Hi John
The D-I dac is rather complicated yes, but there is no other way with that 1541 (clock routing, dem reclock, lot of caps, 3 PS's) But to simplify it could be build in a standard housing (Monacor?) and with only the opamp output. I'll stick to 1541. Not to say i don't use 1543's, today i listened a 1543
What will a complete (octal 1541 D-I dac) pcb kit cost roughly?
The D-I dac is rather complicated yes, but there is no other way with that 1541 (clock routing, dem reclock, lot of caps, 3 PS's) But to simplify it could be build in a standard housing (Monacor?) and with only the opamp output. I'll stick to 1541. Not to say i don't use 1543's, today i listened a 1543
What will a complete (octal 1541 D-I dac) pcb kit cost roughly?
Hi dddac,
Thanks for your reply [post #1035]
Yes, TDA1543 based DAC's can sound very good indeed if more DAC chips are used. As for passive I/V I already tried some setups but I got mixed feelings about it. Either the op-amp I/V stage I am currently using is performing exeptionally well, or passive I/V has it's drawbacks. As I noted before, with the passive I/V the fast attack times and brilliance were gone, but sound was very smooth and natural. After a lot of listening sessions, the op-amp I/V produced the most realistic sound. This was confirmed again yesterday as I visited the VAD audioshow in Veldhoven, listening to a live performance.
Thanks for your reply [post #1035]
Yes, TDA1543 based DAC's can sound very good indeed if more DAC chips are used. As for passive I/V I already tried some setups but I got mixed feelings about it. Either the op-amp I/V stage I am currently using is performing exeptionally well, or passive I/V has it's drawbacks. As I noted before, with the passive I/V the fast attack times and brilliance were gone, but sound was very smooth and natural. After a lot of listening sessions, the op-amp I/V produced the most realistic sound. This was confirmed again yesterday as I visited the VAD audioshow in Veldhoven, listening to a live performance.
Hi maxlorenz,
Thanks for your reply [post #1039]
Yes I am planning to design a single board version TDA5143 D-I DAC, it will have 16 DAC's in balanced D-I mode (virtual sample rate 705.6 KHz, theoretical 20 bits and a resolution enhancement of 256. It will have a DC coupled output as standard (no capacitors, coils or transformers in the signal path). Like the octal D-I DAC it will have both USB and differential I2S inputs.
The DAC chips will be arranged in a 4 X 4 matrix. If more TDA1543's are to be used, it makes more sense to arrange them all in balanced D-I mode. Since each TDA1543 needs a different WS and DATA signal, stacking them would become problematic anyway. In time I could consider designing a 64 X TDA1543 DAC with the DAC chips arranged in a 8 X 8 matrix providing a virtual sample rate of 2.8 MHz, theoretical 22 bits and a resolution enhancement of 4096.
As for passive I/V, I haven't managed to get distortion figures anywhere near the extreme low values obtained with the tuned high-speed opamp I/V. Perhaps the introduced distortion with passive I/V sounds ok, but I like to keep distortion as low as possible. The 25mV output voltage compliance seems to be the absolute maximum when very low distortion is desired.
Maybe it's a incorrect approach, but I prefer a plain wire (DC coupled) instead of coupling cap's, coils or transformers.
Thanks for your reply [post #1039]
Yes I am planning to design a single board version TDA5143 D-I DAC, it will have 16 DAC's in balanced D-I mode (virtual sample rate 705.6 KHz, theoretical 20 bits and a resolution enhancement of 256. It will have a DC coupled output as standard (no capacitors, coils or transformers in the signal path). Like the octal D-I DAC it will have both USB and differential I2S inputs.
The DAC chips will be arranged in a 4 X 4 matrix. If more TDA1543's are to be used, it makes more sense to arrange them all in balanced D-I mode. Since each TDA1543 needs a different WS and DATA signal, stacking them would become problematic anyway. In time I could consider designing a 64 X TDA1543 DAC with the DAC chips arranged in a 8 X 8 matrix providing a virtual sample rate of 2.8 MHz, theoretical 22 bits and a resolution enhancement of 4096.
As for passive I/V, I haven't managed to get distortion figures anywhere near the extreme low values obtained with the tuned high-speed opamp I/V. Perhaps the introduced distortion with passive I/V sounds ok, but I like to keep distortion as low as possible. The 25mV output voltage compliance seems to be the absolute maximum when very low distortion is desired.
Maybe it's a incorrect approach, but I prefer a plain wire (DC coupled) instead of coupling cap's, coils or transformers.
Octal D-I DAC redesigned
Hi Tubee,
Thanks for your reply [post #1041]
Yes, the octal D-I DAC prototype is rather complicated and expensive, but that's going to be changed drastically. I am completely redesigning both PCB's and housing to significantly reduce cost and simplify the construction. I know this will delay PCB production, and I apologize for that, but it will certaily be worth waiting for.
So here is what I have been up to lately, The DAC, I/V diff amp, timing-chain and analog mainboard have been significantly reduced in size. Routing of these boards has already been completed. The result is a very compact analog mainboard, with even better specifications. The new analog mainboard only measures 133 X 219 mm !
The autodetecting is dropped for now, and the Philips format has been used as standard, the new timing-chain is highly optimized and has 11 chip's. But that's not all, it also includes the DEM clock circuitry and the UHS buffers for BCK. This was possible by using one programmable logic device (PALCE16V8). By changing the PAL program, the timing-chain can now be used for either balanced operation (D2,4,6,8 inverted) or single ended operation (none of the D outputs inverted).
Next step is to change the tube-amp module dimensions, the idea is to place all modules directly on the bottom plate, greatly simplifying housing construction. The USB interface will be slightly modified to provide full galvanic insulation.
The complicated backlit pushbutton design will be replaced by a remote control, there will only be one small system controller PCB with integrated indicators and remote control sensor.
Finally a new compact housing design has to be made, I try to make this housing even more beautiful than the prototype housing. It will be a high-quality annodized aluminum housing with high-gloss stainless steel parts. Cost will be reduced even more by using standard aluminum parts, this will significantly reduce milling time. The housing can be annodized in various colours or colour combinations.
A monacor housing for the octal D-I DAC ? oh horror. If you could only hear it, you would realize this DAC deserves a beautiful high-quality housing. Would you prefer off the shelf speaker housings instead of your beautifully handcrafted speaker enclosures?
Hi Tubee,
Thanks for your reply [post #1041]
Yes, the octal D-I DAC prototype is rather complicated and expensive, but that's going to be changed drastically. I am completely redesigning both PCB's and housing to significantly reduce cost and simplify the construction. I know this will delay PCB production, and I apologize for that, but it will certaily be worth waiting for.
So here is what I have been up to lately, The DAC, I/V diff amp, timing-chain and analog mainboard have been significantly reduced in size. Routing of these boards has already been completed. The result is a very compact analog mainboard, with even better specifications. The new analog mainboard only measures 133 X 219 mm !
The autodetecting is dropped for now, and the Philips format has been used as standard, the new timing-chain is highly optimized and has 11 chip's. But that's not all, it also includes the DEM clock circuitry and the UHS buffers for BCK. This was possible by using one programmable logic device (PALCE16V8). By changing the PAL program, the timing-chain can now be used for either balanced operation (D2,4,6,8 inverted) or single ended operation (none of the D outputs inverted).
Next step is to change the tube-amp module dimensions, the idea is to place all modules directly on the bottom plate, greatly simplifying housing construction. The USB interface will be slightly modified to provide full galvanic insulation.
The complicated backlit pushbutton design will be replaced by a remote control, there will only be one small system controller PCB with integrated indicators and remote control sensor.
Finally a new compact housing design has to be made, I try to make this housing even more beautiful than the prototype housing. It will be a high-quality annodized aluminum housing with high-gloss stainless steel parts. Cost will be reduced even more by using standard aluminum parts, this will significantly reduce milling time. The housing can be annodized in various colours or colour combinations.
A monacor housing for the octal D-I DAC ? oh horror. If you could only hear it, you would realize this DAC deserves a beautiful high-quality housing. Would you prefer off the shelf speaker housings instead of your beautifully handcrafted speaker enclosures?
Hi Ecdesigns,
Thanks for your reply.
You know I'm only an amateur, but the idea is adding more DAC chips to existing balanced D-I mode configuration: each stock chip`with its own "tower" of "n" chips, with shared WS and DATA. The goal (ask Doede better) is to increase I and be able to reduce Rload and avoid differential opamp stage (amongst other) to feed directly a transformer (in my case a TVC). The philosophy is using the least active parts on the output stage that I can.
Besides, I recently stole my Dad's Lenco L75 and I'm enjoying his record collection.
The sound that best approaches it in "energy", liveliness and analog warmth is my modded DDDAC. That's why i want to test the previous idea.
Indeed, winter is very cold here!
Don't worry for the delay. We value your time and effort.
Cheers,
M
PS: with passive I/V, sound can be tailored somehow with different R, as we discoused previously. For example, PRP sound balanced, detailed and controlled. Riken sound as the above but with added midrange warmth.
Thanks for your reply.
You know I'm only an amateur, but the idea is adding more DAC chips to existing balanced D-I mode configuration: each stock chip`with its own "tower" of "n" chips, with shared WS and DATA. The goal (ask Doede better) is to increase I and be able to reduce Rload and avoid differential opamp stage (amongst other) to feed directly a transformer (in my case a TVC). The philosophy is using the least active parts on the output stage that I can.
Besides, I recently stole my Dad's Lenco L75 and I'm enjoying his record collection.
The sound that best approaches it in "energy", liveliness and analog warmth is my modded DDDAC. That's why i want to test the previous idea.
Indeed, winter is very cold here!
Don't worry for the delay. We value your time and effort.
Cheers,
M
PS: with passive I/V, sound can be tailored somehow with different R, as we discoused previously. For example, PRP sound balanced, detailed and controlled. Riken sound as the above but with added midrange warmth.
I've been following this thread for some time and perhaps I've missed this, in any case I have a couple of, possibly pretty dumb, questions:
1. Can the DAC be configured for balanced output?
2. Is the DAC restricted to accepting 44.1Khz and 48Khz input rates or can it handle higher rates such as 24/bit 96Khz from from SACD and DVD-A players?
1. Can the DAC be configured for balanced output?
2. Is the DAC restricted to accepting 44.1Khz and 48Khz input rates or can it handle higher rates such as 24/bit 96Khz from from SACD and DVD-A players?
Ohh my what a wonderfull thread.
What about Jfet output stage with a Power Jfet output. Would that be possible. I have great faith in the Tube output, but I have a little trouble as they seem to sound different with the load they see. To me an optimum will be a simple Jfet I/V converter. Like the pass D1 but with Jfets and Pjfet as output.
Cheers Michael Børresen
What about Jfet output stage with a Power Jfet output. Would that be possible. I have great faith in the Tube output, but I have a little trouble as they seem to sound different with the load they see. To me an optimum will be a simple Jfet I/V converter. Like the pass D1 but with Jfets and Pjfet as output.
Cheers Michael Børresen
I'am shure thet the Pass I/V converter would fir perfectly. I just have a little trouble with the vertical fets used in his simple and elegant design. My thought was to make sonehing similar but using jfets or preferably the Pjfet for an output with high current capacity and the ablilty to drive lower impedances. I am shure this can be done. Also this will altogether reduce overall costs as it will do it avay withe the PSU's for the tubestage. I have heard the D1 with vertical fets compared to a state of the art tube I/V converter built with extreme choke bases PSU's. The D1 was not short of. And I belive that with the resoulution and sweetness of jfets its possible to make something that betters the performance.
That's the beauty of it and I'm sure John encourages experimenting with diffrent I/V and output sections to our liking
Dear webchimp,
I didn't mean to be rude or something. In fact your self description suits me well also.
That said, sooner or later one has to take the time and read the data (not that I understand everything neither; I hope rbfw is not around ) to learn and to stop being a burden to all our fine mates.
(some digital experts are very nervous )
That is what I'll do before ordering/building.
Best regards,
M
Dear webchimp,
I didn't mean to be rude or something. In fact your self description suits me well also.
That said, sooner or later one has to take the time and read the data (not that I understand everything neither; I hope rbfw is not around
) to learn and to stop being a burden to all our fine mates. (some digital experts are very nervous
)That is what I'll do before ordering/building.
Best regards,
M
Hi John.
Nice you redesign the mainboard, it could fit in my CDP then. I found first design rather big.
I like the use of SMD parts. (look into a Teac Esoteric: a lot of SMD) Short paths and lower inductancies all over the circuit. Do you use some SMD parts too now? I don't mind to solder in SMD.
Yes you're right, but if some one wants to listen to a great sound and does'nt bother about the design of the dac, it could be done.
Nice you redesign the mainboard, it could fit in my CDP then. I found first design rather big.
I like the use of SMD parts. (look into a Teac Esoteric: a lot of SMD) Short paths and lower inductancies all over the circuit. Do you use some SMD parts too now? I don't mind to solder in SMD.
Yes you're right, but if some one wants to listen to a great sound and does'nt bother about the design of the dac, it could be done.