| Blitz |
I thought it should be promising to combine one of the best DAC (Ultraanlog DACD20400A) with Nelson's Outputstage. Well, only two issues:
1. Does anyone know how much current this DAC produces ? Neslon's D1 is made for 2mA of the PCM63.
2. How should it be connected ? Obviously OUTR is the signal, but why is AL0 /AL1 both grounded ? I would assume that I connect the D1 directly to OUTR, right (see attachment)?
Best Regards
Frank |
|
|
| uli |
Hi Frank,
the Ultra Analog DAC has a buffered lo impedance voltage source output
Guten Rutsch!!!
Uli :nod: :nod: :nod: |
|
|
| Nelson Pass |
In that case, the D1 stage is not appropriate, as it
expects to see a current source. |
|
|
| Blitz |
OK, in that case I will forget the Ultra-analog-DAC and combine the D1-Stage with my TDA1547 (with 1mA output). I know that this has been discussed before iwth changing the current through the 610 from 10mA down to 1mA, but I have not understood if ths has finally worked out or if Nelson sees this as a good approach to change the D1-stage (with only 1mA of current through it, obviously bandwith etc of the stage will change).
By the way: HAs anyone tried to convert the D1-Stage into a tubestage ? |
|
|
| Nelson Pass |
You will get less distortion with the higher current figure.
As to using a tube, there is nothing to stop you. |
|
|
| promitheus |
well you could try to make the D1 output stage with tubes.
Itīs I think a good idea but I am not so good with tubes.
There are even tubes you could use at low voltages. |
|
|
| Klaus |
| quote: | Originally posted by Blitz
OK, in that case I will forget the Ultra-analog-DAC and combine the D1-Stage with my TDA1547 (with 1mA output). I know that this has been discussed before iwth changing the current through the 610 from 10mA down to 1mA, but I have not understood if ths has finally worked out or if Nelson sees this as a good approach to change the D1-stage (with only 1mA of current through it, obviously bandwith etc of the stage will change).
By the way: HAs anyone tried to convert the D1-Stage into a tubestage ? |
Hi Blitz,
you asked my before in the other tread about this issue. I did not answer because my project was postponed...
But right now I start to test the modified D1 stage with the TDA1547 just sourcing 0,1mA. Details will follow here.
I finally biased the circuit not at 1mA but at 8mA using a double JFET 2SK389. CCS like in the Aleph P1.7 everywhere. My listening tests just start...
Regards
Klaus |
|
|
| promitheus |
That sounds very interesting.
Please post more.
;) |
|
|
| Klaus |
| quote: | Originally posted by promitheus
That sounds very interesting.
Please post more.
;) |
More to follow but for the curious guys already some pics... (sorry in another forum in German language) http://www.audiomap.de/forum/index...._id=15097&page=
regards
Klaus |
|
|
| promitheus |
thatīs funny I was just looking at that before I read it here.
what did you use as output caps where Pass uses 220uF? |
|
|
| Klaus |
| quote: | Originally posted by promitheus
thatīs funny I was just looking at that before I read it here.
what did you use as output caps where Pass uses 220uF? |
I use - like everywere in my Aleph4 mono block + Aleph P1.7 chain - 220uF Panasonic FC electrolytics bapassed with blue 10nF Vishay Roederstein MKP1837 (or older ERO KP1830). As written frequently before here I know no better cap. And they are so cheap... |
|
|
| Blitz |
Nice work, Klaus. I just finished the as-symetrical D1-stage to test it first against the Output-Stage of my Sheldon DAC(passive I/V-conversion). When I did that, I will report back the results.
Why did you need 220uF output Cap ? I thought the cap-size is determined by the impedance it sees in the next stage, which should be 22k (in parallel with the impedance of the preamp. which is in my case 300K) in the case of the D1, right ?
Best Regards |
|
|
| promitheus |
Hallo Blitz,
which DAC did you use it on?
I want to do the same thing but I was concerned about the output cap.
What cap did you use?
I also have an unbalance output. |
|
|
| Klaus |
| quote: | Originally posted by Blitz
Nice work, Klaus. I just finished the as-symmetrical D1-stage to test it first against the Output-Stage of my Sheldon DAC(passive I/V-conversion). When I did that, I will report back the results.
Why did you need 220uF output Cap ? I thought the cap-size is determined by the impedance it sees in the next stage, which should be 22k (in parallel with the impedance of the preamp. which is in my case 300K) in the case of the D1, right ?
Best Regards |
The output cap can not be to big, but to small. So I see no reason not to use 220uF like also the one and only is doing in the Aleph amps and the D1. For sure it could be decreased to some uF like many guys are doing. But maybe there are some effects due to "hardest possible coupling" of the cable output load to the active output stage by the very low ESR and inner impedance of a big cap.
| quote: | Originally posted by promitheus
Hallo Blitz,
which DAC did you use it on?
I want to do the same thing but I was concerned about the output cap.
What cap did you use?
I also have an unbalance output. |
I checked the sound losses of coupling caps before I started to build up any Aleph stuff. To my mind these losses are less critical than negative sound effects by untweaked devices ore "basic" cables. The best tested Pass products show that even standard coupling caps can not be so bad... To much prejudices against coupling caps! |
|
|
| promitheus |
well why use 220uF electrolytics when you can use 22-33uF MKP caps?
what you are suggesting is that someone could also use 1000uF electrolytic for better quality. That would be good for power supplies but not for in signal paths. |
|
|
| Blitz |
The size of the output cap is 4.7uF - Mundorf Supreme. I think Klaus agrees with us on the superior quality of the supremes.
I seems that all the theory guys are on vacation to tell us which impedance the C really sees, so once I have it build in, I will use a measuremet CD to see wether we have losses in the bass with 4.7uF. - If 1uF would be suffiencient, I could even use a Supreme with silver/oil which I have here for testing...
The Sheldon-Dac uses the PCM63 as the original D1. Interesting enough Sheldon himself played around with all kind of filters for the HF-garbage and came to the conclusion, that listening to it without any filter is best. I can confirm that. Sofar this DAC has not been beaten by anything. MAybe Nelson or Klaus's output stage will do that job ? |
|
|
| promitheus |
I think with a normal load of 47KOhm you only need 1-2uf.
If you use 10-22uF its more then enough even for loads down to 5KOhm. |
|
|
| Klaus |
| quote: | Originally posted by Blitz
The size of the output cap is 4.7uF - Mundorf Supreme. I think Klaus agrees with us on the superior quality of the supremes.
I seems that all the theory guys are on vacation to tell us which impedance the C really sees, so once I have it build in, I will use a measuremet CD to see wether we have losses in the bass with 4.7uF. - If 1uF would be suffiencient, I could even use a Supreme with silver/oil which I have here for testing...
The Sheldon-Dac uses the PCM63 as the original D1. Interesting enough Sheldon himself played around with all kind of filters for the HF-garbage and came to the conclusion, that listening to it without any filter is best. I can confirm that. Sofar this DAC has not been beaten by anything. MAybe Nelson or Klaus's output stage will do that job ? |
I use the Panasonic FC electrolytics + MKP bypass because they sound extremely open, dynamic and w/o any sonically problems for ca. 2. The Supremes also sound superior (and much better also with 10nF MKP bypass - less sharp and more color), but they cost so much. Not to talk about this silver Supreme stuff... That is the only reason I do not use them.
Maybe the value of the cap regarding bass losses is not the main story, I think it has really something to do with how "hard" the cap couples the cable load and damp all parasitic effects. This is maybe given by the ESR value of the cap. Regarding this a 220uF electrolytic is very good. Bigger ones could get higher inductance and are not useful. But as Nelson and many other producers show 100 or 220uF seem to be suitable. Just test it and do not think to much about theory!
I can also just confirm that after throwing away the OPA filter chain a DAC plays like dynamic fire! Since my TDA1547 works almost without analog filter (just passive 6dB by a cap parallel to the I/U-resistor like also Nelson does) I hear really the music. It blows away almost everything. And I hear no artifacts from IM effects, so how relevant are really this HF distortions? I doubt they are. |
|
|
| Blitz |
Klaus,
This sounds extremely promising, looking forward to see the schematic of the circuit and your sound description of course.
My Marantz CD16 with the TDA1547 inside is waiting for years now to prove that it can make music and I would use your outputstage 1:1.
Best Regards |
|
|
| Klaus |
| quote: | Originally posted by Blitz
Klaus,
This sounds extremely promising, looking forward to see the schematic of the circuit and your sound description of course.
My Marantz CD16 with the TDA1547 inside is waiting for years now to prove that it can make music and I would use your outputstage 1:1.
Best Regards |
I will come back to you and this forum asap...
The sound of the 2SK389 varies extremly with its bias current - I currently search for the best compromize. But the sound improovement still is unbeleavable!!!!! Simply like agin the change from a "normal" pre amp to the Aleph P1.7. And the CCSs seem to give a extreme stability of sound. And this bass...
tweaky regards
Klaus |
|
|
| Blitz |
| In my experience the sound varies especially as well for different Idss of the different devices. So, a 389 GR sounds quite different than a BL or a V (at a given bias). |
|
|
| promitheus |
What is the frequency response of the I/V stage from the D1?
I will be using upsampled data and want the filter to be above 48Khz. |
|
|
| Klaus |
| quote: | Originally posted by promitheus
What is the frequency response of the I/V stage from the D1?
I will be using upsampled data and want the filter to be above 48Khz. |
Pass D1 uses 1,5k parallel to 2,7nF = 40,5kHz f_o for the 2mA PCM63. So 3V peak per branch.
I use 6,5k and 680pF = 44kHz f_o for my TDA1547 just giving 0,1mA. Just 0,65V - so 1,3V on XLR. Enough for no use of the volume adjuster of the D1.
Simply calculate your values for your DAC used. |
|
|
| promitheus |
Thatīs how much I calculated last time so I wasnīt wrong.
I am planing on designing a PCB for the I/V stage. I want to replace the 4 opamps in my DAC to this.
Is anybody interested?
I want to use silver mica for the small caps.
Panasonic FC for the electrolytics.
I probably will use 4,7 uF MKP caps for the output. |
|
|
| Klaus |
Hi,
I put "my" schematic - it is more a schematic of the one and only, because I just combined P1.7 and D1 circuits online http://www.audiomap.de/forum/index...._id=15147&page= . Sorry, but again in German language and in another forum. I have problems to put my stuff into here (donīt know why), so I refer to the other forum which offers somewhat easier data download.
The key massage: find your own bias current! The sound is changing really radical from 3mA up to 10mA through the 2SK389. Also the DAC current will for sure change the best current value. Put one trimming resistor 150Ohm plus 63Ohm in line in for R2 and adjust the bias current. Then set R1 to U_b/2 and listen
Test it, the sound is perfect!!!!!!!!!!!!!!!!!!!!!! Pics of my hardcore P2P work under the link in the quoted words above.
Regards
Klaus |
|
|
| Klaus |
If no pic is seen I have no idea what Iīm doing wrong...
D:\Witte privat\Fotos HiFi\D1-Stufe1r.JPG |
|
|
| Klaus |
Our system admin maybe has changed again the security settings, now it works again. Nice!
The pic shows the mini sized D1 clone I/U + buffer stage. Fully symmetrical (4 "channels") and two voltage regs a la P1.7 in the middle. The 610/9610 lock somewhat pressed in, because I originally layouted for smaller FETs. The sound is unbelievable nice. I trust in Nelsons circuits and just copy them.
Hopefully the other day a nice, small PCB would be available for this stage. Small to bring it into given packaging situations. By adjusteing two Rīs and the filter cap it can easily by adapted to each DAC with symmetrical current output.
Regards
Klaus |
|
|
| Cobra2 |
Great work, I knew the TDA1547 could perform...
-waiting for the PCB... ;)
Arne K |
|
|
| Blitz |
| Looks indeed great, I would be with the pcb as well. |
|
|
| nar |
Don't wait for the PCB
DIY ! |
|
|
| Nelson Pass |
| quote: | | Originally posted by Klaus I use the Panasonic FC electrolytics + MKP bypass because they sound extremely open, dynamic and w/o any sonically problems for ca. 2. |
It is my humble opinion that the Panasonic electrolytic types
are quite good for audio if you need for than about 10 uF, and
of course you bypass them with the film cap of your choice if
you want. And they are cheap and available. :cool: |
|
|
|
