Hi Zero One,
Thanks for your input.
I have verified that EC was correct in every tweak and mod he has proposed to us until now. Plus his descriptions about the sonic effect of those is very accurate.
For example, the other day I attenuated BCK signal (the only one not attenuated before in my D1M project) but chose the wrong resistor values and used a small inductor: the sound was very bad and I removed the circuit when I realized my error. Last night I re-did it with lower resistor values and a DIY choke with 43 material core and OCC copper. I doubt I reached the 10mH that EC recommends but anyway the sound now improved (even that the circuit is attached with alligator clips). It became "more refined" as EC said. Now I think I will become a fan of chokes!
I have still to test it with the scope.
Regards,
M
Thanks for your input.
I have verified that EC was correct in every tweak and mod he has proposed to us until now. Plus his descriptions about the sonic effect of those is very accurate.
For example, the other day I attenuated BCK signal (the only one not attenuated before in my D1M project) but chose the wrong resistor values and used a small inductor: the sound was very bad and I removed the circuit when I realized my error. Last night I re-did it with lower resistor values and a DIY choke with 43 material core and OCC copper. I doubt I reached the 10mH that EC recommends but anyway the sound now improved (even that the circuit is attached with alligator clips). It became "more refined" as EC said. Now I think I will become a fan of chokes!
I have still to test it with the scope.
Regards,
M
Hi Zero One,
http://www.tnt-audio.com/clinica/regulators_noise4_e.html
The main advantage of batteries is that they provide a local current loop (picking-up far less interference) as these are basically "floating" power supplies.
I constructed both charge-transfer and floating charge-transfer power supplies for similar reasons.
http://www.tnt-audio.com/clinica/regulators_noise4_e.html
The main advantage of batteries is that they provide a local current loop (picking-up far less interference) as these are basically "floating" power supplies.
I constructed both charge-transfer and floating charge-transfer power supplies for similar reasons.
HI ECDesigns,
With regards to the transimpedance circuit, T1 provides virtual ground through the Vref at T1 base.
However, the Idac comes from the sum of Ic and Ib of T1. Ibase of T1 can vary between 0 to 13uA for a Ic of 0 to 4 mA (from the datasheet of 2sk2240) . Therefore Ibase is introducing 'noise' into the circuit depending on the magnitud of Idac.
I wonder if you have any idea how to overcome this.
I was thinking to use JFET given the fact JFET gate current max out at 1uA.
With regards to the transimpedance circuit, T1 provides virtual ground through the Vref at T1 base.
However, the Idac comes from the sum of Ic and Ib of T1. Ibase of T1 can vary between 0 to 13uA for a Ic of 0 to 4 mA (from the datasheet of 2sk2240) . Therefore Ibase is introducing 'noise' into the circuit depending on the magnitud of Idac.
I wonder if you have any idea how to overcome this.
I was thinking to use JFET given the fact JFET gate current max out at 1uA.
FQ Tracker
Hi John,
the Memory Card Player is a great idea. I am still interested in a computer based DAC with Toslink interface to Airport Express just for convenience.
How does asynchronous reclocking perform? How big is the buffer of CS141x ICs and how often can buffer overflow/underflow happen? Is that sample drop out/repeat in practice audible?
How well does the frequency tracker in comparison perform?
And the USB interface like used in DI16DAC?
Currently You use one single TDA1543 in the Card Player. Did You test a single TDA1541A too? I tried the new method to implement a TDA1543 too, but like better my implementation of TDA1541A. Which one is your favorite for the ultimate card player?
Hi John,
the Memory Card Player is a great idea. I am still interested in a computer based DAC with Toslink interface to Airport Express just for convenience.
How does asynchronous reclocking perform? How big is the buffer of CS141x ICs and how often can buffer overflow/underflow happen? Is that sample drop out/repeat in practice audible?
How well does the frequency tracker in comparison perform?
And the USB interface like used in DI16DAC?
Currently You use one single TDA1543 in the Card Player. Did You test a single TDA1541A too? I tried the new method to implement a TDA1543 too, but like better my implementation of TDA1541A. Which one is your favorite for the ultimate card player?
There is 550mV at the DAC output on your diagram, and the datasheet prescribes a voltage compliance of +/-25mV. Isn't there any mistake?
BTW, how rigid is the voltage compliance value? I mean there is nominal distortion at +/-25mV, slightly more at +/-30mV, gross distortion at +/-50mV, or is there a kind of hard clipping at a given voltage level? I suppose there are some internal junction diodes, and the distortion characteristics follows some diode law. Has anybody measured the distortion vs. output voltage? It is an easy measurement, the variable parameter is the resistance of the I/V resistor. I always believed it should be around 10 ohms or below.
BTW, how rigid is the voltage compliance value? I mean there is nominal distortion at +/-25mV, slightly more at +/-30mV, gross distortion at +/-50mV, or is there a kind of hard clipping at a given voltage level? I suppose there are some internal junction diodes, and the distortion characteristics follows some diode law. Has anybody measured the distortion vs. output voltage? It is an easy measurement, the variable parameter is the resistance of the I/V resistor. I always believed it should be around 10 ohms or below.
Hi oshifis,
Grossly exceeding output compliance results in asymmetrical clipping (approx. 800mVpp), at lower levels it's merely a bias problem (asymmetric output signal) causing increased distortion as specified in the data sheet. This is caused by TDA1541A output circuit (diode-transistor bit switches).
The bias circuit around R2 and R3 fixes this, and allows for hihger ac output amplitude while maintaining low THD. The actual passive I/V resistor consists of both R1 and R2 in series (136 Ohm), providing approx. 544mVpp. So R2 is part of both, I/V resistor and bias circuit.
The resistive bias circuit adds noise, this ends-up at the I/V resistor, and the signal does look noisy on the oscilloscope. But the same happens in the JFET amplifier stage, but now inverted as the JFET amp inverts. So now it's possible to neutralize both effects, leading to a clean output signal despite power supply noise levels (increased PSRR).
So the 544mV ac signal is no mistake for sure.
I attached some oscillograms, the one on the left shows how the distortion looks like when ac signal amplitude is too high (I/V resistor tied to GND), this is a 450mVpp signal, on the right a 540mVpp signal, also on the DAC output, but now using the additional bias circuit.
The distortion sets-in slowly, starting with distortion (soft clipping) at the lower part of the 1 KHz 0dB test sine wave. When the bias circuit is adapted, and I/V resistor value is significantly increased, the signal clips symmetrical at approx. 800mVpp.
Grossly exceeding output compliance results in asymmetrical clipping (approx. 800mVpp), at lower levels it's merely a bias problem (asymmetric output signal) causing increased distortion as specified in the data sheet. This is caused by TDA1541A output circuit (diode-transistor bit switches).
The bias circuit around R2 and R3 fixes this, and allows for hihger ac output amplitude while maintaining low THD. The actual passive I/V resistor consists of both R1 and R2 in series (136 Ohm), providing approx. 544mVpp. So R2 is part of both, I/V resistor and bias circuit.
The resistive bias circuit adds noise, this ends-up at the I/V resistor, and the signal does look noisy on the oscilloscope. But the same happens in the JFET amplifier stage, but now inverted as the JFET amp inverts. So now it's possible to neutralize both effects, leading to a clean output signal despite power supply noise levels (increased PSRR).
So the 544mV ac signal is no mistake for sure.
I attached some oscillograms, the one on the left shows how the distortion looks like when ac signal amplitude is too high (I/V resistor tied to GND), this is a 450mVpp signal, on the right a 540mVpp signal, also on the DAC output, but now using the additional bias circuit.
The distortion sets-in slowly, starting with distortion (soft clipping) at the lower part of the 1 KHz 0dB test sine wave. When the bias circuit is adapted, and I/V resistor value is significantly increased, the signal clips symmetrical at approx. 800mVpp.
Attachments
Thanks for the explanation. One thing is not quite clear: is your bias circuit providing the 2 mA DC bias current flowing into the DAC output pin? If this is the case, Philips could have relaxed the +/-25mV restriction on their data sheet, in the special case when an external 2mA bias is applied at the output pin.
u can browse his website
http://www.ecdesigns.nl/
and can pm him for the products that suit your need.
http://www.ecdesigns.nl/
and can pm him for the products that suit your need.
Re: Speaker attenuation experience
Hi Zero One, John
i had similar experiences.
It is easy to test how speaker attenuator works.
I tried with 100 ohms 3-4W mono wirewound potentiometer. Just connected between speaker + terminal and speaker cable.
Thats it.
Sound is MUCH better compare to volume control unit with Vishay bulk foil resistors.
Even in low levels sound is full, live, clean, especially dynamic is increased.
So in future i would go for remote relay speaker attenuator with Mills 5W noninductive resistors.
Thanks Zero One and John for sharing experiences.
I now what John put on thread is based on many measurents, listenings, also brothers sugestions with his experiences of living concerts.
It concern also DEM mods and attenuator for DI4T, which works really good.
Zero One said:
Hi Zero One, John
i had similar experiences.
It is easy to test how speaker attenuator works.
I tried with 100 ohms 3-4W mono wirewound potentiometer. Just connected between speaker + terminal and speaker cable.
Thats it.
Sound is MUCH better compare to volume control unit with Vishay bulk foil resistors.
Even in low levels sound is full, live, clean, especially dynamic is increased.
So in future i would go for remote relay speaker attenuator with Mills 5W noninductive resistors.
Thanks Zero One and John for sharing experiences.
I now what John put on thread is based on many measurents, listenings, also brothers sugestions with his experiences of living concerts.
It concern also DEM mods and attenuator for DI4T, which works really good.
Good morning a333BT
The more I look at ECs ideas and implementation and my own experience the more convinced I am it is the way to go for best sound, re Dacs, volume control, and I would say the SD card player and of course battery power.
Over the last couple of days I have been sorting through different settings on my system and A/Bing between using the pre amp with its volume control and direct buffered input into the power amp modules with L pads controlling the show.
The difference is really quite stunning, as EC says the amps end up sounding pretty much like class A, smooth and detailed, controlled....just lovely.
The result are the same playing from vinyl via the phono pre direct as well, in fact the difference is probably even greater.
Yesterday I had a friend over for couple of hours to audition the system, he was in his own words "utterly amazed" and I have some other guys in over the next week or so for a listen so it will be interesting to get their feedback.
The really great thing is that we have people on these forums who are actually prepared to share for all of us to benefit when they could just shut up and reap commercial profits, so a big thanks to EC.
For the moment I will apply the EC mods to my basic NOS dac and hopefully somewhere down the track be able to buy one of ECs card players.
I too would love to be able to buy the player without the case if possible, basically because I would need to junk the case so I could integrate it into the rest of the system which all has my own custom built casework.
I like others here look forward to the pricing of the player.
The more I look at ECs ideas and implementation and my own experience the more convinced I am it is the way to go for best sound, re Dacs, volume control, and I would say the SD card player and of course battery power.
Over the last couple of days I have been sorting through different settings on my system and A/Bing between using the pre amp with its volume control and direct buffered input into the power amp modules with L pads controlling the show.
The difference is really quite stunning, as EC says the amps end up sounding pretty much like class A, smooth and detailed, controlled....just lovely.
The result are the same playing from vinyl via the phono pre direct as well, in fact the difference is probably even greater.
Yesterday I had a friend over for couple of hours to audition the system, he was in his own words "utterly amazed" and I have some other guys in over the next week or so for a listen so it will be interesting to get their feedback.
The really great thing is that we have people on these forums who are actually prepared to share for all of us to benefit when they could just shut up and reap commercial profits, so a big thanks to EC.
For the moment I will apply the EC mods to my basic NOS dac and hopefully somewhere down the track be able to buy one of ECs card players.
I too would love to be able to buy the player without the case if possible, basically because I would need to junk the case so I could integrate it into the rest of the system which all has my own custom built casework.
I like others here look forward to the pricing of the player.
asking about DAC design available
Hi Ecsdesign
Before all the discussion go to long
1
What is your DAC design available for us [ DIY Audio Member ]
you got a lot of fans and still waiting for the ultimate nos dac
2
Please Describe one by one, what is your product available
- DI 8 ???
- DI 16 ???
- DI 4T ???
- DI 4MJ ???
- DI 1 ???
- DI 2 ???
3
Because it is a forum [some people didnt like to talk about trade]
so make a new thread in trading post [ market place ] or make a
good one in vendor bazaar, you can contact moderator & donate
4
Dont forget photo
im still waiting too
Hi Ecsdesign
Before all the discussion go to long
1
What is your DAC design available for us [ DIY Audio Member ]
you got a lot of fans and still waiting for the ultimate nos dac
2
Please Describe one by one, what is your product available
- DI 8 ???
- DI 16 ???
- DI 4T ???
- DI 4MJ ???
- DI 1 ???
- DI 2 ???
3
Because it is a forum [some people didnt like to talk about trade]
so make a new thread in trading post [ market place ] or make a
good one in vendor bazaar, you can contact moderator & donate
4
Dont forget photo
im still waiting too
i also have made this ultimate DEM reclocking published by EC in one of my TDA1541A cd-players
my first experiment was not very good because having a Kwak Clock 7, I took both the normal and inverted clock signal directly from the AD comparator telling to myself that the inverted signal might work very well without putting an extra load (the hex inverter chip) to the normal clock signal thus not introducing
extra jittery loads. But this turned not to be the best choice because i obtained sound distorsions on the low level signals (the quiet passages and the end of tracks)...after this i experimented by changing some values for the 2,2k resistors going to the -15V but distorsion still there!
After this i decided to add a hex inverter (74HCT04) to the normal clock signal and obtain the inverted signal from this.
I did this setup and distorsion still there (with initial 1k resistors to -15V) puting 1,2k in series with the 1k and obtaining the Magical 2,2k value - Distorsion gone !!
It seems that EC tuned very well this circuit obtaining the optimal values because every small deviation produces distorsions !
After that i decided not to extra load the BCK going to the NOS DAC (masterclock/4) and i sent to the hex inverter chip the masterclock/2 coresponding to a 5,8 mhz signal from the divider chip.
5,8mhz and inverted 5,8mhz going through 10k to the DEM pins seems to me the BEST Solution!
It gives more focus to the sound and crisper highs just like changing a plain TDA1451 TO a S2 double crown version of the same chip.
I'm very happy with this mod together with short I2S path and I2S attenuators so BIG thanks to EC for sharing with us his great findings
my first experiment was not very good because having a Kwak Clock 7, I took both the normal and inverted clock signal directly from the AD comparator telling to myself that the inverted signal might work very well without putting an extra load (the hex inverter chip) to the normal clock signal thus not introducing
extra jittery loads. But this turned not to be the best choice because i obtained sound distorsions on the low level signals (the quiet passages and the end of tracks)...after this i experimented by changing some values for the 2,2k resistors going to the -15V but distorsion still there!
After this i decided to add a hex inverter (74HCT04) to the normal clock signal and obtain the inverted signal from this.
I did this setup and distorsion still there (with initial 1k resistors to -15V) puting 1,2k in series with the 1k and obtaining the Magical 2,2k value - Distorsion gone !!
It seems that EC tuned very well this circuit obtaining the optimal values because every small deviation produces distorsions !
After that i decided not to extra load the BCK going to the NOS DAC (masterclock/4) and i sent to the hex inverter chip the masterclock/2 coresponding to a 5,8 mhz signal from the divider chip.
5,8mhz and inverted 5,8mhz going through 10k to the DEM pins seems to me the BEST Solution!
It gives more focus to the sound and crisper highs just like changing a plain TDA1451 TO a S2 double crown version of the same chip.
I'm very happy with this mod together with short I2S path and I2S attenuators so BIG thanks to EC for sharing with us his great findings