syyma,
advised to use Panasonic polymer LE caps for C22 (or a old SP Sanyo : with bader treble but better bass & mid-bass : see my pdf above). APSE caps is good for me at C18 but not also with C22. I believe it depends on each systems : make your own tests !
Even better with the good caps before the ldo regs, C21 & C32 or C35. Maybe try this as it's a very better DAC according to me and on my sytem befor going with bad I2S with bad wires and no impedance matching between boards !
Just two cents...
advised to use Panasonic polymer LE caps for C22 (or a old SP Sanyo : with bader treble but better bass & mid-bass : see my pdf above). APSE caps is good for me at C18 but not also with C22. I believe it depends on each systems : make your own tests !
Even better with the good caps before the ldo regs, C21 & C32 or C35. Maybe try this as it's a very better DAC according to me and on my sytem befor going with bad I2S with bad wires and no impedance matching between boards !
Just two cents...
Last edited:
How much current will you need to deliver? That will determine how much needs to be changed. The PS is only good for less than ~250ma as designed. If you only need to deliver similar current as the Subbu Dac, then changing R1 and R2 is probable enough. Change R2 from 2.15K to 820 ohms and R1 from 4.99K to 6.8K should give 8v output. I'd also use a 10v transformer to give enough input voltage, again with a VA rating high enough for your application.
The LM723 datasheet has all the numbers needed for other voltages such as the 9v that Eldam asked about.
---Gary
But it doesn't work with the Subbu PSU at home ! I went for a tamura with 18 V insteadf 12 V for the SB to have more than 3v between input of the LM723 and output, took the good value of R1 & R2 on the datasheet... It doesnt'work : have 5.5 V ! Too much imput voltage ?
Last edited:
Genuine pcb and genuine datasheet with the values for 9V (sheme n°20 or 22 IIRC) : where is the error according to you ?
R1 & R2 are inverted in relation to the Subbu shematic as you know your self (your post 404) so did you read just the LM723 shematic or did you try it with the genuine Subbu PCB ? I try two different LM723 and the values are the genuine BOM but R1& R2 changed for 9V !
R1 & R2 are inverted in relation to the Subbu shematic as you know your self (your post 404) so did you read just the LM723 shematic or did you try it with the genuine Subbu PCB ? I try two different LM723 and the values are the genuine BOM but R1& R2 changed for 9V !
Eldam,
I don't have enough data to debug the circuit for you. There are many places it could go wrong - not just the values of R1 and R2. I was only stating the fact that this is a very well proven design including the values of R1 and R2 needed to give 9v output. If it's not working then you failed to do something else that was needed. Maybe not enough input voltage? Or maybe not sizing other components for the current needs of your load.
---Gary
I know all these things...and I agree: of course this is the first thing I thought : an error.
But did you try yourself with the values of R1& R2 of the Datasheet for 9V ? Or are you just surmising that I maid an error without really checking yourself?
As I said, all the component were tested both times, desoldered, resoldered with the same values and components that my first pcb but with the second pcb and a bigger Tamura transformer for bigger voltage (with care of two secondary to match the design of Subbu) but have always 5.5 V.
Maybe this design is not valid for more than 5 V as there is a lot of possibilities and different shematic for the LM723... but scheme n20 or 22 is the same but the diode bias... so the only possibility I see is an error of printing with the values of the 9V or a too big input voltage for this scheme ! Maybe I will try on the first pcb with the 12 V transformer... But I can't see why 18V input can not work here !
That's why I asked to Subbu if he could help me as him understanding are far better than ours both.
But did you try yourself with the values of R1& R2 of the Datasheet for 9V ? Or are you just surmising that I maid an error without really checking yourself?
As I said, all the component were tested both times, desoldered, resoldered with the same values and components that my first pcb but with the second pcb and a bigger Tamura transformer for bigger voltage (with care of two secondary to match the design of Subbu) but have always 5.5 V.
Maybe this design is not valid for more than 5 V as there is a lot of possibilities and different shematic for the LM723... but scheme n20 or 22 is the same but the diode bias... so the only possibility I see is an error of printing with the values of the 9V or a too big input voltage for this scheme ! Maybe I will try on the first pcb with the 12 V transformer... But I can't see why 18V input can not work here !
That's why I asked to Subbu if he could help me as him understanding are far better than ours both.
Eldam,
Did you check the PS without load? If it's not giving 9v then you made an error in the parts. If the PS gives 9v with no load and 5.5v with load then you are trying to supply more current than the design can handle. Since the SB needs >1A, I suspect that's part of the problem.
---Gary
Thanks for the advices Gary. Yes I just tried with no load... (don't want to risk my SBs Duet)
<<i think the simplier is to test the 9 V values with my first pcb as it works perfectly and give 5V. If this one give 5.5 with 4.99 & 7.xx ohms R values, I will have the proof of concept than I make an error on the second pcb or that the design doesn't work for higher values. certainly the first solution, i be blind... but where, strange !
Not so important, maid few diy these times, more listening music with the DAC (like with both LE and SP caps...though it's different -LE is softer after 100 hours and win now with classic music on my system! retry the SPC : too much acid at home) and spend time on TDA1541 project (AYA2 2014 + RyanJ coreboards... it will be fun with the silmutanous IS mod on the both pcbs...). Have you got time to try acrylics before the LDOs ? with C21 SP + C22 LE ?
<<i think the simplier is to test the 9 V values with my first pcb as it works perfectly and give 5V. If this one give 5.5 with 4.99 & 7.xx ohms R values, I will have the proof of concept than I make an error on the second pcb or that the design doesn't work for higher values. certainly the first solution, i be blind... but where, strange !
Not so important, maid few diy these times, more listening music with the DAC (like with both LE and SP caps...though it's different -LE is softer after 100 hours and win now with classic music on my system! retry the SPC : too much acid at home) and spend time on TDA1541 project (AYA2 2014 + RyanJ coreboards... it will be fun with the silmutanous IS mod on the both pcbs...). Have you got time to try acrylics before the LDOs ? with C21 SP + C22 LE ?
Thanks, Gary. I should've checked the specs on the lm723. I will need to look at an lm317 or similar design. Regards.
Peterma,
This DAC is not a 3D avenue but have suffisant 3D stage and depend greatly of your own devices around as I understood the little difference between my experiences and the Gary's ones (which is a good sign about this DAC).
If you have a another DAC with more deep, try too understand its tonal balance : more treble, less medium, more bass, more clearness, etc : and try to adjust the caps in the Subbu inrelation to that for the more qdequete tonal balance in relation to the rest of your devices. If you had never a real 3D space at home, its maybe, the room, speakers positions, or unluckily the devices itselves (from the most important to the less
speakers, DAC, pre, amp : IMHO with my two cents experience). Notice than some caps arrengement can give a more hairy sound (sort of 3D : e;g. at home the Panasonic LE cap in C22 position; and others a more deep but less hairy as the Sanyo SP) : try my conf (pdf page 37).
Are you 100% 75 ohms for spidf between source and Dac ? Can help but it's not the most impotant factor : you have to set up your tonal balance if you want to have a gain in deep... Have you not something out of phase also : a speaker, a driver ? Did you try a different source or in a buddy's system ?
This DAC is not a 3D avenue but have suffisant 3D stage and depend greatly of your own devices around as I understood the little difference between my experiences and the Gary's ones (which is a good sign about this DAC).
If you have a another DAC with more deep, try too understand its tonal balance : more treble, less medium, more bass, more clearness, etc : and try to adjust the caps in the Subbu inrelation to that for the more qdequete tonal balance in relation to the rest of your devices. If you had never a real 3D space at home, its maybe, the room, speakers positions, or unluckily the devices itselves (from the most important to the less
speakers, DAC, pre, amp : IMHO with my two cents experience). Notice than some caps arrengement can give a more hairy sound (sort of 3D : e;g. at home the Panasonic LE cap in C22 position; and others a more deep but less hairy as the Sanyo SP) : try my conf (pdf page 37). Are you 100% 75 ohms for spidf between source and Dac ? Can help but it's not the most impotant factor : you have to set up your tonal balance if you want to have a gain in deep... Have you not something out of phase also : a speaker, a driver ? Did you try a different source or in a buddy's system ?
Eldam,
I need to correct my comments from yesterday. The PS board for the Subbu DAC is only good for voltages < 7v. That's because the internal reference voltage of the LM723 is 7.15v. You can change the wiring of the LM723 to support voltages >7v but it requires major surgery to the board. If you look at the LM723 datasheet http://www.ti.com/lit/ds/symlink/lm723.pdf you can see examples of wiring for Vout<7v and Vout>7v. The Subbu DAC PS is very similar to the circuit in Figure 20 of the datasheet. For Vout>7v, you need to reconfigure R1 and R2 to look the schematic in Figure 19. Sorry for my earlier misinformation.
---Gary
Well done Gary, thank you ! How did you see that with the sematic ? figure 4 ? I need to understand because figure n°20 is in the 9 V (and more) spreadsheet of the official datasheet.
My understanding is th Subbu Ps is exactly the same than figure n°20 but the easter eggs of the led reg... (which change nothing to the output voltage after my own tests !).
Well I'm unable to surgery a Teddybear so I would not try with the Subbu PS : it will stay like that for the other Subbu DAC I have and will go with one of the super reg pcbs you link but for the SB Duet... Or maybe I will switch with TDA1541A and network streaming with async USB to I2S with the "AYA2 2014" & the well born TDA1541A "Distinctive" core board.
I would like to hear the difference with an USB input on the Subbu to hear by myself the importance of jitter between a good spidf and an async USB... Hé Maybe Marce fellow has finisched him own revision of the Subbu's layout and put an USB input ! If you read me by chance Marce, send me a mail please between one of your travels !
Finally Gary your silence is heavy of consequence about the acrylic caps before the LDOs :you don't like it on your system or don't hear a difference with them instead of the 10 uf X7r ? Maybe not have time or will to taste it and now switch to other projects ?
If so, did you try the whole mod I maid and/or could you say if you an input between MKS2 2 mm pitch at C32 against Black-Gate ? You say above you prefer the BG, but my understanding is it was a MKS2 5 mm pitch at C35 and JP seems very confident with the little Wima at C32...maybe both the layout and the inductance but it surprise me if it would be huge difference between the two Wimas ! Don't have suffisant buying at Mouser since my last one to put it on my list !
Thanks again for your help about the shematic.
cheers
Eldam
My understanding is th Subbu Ps is exactly the same than figure n°20 but the easter eggs of the led reg... (which change nothing to the output voltage after my own tests !).
Well I'm unable to surgery a Teddybear so I would not try with the Subbu PS : it will stay like that for the other Subbu DAC I have and will go with one of the super reg pcbs you link but for the SB Duet... Or maybe I will switch with TDA1541A and network streaming with async USB to I2S with the "AYA2 2014" & the well born TDA1541A "Distinctive" core board.
I would like to hear the difference with an USB input on the Subbu to hear by myself the importance of jitter between a good spidf and an async USB... Hé Maybe Marce fellow has finisched him own revision of the Subbu's layout and put an USB input ! If you read me by chance Marce, send me a mail please between one of your travels !
Finally Gary your silence is heavy of consequence about the acrylic caps before the LDOs :you don't like it on your system or don't hear a difference with them instead of the 10 uf X7r ? Maybe not have time or will to taste it and now switch to other projects ?
If so, did you try the whole mod I maid and/or could you say if you an input between MKS2 2 mm pitch at C32 against Black-Gate ? You say above you prefer the BG, but my understanding is it was a MKS2 5 mm pitch at C35 and JP seems very confident with the little Wima at C32...maybe both the layout and the inductance but it surprise me if it would be huge difference between the two Wimas ! Don't have suffisant buying at Mouser since my last one to put it on my list !
Thanks again for your help about the shematic.
cheers
Eldam
Peterma,
It's interresting !
Ok the sound is good but di you try several caps to see if difference exist with soundstage before to involve the layout or the dac chip itself !
My experience is there is there are many difference (and not just subtilities= with different caps devices and arrengements if your system has enough resolution. As it seems to be the case with yours : has you on C4 a 1 uf X7r 604 case standalone (no // with an another cap), BNCs on your streamer and the DAC (no RCAs plugs) ?
Can you tell which DAC has more deep soundstage on your system please ?
Eldam,
Please look at the datasheet for the LM723 and study Table 1 and Table 2. Footnote #2 of table 1 explains the changes that I discussed in R1 and R2. Table 2 includes the equations for calculating R1 and R2 for Vout<7v and Vout>7v.
Sometimes silence is just silence.
Regards,
---Gary
: