definitively not! look at left and right channel as two identical DACs. They both need the same signal structure. One signal should contain left data and one right data...
Within the 64 clock cycles of one "FS" like 44.1Khz or 96kHz related sample time, you will need to right justify the 24 bits data complete to the right: so 7 clocks delay for R and 39 clocks delay for L
now you need to make sure the data arrives "on time"
that's all Folks
Thanks again Doede
Jean-Denis, yes! I fitted them only last week, but I'm very pleased with the sound
Rhopoint GG102D 120R coming out, Tx2725 133r going in.
Jean-Denis, yes! I fitted them only last week, but I'm very pleased with the sound
An externally hosted image should be here but it was not working when we last tested it.
Rhopoint GG102D 120R coming out, Tx2725 133r going in.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I’ve made some more changes to my unregulated power supply. Up till now I had been playing with a bridge rectifier and was skeptical about the possible improvement of a two diode full wave rectifier. However, after I bit the bullet to upgrade my toroidal transformer to a R-core transformer and learned what difference this made, I got to wonder. A two diode full wave rectifier puts a much bigger demand on the transformer as it draws considerably more current in comparison to a Bridge, for the same DC output. It requires a centre tap transformer, or one with equal secondaries that can be used to make a centre tap with.
First, I noticed that the transformers heats up a lot more, it’s a shame to waste power on heat, but I understand transformers “like” a bit of a load. I measured power input at the powersocket (for all the components in my case), before it used to draw 27Watt and after this rectifier change it is drawing 40Watt, which is still reasonable. Also my voltage after choke and what not, at the dac, has gone up by 0.5Vdc, which might be usefull for some of you who are on the low side.
But.., I must say the sound has come alive to another level. The box standard DDDAC is no slouch in terms of dynamics, but it’s gone through the roof now. Things like decaying sounds seem to last forever and spacial presence is intense. It’s almost too much of a good thing, as if I need to find a way to reign it back in a bit.
Food for thought.
First, I noticed that the transformers heats up a lot more, it’s a shame to waste power on heat, but I understand transformers “like” a bit of a load. I measured power input at the powersocket (for all the components in my case), before it used to draw 27Watt and after this rectifier change it is drawing 40Watt, which is still reasonable. Also my voltage after choke and what not, at the dac, has gone up by 0.5Vdc, which might be usefull for some of you who are on the low side.
But.., I must say the sound has come alive to another level. The box standard DDDAC is no slouch in terms of dynamics, but it’s gone through the roof now. Things like decaying sounds seem to last forever and spacial presence is intense. It’s almost too much of a good thing, as if I need to find a way to reign it back in a bit.
Food for thought.
Last edited:
Nice!
I'm nearly done with mine
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Our supplies are turning out very similair. I like the rubber spacers, very nice. I'm thinking about putting mine on felt. Although, I think you have a single secondairy, so full bridge, (right?). What's the black wire going back to the R-Core?
Just a thought. I wonder if it would be better to put the bleeder after the second choke, so it gets a bit more current going though it. I've not tried your setup though, so just speculating.
Last edited:
Why the bleeder between the Lundahls and not after James ?
I have it after both Lundahls . Have never had better sound than now with the finished cLCLC choke supply. With 15v Trafo i needed a small cap ( 220uf )after the Shottcky rectifiers and before the first choke. Without this cap voltage was a little bit to low.
My plan was to copy Supersurfers psu, but since I don't have the shunt regs on my dac with 4 boards I got 16vdc at output. Have to replace the two 2200uf caps before the first choke with a 220uf cap and a bleeder ( 47R ) after both chokes,,voila,,,,12vdc.
Some say the first cap(s) in a clclc setup must be the biggest, maybe this is correct but my dac sings as never before now. I will bring it over to a friend who has a scope to see if there is any ripple. Honestly, I doubt it. It just sounds to good.
I have it after both Lundahls . Have never had better sound than now with the finished cLCLC choke supply. With 15v Trafo i needed a small cap ( 220uf )after the Shottcky rectifiers and before the first choke. Without this cap voltage was a little bit to low.
My plan was to copy Supersurfers psu, but since I don't have the shunt regs on my dac with 4 boards I got 16vdc at output. Have to replace the two 2200uf caps before the first choke with a 220uf cap and a bleeder ( 47R ) after both chokes,,voila,,,,12vdc.
Some say the first cap(s) in a clclc setup must be the biggest, maybe this is correct but my dac sings as never before now. I will bring it over to a friend who has a scope to see if there is any ripple. Honestly, I doubt it. It just sounds to good.
James, Great ! If you can, tell us how do they compare to Rhopoints...
Great case work too, I hope mine turns out that nice.
Smooth Dancer : great upgrade of yours. Sad you don't have enough voltage to use it as a true choke input, as it's even more relaxed and real without a cap after the diodes (at least in my experience).
Great case work too, I hope mine turns out that nice.
Smooth Dancer : great upgrade of yours. Sad you don't have enough voltage to use it as a true choke input, as it's even more relaxed and real without a cap after the diodes (at least in my experience).
Thanks again Doede
Jean-Denis, yes! I fitted them only last week, but I'm very pleased with the sound
An externally hosted image should be here but it was not working when we last tested it.
Rhopoint GG102D 120R coming out, Tx2725 133r going in.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I did exactly the opposite with my AN DAC 4.1x and didn't regret it at all. Too "sterile" sounding with the TX2725.... But in this DAC with those mods and for your taste, it can be just right.
Last edited:
Hello everyone,
Speaking of rectifiers. Supersurfer, would you be so kind to post some scratch on how to connect four Shottkey diodes as a full wave bridge using two transformers. I did it with one transformer and that is easy. Just follow the full bridge scheme. But I'm not getting to the point how to make it using two trafos and ending with one output line.
Is it similar to this scheme, except instead center tap I use two shorted negative lines from both trafos?
What vantages are using this sort of supply?
Speaking of rectifiers. Supersurfer, would you be so kind to post some scratch on how to connect four Shottkey diodes as a full wave bridge using two transformers. I did it with one transformer and that is easy. Just follow the full bridge scheme. But I'm not getting to the point how to make it using two trafos and ending with one output line.
Is it similar to this scheme, except instead center tap I use two shorted negative lines from both trafos?
An externally hosted image should be here but it was not working when we last tested it.
What vantages are using this sort of supply?
Last edited:
Or it is just as simple as that? Using two diodes is enough? Sorry just noobish questions.
Two diodes full wafe bridge.png - DivShare
Two diodes full wafe bridge.png - DivShare
I cannot see how the 2 diode full wave rectifier can be better than a 4 diode bridge taken in its narrowest sense. It must be either a misperception ( we all do this inadvertently) or some other incidental change in the actual configuration in the circuit.
I know transformers can be subject to a little mechanical vibration at times, so I thought I'd mount them in such a way that it's not transmitted to anything else.
The black wire is the centre tap, I've got 2 diodes on that tag strip.
I can't remember who suggested the bleeder after the first choke.... Probably Eduard? It made sense to me at the time, but was a while ago...
I only have a 12v trafo at the moment. I noted that you can raise the voltage a little by adding a small cap before the first choke, but I checked the Morgan Jones Valve Amplifiers book and he shows that using this method can increase the impedance of the supply quite a lot and part of my aim was to make this very low impedance.
It's 2 of 1NGXA288 which is 80mm high, 230mm wide and 280mm deep with a 10mm aluminium faceplate.
I looked at loads of other cases, but I'm really pleased with these and would definitely recommend them
It's definitely a bright and detailed sound, but it seems to work well as a whole and although I haven't listened for long periods, it seems it's not fatiguing, so I'm happy for nowI did exactly the opposite with my AN DAC 4.1x and didn't regret it at all. Too "sterile" sounding with the TX2725.... But in this DAC with those mods and for your taste, it can be just right.
Yeah, it was fine with 120R, it just means the output is a little quieter compared to 133R. Doede says you can go below a little, but not above as that will risk distortion.
Yes indeed
I did a lot of reading around looking at power supplies and I read a few times about people preferring the 2 diode sound. For example Martin Clarke here: Modifying and building audio power supplies
I guess it's less diodes to make nasty diode noise and less in the ground line as Martin says.
Here's a VERY rough sketch of the layout of my supply.
I haven't put in the snubber (in blue) across the transformer secondaries yet as I want to build a Cheapmodo circuit to work out the best values to use.
Each Lundahl LL1694 choke has 2 distinct halves to them, so there's a few ways you can wire them.
I wired the first choke in serial mode, and the second one in common mode rejection mode. My scope shows that only using half of the second choke is sufficient for removing ripple in the supply. The first choke seems to do almost all the work of smoothing the supply.
I've done 2 star grounds: a dirty one which contains most of the charge pulses and noisy stuff, and a second, clean one for connecting to the audio circuits, with the half of the choke between them. I've connected the gnds from my other supplies like the S03 reclocker and the bufffer stage to the clean star, but not the one for the BBB/rPi as that would effectively bridge the isolator in the S03.
I'm no expert and I'm certainly not saying this is the best magic solution for everybody, but it works well for me. It's very quiet in terms of background noise (especially now it's in a different case) and I like the sound
Last edited:
James, Curious to learn what snubber you'll come up with, I'm using a few, 22nF's after the diodes and allong with the caps..., seems to take a bit of the course edge of the highs. I've come back on using the common mode on the chokes. Contra to what I expected, the move from (one choke) common back to serial mode, made the SQ a bit more relaxed and controlled. I'm starting to think there is some sense in keeping the impedence back to ground as low as possible, but the quality of the transfomer has big role to play in this.