Id doesnT matter at all, because you don't use opamps at all. You should hard-wire DAC's current output to tube grid, and only add an I/V conversion resistor to the ground. And it sounds heavenly.
Hi,
First, zenners or Diodes vary their voltage drop with current flow. The Digital supply on the PCM1742 running at 44.1KHz is 6..10mA, so use a resistor that creates 8mA @ 3.3V to test, that is 410 Ohm, NPV is 430 Ohm, so measure using 430 Ohm to load the PSU.
To be honest, you can easily find 3-pin 3.3V regulators, it seems a little suboptimal trying to wrangle a really bad design into submission, instead of tearing it out and fitting a serious solution in the space.
Two diodes will drop around 1.2V at their rated current (these diodes are serioes and normal, not zenner), so you will get around 3.8V under normal operation (guess) or more...
Why not fit a TL431 as shunt to ground and replace the diodes with resistors, if you insist on keeping this $ 1.35 DAC Chip?
Also, you may wish to check that the Decoder chip feeding this DAC uses 3.3V Logic, not 5V, otherwise you have an even greater issue in the design...
I would suggest taking the route with the PCM1798 DAC Board and passive IV is a good choice.
Incidentally, if you do this, do not feed the DAC via SPDIF.
Ciao T
First, zenners or Diodes vary their voltage drop with current flow. The Digital supply on the PCM1742 running at 44.1KHz is 6..10mA, so use a resistor that creates 8mA @ 3.3V to test, that is 410 Ohm, NPV is 430 Ohm, so measure using 430 Ohm to load the PSU.
To be honest, you can easily find 3-pin 3.3V regulators, it seems a little suboptimal trying to wrangle a really bad design into submission, instead of tearing it out and fitting a serious solution in the space.
Two diodes will drop around 1.2V at their rated current (these diodes are serioes and normal, not zenner), so you will get around 3.8V under normal operation (guess) or more...
Why not fit a TL431 as shunt to ground and replace the diodes with resistors, if you insist on keeping this $ 1.35 DAC Chip?
Also, you may wish to check that the Decoder chip feeding this DAC uses 3.3V Logic, not 5V, otherwise you have an even greater issue in the design...
I would suggest taking the route with the PCM1798 DAC Board and passive IV is a good choice.
Incidentally, if you do this, do not feed the DAC via SPDIF.
Ciao T
You were right on the money - 470R produced a drop to 3.51V
I will put back 1742 without any further tweaking and see if it works. As I mentioned before this player is a BIG WONDER because of its circuit design.
I think the Decoder is NXP P89V51RB2FN that is powered by 5V. Spec sheet says it should have 4.5V low level output (measures 5.08V on mine). Output is fed to DAC chip via 100 Ohm resistors...
(The deeper the forest - the taller the trees)
I bought CS8416 board and will fit it in using existing tube output stage (modified).
How do you suggest I feed the DAC?
Last edited:
Hi,
That may work(ish) but it is really shitty design. The 6mA are listed as typical, so digital supply current may be even lower or as high as 10mA. That is quite irresponsible design, just to shave of a few cent of the BOM.
I am more and more inclined to agree with you. It is an absolute marvel, that it works at all, I mean...
That is the Microprocessor, the decoder is a big SMD chip below the PCB behind the drive.
Incidentally, the PCM1742 Datasheets suggests the inputs are 5V compatible, however this often includes clamping the input to the supply rail, which will throw massive amounts of noise into the digital supply.
CS8416 is a digital receiver (SPDIF to 3/4 Wire serial).
You need a DAC.
You need to tap the 4-Wire signals from the PCM1742 inputs (MCK, BCK, WCK and DATA) and send them to the same pin's on the new DAC Chip. You need to see which input configuration sets the new DAC into the correct data mode.
If the decoder chip is Philips the format is likely I2S while it will be EIJA if it is a Sony or other japanese decoder chip.
Ciao T
That may work(ish) but it is really shitty design. The 6mA are listed as typical, so digital supply current may be even lower or as high as 10mA. That is quite irresponsible design, just to shave of a few cent of the BOM.
I am more and more inclined to agree with you. It is an absolute marvel, that it works at all, I mean...
That is the Microprocessor, the decoder is a big SMD chip below the PCB behind the drive.
Incidentally, the PCM1742 Datasheets suggests the inputs are 5V compatible, however this often includes clamping the input to the supply rail, which will throw massive amounts of noise into the digital supply.
CS8416 is a digital receiver (SPDIF to 3/4 Wire serial).
You need a DAC.
You need to tap the 4-Wire signals from the PCM1742 inputs (MCK, BCK, WCK and DATA) and send them to the same pin's on the new DAC Chip. You need to see which input configuration sets the new DAC into the correct data mode.
If the decoder chip is Philips the format is likely I2S while it will be EIJA if it is a Sony or other japanese decoder chip.
Ciao T
I meant I got DIY DAC board with CS8416+CD4397 (Lampucera). I think I will just run the SPIDF signal from the controller board.
Seeing what I discovered in this player I just can't understand how people could praise it. The failure rate of DAC and other components do to faulty design should be very high. It failed on me and I have a good idea why...Even when it works the voltage oscillations should make it sound like crap smoothed out by tubes working at wrong parameters...
Seeing what I discovered in this player I just can't understand how people could praise it. The failure rate of DAC and other components do to faulty design should be very high. It failed on me and I have a good idea why...Even when it works the voltage oscillations should make it sound like crap smoothed out by tubes working at wrong parameters...
Where do you tap the signal on the "Lampucera" DIY DAC board, could you give more details on how to wire it to the tube stage?
Thanks
D.
Last edited:
I wrote about PCM1798 DAC board, wchich is current output chip.
And You have another one based on CS4397 wchich is voltage output chip. I suggest to use signal transformers to desymmetrize signal from the DAC, followed by a cathode follower existing in this CD player. It will give You about 2V signal level at RCA outputs.
A pair of transformers for 1798 would cost more than the entire DIY DAC kit. It's either 1798 with conversion OpAmps or 4397 tapped directly. I'd say 4397 is most cost effective, besides there are some clear instructions on Lampizator website.
You don't understand. PCM1798 requires no signal transformer at all.
But CS4397 requires. I tried lampizator's site way of connecting tube analogue stage, and it doesn't work good. It gives about 20 (TWENTY!) volts of signal level at the output, and it overdrives almost every solid state amplifier.
Last edited:
Yes, PCM1798 is a current output, balanced DAC. It doesn't require any signal transformer. It requires a shunt resistor for I/V conversion. And you can tap it's output directly to the tube grid.
Since you've installed this board in several CD players maybe you have a page with description and schematics on how to connect Lampucera PCM1798 board to cathode follower?
I am trying to decide what to do with my MHZS 66 machine:
1. Repair it with 1742 and sell it
2. use it as a transport
3. put new DAC board in it (need some help)
4. Use it as a door stop.
If I can suggest, repair it with PCM1742 and sell it. Then build a standalone DAC based on PCM1798 PCB.
The standalone DAC gives you possibility to play not only compact discs, but also a music files from your computer and a sound from your DVD/blue ray player or television. The DAC is far more universal and flexible. And my DIY DAC based on PCM1798 chip and tube output stage sounds just marvellous and I don't want any other device.
I am going to try and put in a separate 3.3V regulator (78L33 with couple caps) for DAC Vdd+ and see what happens. It should improve things a bit.
Just don't understand why people at MHZS didn't put $0.60 part into this critical circuit.
Hi,
The CS8416 is probably the worst jitter performance SPDIF Receiver Chip. If you want to use this board and SPDIF you should put the CS8416 into slave mode and feed it the bit and wordclocks from the CD-Player, plus the MCK from the CD-Player direct to the DAC Chip. Otherwise you are looking at very poor jitterperformance.
I still think you are better off using the 4-Wire serial signal (often called I2S even if the format is not) instead of SPDIF.
Ciao T
The CS8416 is probably the worst jitter performance SPDIF Receiver Chip. If you want to use this board and SPDIF you should put the CS8416 into slave mode and feed it the bit and wordclocks from the CD-Player, plus the MCK from the CD-Player direct to the DAC Chip. Otherwise you are looking at very poor jitterperformance.
I still think you are better off using the 4-Wire serial signal (often called I2S even if the format is not) instead of SPDIF.
Ciao T
Hi,
I would suggest:
1), followed by 4), followed by 3) using 4-Wire signals followed by 2).
Ciao T
I would suggest:
1), followed by 4), followed by 3) using 4-Wire signals followed by 2).
Ciao T
Hi,
You may not need extra cap's, there are ones already on the 5V and 3.3V line quite close by.
I might worry about the voltage though, have a look for a "low dropout type", like LM1085-3.3
To them it would have probably been a 6 cent part, but the diodes are under 1 cent each. I would suspect if you spend enough time you can find more such instances of "cost savings"...
Ciao T
You may not need extra cap's, there are ones already on the 5V and 3.3V line quite close by.
I might worry about the voltage though, have a look for a "low dropout type", like LM1085-3.3
To them it would have probably been a 6 cent part, but the diodes are under 1 cent each. I would suspect if you spend enough time you can find more such instances of "cost savings"...
Ciao T
Took out two diodes and replaced them with voltage regulator for Vdd - got stable +3.3V for DAC Vdd, replaced filtering lytic caps with oscons and nichicon muse, added wima bypass caps to both +5 and +3.3 supplies - no more crazy oscillations. The player no sounds really good, even with the original (although replaced) DAC.
will post the pics later...
will post the pics later...
Last edited: