It is not a competition to make the lowest footprint DAC here with those old chips. Those modern reg chips are needing very low esr caps that are so so on the analog output side of the crcuit IME. Ceramic sounds weird but for values away from the ears frequencies. Also that vey fast regs makes not confortable some op amps.
Plus the old TH caps are often mandatory I have found for setuping a good sounds !
What are according to you the noise floor needed with such PCM chips ? 120 dB at best for the AD1862 / PCM63 ; after it is diminishing return or just unheard ?
I agree for the very low AND flat impedance of regs, but which integrated chips is better than a good designed discrete here ?
But imo it is digression here, certainly asks a thread for benchmark as made Linear Audio from @jan.didden back then with the often used usual supects, imho !
As for shunt vs linear for digital side, not so clear to me if always about the topology of the parts used and the weigth of each characteristics : output, impedance, PSRR, phase, speed, etc
It is not like it was a 100 Mhz crystal modern AK and ESS DAC chips circuitries in this thread ! But the topic is crucial indeed ! For instance I had liked when everyone jumped to the CFA op amp moode, we had compared the other very good VFA op amps with the same Walt Jung supply !
A britanic member who asked me some tips about a TDA1541A DAC is comparing 2 type of discrete regs and parts and I am looking forward to have his conclusions. The shops are full of DACS at 100 bucks that measure sota and sounds bad ! Well some would say as the noise floor is ultra low this is not the 100 bucks DAC that is culpritt but all the rest, the owner included ! I think it is more complex than this ! But okay, we like to focus on what we have a control, it is well known biased behavior than we know and that has to be keeped under control btw when a group of people are involved in a project ! Did you not ever seen in your carrer engineers shooting each others in the company corridor, ahahah ! And the project managers or the executives manager ending the end of the school playground, lol ! 😛
🤣
EDIT : but we should re open really a new topic thread, how interresting subject than power supplies.
Plus the old TH caps are often mandatory I have found for setuping a good sounds !
What are according to you the noise floor needed with such PCM chips ? 120 dB at best for the AD1862 / PCM63 ; after it is diminishing return or just unheard ?
I agree for the very low AND flat impedance of regs, but which integrated chips is better than a good designed discrete here ?
But imo it is digression here, certainly asks a thread for benchmark as made Linear Audio from @jan.didden back then with the often used usual supects, imho !
As for shunt vs linear for digital side, not so clear to me if always about the topology of the parts used and the weigth of each characteristics : output, impedance, PSRR, phase, speed, etc
It is not like it was a 100 Mhz crystal modern AK and ESS DAC chips circuitries in this thread ! But the topic is crucial indeed ! For instance I had liked when everyone jumped to the CFA op amp moode, we had compared the other very good VFA op amps with the same Walt Jung supply !
A britanic member who asked me some tips about a TDA1541A DAC is comparing 2 type of discrete regs and parts and I am looking forward to have his conclusions. The shops are full of DACS at 100 bucks that measure sota and sounds bad ! Well some would say as the noise floor is ultra low this is not the 100 bucks DAC that is culpritt but all the rest, the owner included ! I think it is more complex than this ! But okay, we like to focus on what we have a control, it is well known biased behavior than we know and that has to be keeped under control btw when a group of people are involved in a project ! Did you not ever seen in your carrer engineers shooting each others in the company corridor, ahahah ! And the project managers or the executives manager ending the end of the school playground, lol ! 😛
🤣
EDIT : but we should re open really a new topic thread, how interresting subject than power supplies.
Last edited:
Indeed.
But in general if the goal is to make very compact and small pcb then use IC devices. If it is not a main design concept then go for diskrete elements. My experiences that use a just one modue in regulator is not the solution..
"Compactness" is not a goal, the goal is to improve performance, and typically IC regulators are much better than discrete ones.
Like any discrete OP - they will never be as good at high frequencies as IC OP (which is very important for I/V in DACs, where “compactness” is also useful, because it reduces the area of circuits, potentially sensitive to HF external fields).
Alex.
Like any discrete OP - they will never be as good at high frequencies as IC OP (which is very important for I/V in DACs, where “compactness” is also useful, because it reduces the area of circuits, potentially sensitive to HF external fields).
Alex.
Very interesting.
So can you kindly point us to one, ideally with measurements, that has better performance than an LT3045 at 1MHz ?
What is your definition of high frequencies ?
Thanks,
Patrick
HF performance more depends of a series inductance (beads, etc.) and bypass capacitors after regulator.
Alex.
Alex.
Component arrangement, commons sense check:
1. Transformers away from DAC, usually in line with iec inlet and on off switch. One more R core not shown but will be above the bottom one.
2. Twisted wires throughout
3. I will add CEN IV later and plan to hang it on the back panel beside the output ports.
4. That will add 1 more power supply board.
1. Transformers away from DAC, usually in line with iec inlet and on off switch. One more R core not shown but will be above the bottom one.
2. Twisted wires throughout
3. I will add CEN IV later and plan to hang it on the back panel beside the output ports.
4. That will add 1 more power supply board.
That layout makes sense, it's how I did my two builds of Miro AD1862 dacs and have zero issues with them.
Turn your shunts around 180deg so that the AC wiring can be tucked along the front bottom edge of the case?
It is not the rule... When properly designed, discrete are better. With IC regulators all the elements are exposed to unnwanted heat... 🙁 At least use (old but still excellent) 723 with external pass transistor to keep the heat where it shoud be...
Not true either?
🙁 sorry. All OPs are Push-Pull, complementary and too-much-elements-in small room devices.
With discrete it is possible to design Single ended, very low distortion and low Zout IV conversions. 🙂
But we dont want the super high frequency to pass? 🤢
.
I will post some example tomorrow.
Cheers.
Yes.
Some dual OP have 200-300 transistors inside.
Try do do the same with discrets...
We are talking about tens and hundreds of MHz.
No, we don't want to.
And not only "not to pass”, but also so do not influence to the lower frequencies.
Alex.
Yes, but try first one simple OP with mosfets from Pass article of OP amps. ZVP and ZVN transistors. 🙂
.
But these small ICs are designed exactly for reducing space and make the devices more compact? That is the main goal for IC regulators. You cant put discrete reg in a lap top or mobile phone... 🙁
I still need to choose a regulator for the output stage.
I am intentionally using different regulators just to practice how to assemble them. Shunt and 78xx LDOs are completed.
Would any of TPS7A4701 and a negative equivalent suit the discreet stage in this case?
There are even ready-to-run versions on ali:
https://vi.aliexpress.com/item/4001138509950.html
I am intentionally using different regulators just to practice how to assemble them. Shunt and 78xx LDOs are completed.
Would any of TPS7A4701 and a negative equivalent suit the discreet stage in this case?
There are even ready-to-run versions on ali:
https://vi.aliexpress.com/item/4001138509950.html
I think the dual version would be a good one to go with.
I also like these.
https://a.aliexpress.com/_EzC81wd
The ones with the pass transistor and opamp buffer stage. And it will give you a spare single rail for something else maybe..Robust little things. I either shorted or reversed one a while ago. The pass transistor got so hot it floated away on its molten solder. Re aligned and it worked again!
I also like these.
https://a.aliexpress.com/_EzC81wd
The ones with the pass transistor and opamp buffer stage. And it will give you a spare single rail for something else maybe..Robust little things. I either shorted or reversed one a while ago. The pass transistor got so hot it floated away on its molten solder. Re aligned and it worked again!
