Sigma ?
  Reply With Quote

Reply


Valve DAC from Linear Audio volume 13Hide 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 tradition that we use linear PSUs for valve amps? Paul Uszak Power Supplies 9 2nd September 2016 10:58 PM
13.8V 80A Linear PS circuit design analysis for Ham Radio Use FJHookah Power Supplies 69 14th June 2016 02:46 PM
Audio Valve Eklipse volume pot value? Ki Choi Tubes / Valves 0 27th March 2015 05:22 PM
Londog Audio VDt1 Valve Output 24/192 DSD DAC mainscablesrus Mains Cables R Us 1 1st October 2013 06:52 PM
Audio Note DAC 2.1x valve DAC juancho Swap Meet 0 6th July 2005 09:16 PM


New To Site? Need Help?

All times are GMT. The time now is 02:39 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki