Hi Marce,
I saw most of the time very thin traces on the pcbs are sufisant for those DACS. Looking at some links you provided elswhere, it seems more important to have a good separation width between conductive signals traces (but their gnd return of course : close trace or continuous ground plane at the bottom of the conductive trace) to isolate them from each others. (article with guard traces, width, etc you provided one day
in the past. thanks again for your didactic each time, help a lot a noob enthusiat like me
So less Impedance with no more widhter traces on digital pcbs seems not to be usefull (short length, etc), so I understand changing copper for silver don't matter on such pcbs...
Is the insulation dielectric change something on the copper trace of pcbs in our hobby as the traces are very thin ?
I saw most of the time very thin traces on the pcbs are sufisant for those DACS. Looking at some links you provided elswhere, it seems more important to have a good separation width between conductive signals traces (but their gnd return of course : close trace or continuous ground plane at the bottom of the conductive trace) to isolate them from each others. (article with guard traces, width, etc you provided one day
in the past. thanks again for your didactic each time, help a lot a noob enthusiat like me So less Impedance with no more widhter traces on digital pcbs seems not to be usefull (short length, etc), so I understand changing copper for silver don't matter on such pcbs...
Is the insulation dielectric change something on the copper trace of pcbs in our hobby as the traces are very thin ?
The majority of digital designs (if not all) that don't involve DDR memory, Gigabit Ethernet and other such very high speed traces (we wont get into RF) the trace width is chosen to allow the design to be routed whilst keeping the design easy to manufacture, ie you choose the widest trace widths possible, common sizes are from 0.1mm - 0.3mm, with the majority these days 0.15mm - 0.2mm (6 - 8 thou).
Separation, as a rule of thumb the majority of the electro-magnetic fields extend 1W either side of the trace (W = the width of the trace), when routing a board the standard track and gap size is usually the same...i.e. 0.1/0.1mm, where possible it is always better to have as large a gap as possible and on clocks (and periodic changing signals, a fancy name for clocks) always try and us 3W for the spacing (0.1/0.3mm trace/gap).
Guard traces can be a good thing and a bad thing! mainly a good thing, a lot of PCBs (both analogue and digital) will have copper pours on both outer layers (with many stitching vias to the contiguous ground plane).
Use a good quality 170 deg C Tg rated FR4 and you wont have any problems don't worry about fancy dielectric's just adds cost for no benefit.
Do nice pretty neat routing, if you are going to play about changing PTH components then always tear-drop the trace when it enters the pad for added strength and to avoid the trace cracking where it enters the PTH pad (a week spot).
Look on the bright side, I am doing a 1600 component board with 4000+ connections, analogue/digital fast Ethernet, I have spent over 1 1/2 weeks on placement so far.....
Separation, as a rule of thumb the majority of the electro-magnetic fields extend 1W either side of the trace (W = the width of the trace), when routing a board the standard track and gap size is usually the same...i.e. 0.1/0.1mm, where possible it is always better to have as large a gap as possible and on clocks (and periodic changing signals, a fancy name for clocks
) always try and us 3W for the spacing (0.1/0.3mm trace/gap).Guard traces can be a good thing and a bad thing! mainly a good thing, a lot of PCBs (both analogue and digital) will have copper pours on both outer layers (with many stitching vias to the contiguous ground plane).
Use a good quality 170 deg C Tg rated FR4 and you wont have any problems don't worry about fancy dielectric's just adds cost for no benefit.
Do nice pretty neat routing, if you are going to play about changing PTH components then always tear-drop the trace when it enters the pad for added strength and to avoid the trace cracking where it enters the PTH pad (a week spot).
Look on the bright side, I am doing a 1600 component board with 4000+ connections, analogue/digital fast Ethernet, I have spent over 1 1/2 weeks on placement so far.....
.....Separation, as a rule of thumb the majority of the electro-magnetic fields extend 1W either side of the trace (W = the width of the trace), when routing a board the standard track and gap size is usually the same...i.e. 0.1/0.1mm, where possible it is always better to have as large a gap as possible and on clocks (and periodic changing signals, a fancy name for clocks
Look on the bright side, I am doing a 1600 component board with 4000+ connections, analogue/digital fast Ethernet, I have spent over 1 1/2 weeks on placement so far.....
Thanks & I believe it could help some fellows whom are designing clock boards in this moment
That's crasy, a computer have less more components....
Purely digital boards these days such as a "computer" (processor, memory, communications etc.) might have 5-20 large ICs, a few smaller ICs and a few hundred decoupling caps and pull up resistors. These type of boards will have a few thousand connections....
What a lot of people don't see are the control boards and similar, we still live in a analogue world and need analogue control and measurement to control things, so these sort of boards have lots of analogue front ends and outputs, pre-amps frequency counters, output amps analogue to digital, digital to analogue etc. The same sort of stuff you play with in audio with the same problems (ground loops, noise etc. etc.) only often more complex an]]with many channels ADC/DAC on the same board, a few DSPs also. I have recently done boards for testing ICs, sensors (for space), diodes, power station, flight refuelling, recycling machinery, communications, rally cars and soon going to the Netherlands to do some work for the ESA. Not everything is digital as I said it is an analogue world
What a lot of people don't see are the control boards and similar, we still live in a analogue world and need analogue control and measurement to control things, so these sort of boards have lots of analogue front ends and outputs, pre-amps frequency counters, output amps analogue to digital, digital to analogue etc. The same sort of stuff you play with in audio with the same problems (ground loops, noise etc. etc.) only often more complex an]]with many channels ADC/DAC on the same board, a few DSPs also. I have recently done boards for testing ICs, sensors (for space), diodes, power station, flight refuelling, recycling machinery, communications, rally cars and soon going to the Netherlands to do some work for the ESA. Not everything is digital as I said it is an analogue world
What a pity you have so few time to developp for communauty a super DAC as we saw recently in the GB section from a fellow with 0.1 to 0.2 % precision resitors or make to us a super FIFO, simultaned mode in one board without 2 meters of uf.l cables between pcbs to have a standalone pcb to feed our multibit dacs : tda-1541, PCM & AD family dacs.
Did you have any time to finish your Subbu revision pcb ?.... I will ask to Railways and plane companies to put solder station on their seats
Hope you a good travel, you need now a very good transportable system : good headphone and compact player like the IFI Nano transportable usb dac with battery !
see you... (try my huge MKZ muse caps for the Wolfson on the Subbu if you have time to finish your pcb....better subjective listening curve than the better inductance of a 2.5 mm pitch with all the polymers caps I tryed before.... some things are strange in the analog domain when we check it with ears )
Did you have any time to finish your Subbu revision pcb ?.... I will ask to Railways and plane companies to put solder station on their seats
Hope you a good travel, you need now a very good transportable system : good headphone and compact player like the IFI Nano transportable usb dac with battery !
see you... (try my huge MKZ muse caps for the Wolfson on the Subbu if you have time to finish your pcb....better subjective listening curve than the better inductance of a 2.5 mm pitch with all the polymers caps I tryed before.... some things are strange in the analog domain when we check it with ears
)
Last edited:
I don't seem to have any time these days for anything fun, rather doing to much work (done a few weekends)!!! but it pays the bills. I am trying to find time to learn Lightroom and photoshop properly, bought my self a set of training videos and books (PDF) from Adobe last year ($19) and haven't got round to using them, as to getting chance to actually take any photos
I don't fly so I'll be going via North Sea Ferries and driving.
But I would rather be busy and employed.
I don't fly so I'll be going via North Sea Ferries and driving.
But I would rather be busy and employed.
Just asked my boss, I'm going to be working here:
Hogenakkerhoekstraat
9150 Kruibeke
Belgium
Was told it was only a few miles from Rotterdam...My main concern is getting to a Christmas market. Going to be fun, overnight Ferry Sunday night, work full week in 4 days and overnight ferry back to UK Thursday night!!! Looking forward to it though, I like getting away from the office and do quite a lot of work where I have to be on site due to certain data restrictions.
Hogenakkerhoekstraat
9150 Kruibeke
Belgium
Was told it was only a few miles from Rotterdam...My main concern is getting to a Christmas market. Going to be fun, overnight Ferry Sunday night, work full week in 4 days and overnight ferry back to UK Thursday night!!! Looking forward to it though, I like getting away from the office and do quite a lot of work where I have to be on site due to certain data restrictions.
Ouch. I guess it is the classic British understatement usage of "a few". It is 70 miles, and at least 1 h by car on a good day. And there are not many good days on the Antwerp ring road - you are right on the other side of Antwerp.
Hope you don't have to visit the main ESA site in Nordwijkerhout too often from there - that is another hour.
The one in Antwerp isn't too bad.
The overnight ferry isn't too bad, but that doesn't sound like much spare time for being a tourist... Would you cross from/to Hull?
Last edited:
Yep, quick drive from one side of the UK to the other on the M62...
Thanks for the tip about the Antwerp ring road. No I never get much time to look round, and the wife gets a bit upset if I don't come home at weekends to sort the Kids/grandkids and stray daughters (My oldest daughter and hubby have decided to split up so she's staying with us at the moment! with 3 grandkids 2 to 11 years, add my wonderful 15 year old twins (boy and girl) into the equation and you can understand why I cant wait to go)
Thanks for the tip about the Antwerp ring road. No I never get much time to look round, and the wife gets a bit upset if I don't come home at weekends to sort the Kids/grandkids and stray daughters (My oldest daughter and hubby have decided to split up so she's staying with us at the moment! with 3 grandkids 2 to 11 years, add my wonderful 15 year old twins (boy and girl) into the equation and you can understand why I cant wait to go
)As long as you avoid Leeds and Hull rush hour...
With 3 grandkids 2 to 11 years, add my wonderful 15 year old twins (boy and girl) into the equation and you can understand why I cant wait to go
I can imagine it is not an environment that is very conductive to quiet, methodological thinking...
If you do get a chance to come up to Netherlands, give me a shout! There is also a possibility my wife wants to drag me down to the Antwerp Christmas market too...
I am sort of use to the madness at home, we have had 7 kids and now have about a million grandkids there is never a dull moment
Hi all, I spoke a few times of a new 5V PSU intended for Squeezebox Touch players etc. and it could also feed the V3 (it was not intended for that originally). Last week the boards came in after a very long wait. I assembled a few and it performs quite good. I needed only 2 of them myself but a Group Buy maybe is possible. I will open a separate thread.
Since many posts are off topic I think this post would not hurt
Since many posts are off topic I think this post would not hurt
Last edited:
Hi All
Reminder for limited 2nd run : http://www.diyaudio.com/forums/grou...-subbu-dac-v3-gb-interest-98.html#post4112114
Regards
Phil
Reminder for limited 2nd run : http://www.diyaudio.com/forums/grou...-subbu-dac-v3-gb-interest-98.html#post4112114
Regards
Phil
Hi
Links updated in the initial Subbu DAC V3 GB first Post :
http://www.diyaudio.com/forums/grou...-wm8804-spdif-power-supply-pcb-group-buy.html
GB is limited to 50 units of each PCBs, DAC & PSU.
Assembly Service form added.
Regards
Phil
Links updated in the initial Subbu DAC V3 GB first Post :
http://www.diyaudio.com/forums/grou...-wm8804-spdif-power-supply-pcb-group-buy.html
GB is limited to 50 units of each PCBs, DAC & PSU.
Assembly Service form added.
Regards
Phil
Finally I could resolve my problems. Was a bad solder somewhere, finally it works like a gem.
Exchanged my (hardcore modified, thanks to Daario) 2496 AK4396 Chinese DAC (which already was a quite good performer) with the Subbu / JP DAC for my Shigaclone. Acoustic became much clearer and transparent. I would say its more "balanced" as I lost a little bit of bass and mids, thats where the Chinese DAC was a real flavor adder ;-)
Thanks for making this DAC Subbu and JP! You really did a great job here!
Regards,
Fabian
Exchanged my (hardcore modified, thanks to Daario) 2496 AK4396 Chinese DAC (which already was a quite good performer) with the Subbu / JP DAC for my Shigaclone. Acoustic became much clearer and transparent. I would say its more "balanced" as I lost a little bit of bass and mids, thats where the Chinese DAC was a real flavor adder ;-)
Thanks for making this DAC Subbu and JP! You really did a great job here!
Regards,
Fabian