Hi James,
Your enthusiasm and contributions to this thread were and still are invaluable! I personally appreciated and thankful for yours and All. Here by no means of being sarcastic but positively recognised at a personal level.
Thanks for the buffer stage explanation. It is a pity that I live other side of the Globe, else, I love to have a gtg with you guys for the listening opportunity. May be when I have this DAC housed proper, I would be delighted to have it sent to you for evaluation?
Best,
Chanh
Your enthusiasm and contributions to this thread were and still are invaluable! I personally appreciated and thankful for yours and All. Here by no means of being sarcastic but positively recognised at a personal level.
Thanks for the buffer stage explanation. It is a pity that I live other side of the Globe, else, I love to have a gtg with you guys for the listening opportunity. May be when I have this DAC housed proper, I would be delighted to have it sent to you for evaluation?
Best,
Chanh
Actually started my build. I got a few "nice" parts for what will mainly be a vanilla build. No shunts, nor pin 20 shenanigans.
I did get a set of 30 VA toroids from Plitron, and a pretty much full set of Silmic II caps. One Watt Audio Note Tantalum for the output, along with Sowter transformers.
I have the Wave I/O board, and a fanless DIY computer for it. I also have an rPi, and the new G5 isolator board to try, but setting up the rPi will be a learning experience for me, and I'd like the DDDAC to be test first with the known Wave board.
I'm set up for 2 chassis syetem, PSU in one box, and the input, DAC and output systems in another.
Building the boards, I'll start pictures once the chassis gets started.
I did get a set of 30 VA toroids from Plitron, and a pretty much full set of Silmic II caps. One Watt Audio Note Tantalum for the output, along with Sowter transformers.
I have the Wave I/O board, and a fanless DIY computer for it. I also have an rPi, and the new G5 isolator board to try, but setting up the rPi will be a learning experience for me, and I'd like the DDDAC to be test first with the known Wave board.
I'm set up for 2 chassis syetem, PSU in one box, and the input, DAC and output systems in another.
Building the boards, I'll start pictures once the chassis gets started.
Are you building a single board DDDAC? If you do I would highly recommend adding at least the two 8v shunts at the Vcc side of the board.
IMO, you'll unlikely better your AN 4.1 SE (with a single board and no shunts), Although sound signatures are quite different, so it all very much depends.... I had the privilege to hear 4 board DDDAC+shunts, 4 board DDDAC (without) and AN 4.1 SE, side by side. I thought the AN came second best in that setup.
Last edited:
I have moved from a four board vanilla DDDac to a one Board vanilla in preparation to add Tent 8v and Belleson 3.3v along with better psus.
I was prepared for a big drop in sq, but used capacitorless driven from the pos and neg I think the sound is 'preferable' to my 4 board version. How? The 4 board had more body but the one board seems to sound 'purer' and 'crisper' by comparison.
I don't need much current drive for my sensitive valve pre /power combo.
David
I was prepared for a big drop in sq, but used capacitorless driven from the pos and neg I think the sound is 'preferable' to my 4 board version. How? The 4 board had more body but the one board seems to sound 'purer' and 'crisper' by comparison.
I don't need much current drive for my sensitive valve pre /power combo.
David
Confirms what I expected. That is why I wold like to see DAC designers make a DAC that delivers say 100mA or more. I bet it will come Anybody know of a discrete parts DAC i.e the DAC chip done as a discrete miniboard
Last edited:
Simple explanation of why LM334 doesn't sound good:
the silicone capacitor within the chip.
ICs do not sound as good as discrete circuits because of these necessary things.
The tight packaging allows great frequency response but at the expense of great sound. I know they can sound good, but never great.
HARRIS SEMICONDUCTOR before they went away (were they absorbed into TI?) had come up with a way to minimize the problems of chip capacitors and I am sure the technique is uniform now but there was a lot of excitement about these chips decades ago - they were the first better sounding ICs.
No way around it, unfortunately.
the silicone capacitor within the chip.
ICs do not sound as good as discrete circuits because of these necessary things.
The tight packaging allows great frequency response but at the expense of great sound. I know they can sound good, but never great.
HARRIS SEMICONDUCTOR before they went away (were they absorbed into TI?) had come up with a way to minimize the problems of chip capacitors and I am sure the technique is uniform now but there was a lot of excitement about these chips decades ago - they were the first better sounding ICs.
No way around it, unfortunately.
What is the problem with the capacitors, and how do they affect the sound?
I've just tried a crude but very successful experiment! I removed the 2nd mini choke after the LF80 and replaced it with a 15H choke with a Elna Silmic 100uf across the output. This gives a quite remarkable step up in SQ.......
I will do more comparisons, when /if my Tent Regs ever turn up. 2 months and counting.....
David
I will do more comparisons, when /if my Tent Regs ever turn up. 2 months and counting.....
David
Hey! Who's tossing around technical discussions (= articles)!?... Is that you, Ross?...
Good find on your part, actually. The 2SK246 is in fact designed for constant current sourcing (maybe a long-gate JFET?), and the capacitances give it a likely excellent high frequency performance.
I can't believe that I had that same article open on my pile of papers at that particular page for the entire week (for another project) and the tempco plots didn't jump at me. I am a distracted reader these days... Borbely also has (in that same article - same page, to be exact) a circuit for fitting JFETs for various operating points and conditions; may be a good little appliance to build and use for whoever wants to finely tweak and select parts for this.
Radu.
Last edited:
Enjoy! Hope that project goes well for you
Are you comparing 1 deck capless pos+neg output to a 4 deck with output caps?
Looking forward to seeing how you attack this. I really want to make mine a 2 box affair with a separate box to house all the ac/transformers/mains etc and ideally with a nice multi connector to carry the various dc supplies into the other box, but I've been waiting for my build to stop evolving and also for some inspiration
Interesting stuff, thanks
I hadn't spotted the internal capacitor An externally hosted image should be here but it was not working when we last tested it.
Looks like the lt3092 doesn't have one of those, so worth keeping that in the list of possible solutions.
On the subject of lt3092, I notice that in the datasheet, there is an example layout for a high accuracy application using a diode instead of one of the resistors, but unless I missed something, that particular setup isn't covered in the text there. Anyone care to comment?
An externally hosted image should be here but it was not working when we last tested it.
Cheers,
James
The point being done there - correct me if I'm wrong, Rick - is that this is a compromised capacitor being built on the chip dye etc. Not something anywhere close to a well performing cap for audio duty.
LT3092 Comments
dwjames: "Looks like the LT3092 doesn't have one of those, so worth keeping that in the list of possible solutions. On the subject of LT3092, I notice that in the datasheet, there is an example layout for a high accuracy application using a diode instead of one of the resistors, but unless I missed something, that particular setup isn't covered in the text there. Anyone care to comment?"
I should be receiving 4 of these on Monday. We only have about 2.45V from pin 20 to ground so we can't use up 1.25Vof this in the voltage regulating resistors part of the circuit. I plan to use 18-20K adjustable resistance on the left side and a 470ohm fixed resister on the right side for a 0.188 voltage drop at 0.4ma. The LT3092 device will have about 2.26V to work with. It will be interesting how the scope display for this compares with other circuits I have tested. I should also be getting ten 2SK246 JFETS on Monday. The fun never stops.
dwjames: "Looks like the LT3092 doesn't have one of those, so worth keeping that in the list of possible solutions. On the subject of LT3092, I notice that in the datasheet, there is an example layout for a high accuracy application using a diode instead of one of the resistors, but unless I missed something, that particular setup isn't covered in the text there. Anyone care to comment?"
I should be receiving 4 of these on Monday. We only have about 2.45V from pin 20 to ground so we can't use up 1.25Vof this in the voltage regulating resistors part of the circuit. I plan to use 18-20K adjustable resistance on the left side and a 470ohm fixed resister on the right side for a 0.188 voltage drop at 0.4ma. The LT3092 device will have about 2.26V to work with. It will be interesting how the scope display for this compares with other circuits I have tested. I should also be getting ten 2SK246 JFETS on Monday. The fun never stops.
Ahh, voltage regulating resistor. That's a new symbol on me, thanks
Look forward to hearing your thoughts on the other solutions.
I guess having to be made really cheap as a small part of another cheap device, having to be minutely small and having no choice whatsoever on the type of dielectric sound like significant compromises to me.