Hi,
I`m still in the procces of building my own cd player, with a cdpro2 transport from Philips.
The black gates I orderd are still not in, so I guess I have to wait a bit more
This day I had a discussion with a colleage at work, who is quite a clever person, and he is a skilled electrician. We were talking about my cd player, and he claimed that a power supply in a cd player is completly not important. That is; that it doesn`t have to be clean, and adding multiple ps is pure nonsense, cos digital= digital.
Now I have some understanding of analogue, and not much of digital, whats the deal with the power supplìes in cd players?
What are the arguments -which I can use
to convince him -
for having a clean ps/multiple ps`s.
tnx!
I`m still in the procces of building my own cd player, with a cdpro2 transport from Philips.
The black gates I orderd are still not in, so I guess I have to wait a bit more
This day I had a discussion with a colleage at work, who is quite a clever person, and he is a skilled electrician. We were talking about my cd player, and he claimed that a power supply in a cd player is completly not important. That is; that it doesn`t have to be clean, and adding multiple ps is pure nonsense, cos digital= digital.
Now I have some understanding of analogue, and not much of digital, whats the deal with the power supplìes in cd players?
What are the arguments -which I can use
to convince him -for having a clean ps/multiple ps`s.
tnx!
Member
Joined 2002
A unclean power-supply will induce noise and Problems. Most chips want ripple free and constant power flow not up and down Fluctuating.
Think of it like this. say your car is fuel injected. now say your injectors are dirty and your engine runs like crap now it is because they don't have a constant flow of fuel. Now you clean them and there flowing smoothly and cleanly you stomp on the gas and your roars with power with out sputtering to et to speed and power. : O ) seem simple.
Think of it like this. say your car is fuel injected. now say your injectors are dirty and your engine runs like crap now it is because they don't have a constant flow of fuel. Now you clean them and there flowing smoothly and cleanly you stomp on the gas and your roars with power with out sputtering to et to speed and power. : O ) seem simple.
thanks for the posts.
But how does it really influence the digital part? Is it worse because it will oscillate with a bad supply?
My collegue says this "do the parts work at 4.9 v?"
me: "yes"
He "does it affect the digital bitstream?"
me "Maybe, most likely not ideal"
he "If it works, it works"
me "ermmm"
I must kick his *** with some proven facts!
But how does it really influence the digital part? Is it worse because it will oscillate with a bad supply?
My collegue says this "do the parts work at 4.9 v?"
me: "yes"
He "does it affect the digital bitstream?"
me "Maybe, most likely not ideal"
he "If it works, it works"
me "ermmm"
I must kick his ***
with some proven facts!
why don´t you read the documents provides on Nicos homepage? It´s not so important If you have 4,8 or 5,2 V instead of 5V.
The 9 V supply needs to able to deliver short times enough current for moving the actuators, and should not be more than 10V or so, else the actuators may overshoot. (to are to fast)
The 5V is not only a digital supply. because there is an analog circuit (2 OpAmps and 2 capcitors) also on the 5V supply. If the 5V is noisy it may affect the OpAmps (I´m sure they will attenuate the noise 60 or 90 dB or so).
Nico told me there is no voodoo with the power supply and the sound, the supply should only be stabel and low impedance (every 7805 with enough small capacitors around should be)
http://www.daisy-laser.nl/homeoptics/page13.html
important is there should not be 9V without 5V else the actuator of the lens may move uncontrolled and cause damage.
The 9 V supply needs to able to deliver short times enough current for moving the actuators, and should not be more than 10V or so, else the actuators may overshoot. (to are to fast)
The 5V is not only a digital supply. because there is an analog circuit (2 OpAmps and 2 capcitors) also on the 5V supply. If the 5V is noisy it may affect the OpAmps (I´m sure they will attenuate the noise 60 or 90 dB or so).
Nico told me there is no voodoo with the power supply and the sound, the supply should only be stabel and low impedance (every 7805 with enough small capacitors around should be)
http://www.daisy-laser.nl/homeoptics/page13.html
important is there should not be 9V without 5V else the actuator of the lens may move uncontrolled and cause damage.
Ask your friend how much audio equipment he has built. If he can show you something that you think sounds good then look at it carefully. Anytime you are dealing with analog audio, and that's what the output of the cd player is, the power supply needs to be clean and low Z.
Yes the cd player might work with any junk power supply but it will sound like that power supply, or more correctly the noise that gets past the power supply and is coupled into the analog audio part from the other circuits in the unit.
Put an AM radio near a computer and listen to the hash. That signal is on the power supply lines of you cd player. Why make the opamps in the audio ourput of the cd player deal with that?
Later BZ
Yes the cd player might work with any junk power supply but it will sound like that power supply, or more correctly the noise that gets past the power supply and is coupled into the analog audio part from the other circuits in the unit.
Put an AM radio near a computer and listen to the hash. That signal is on the power supply lines of you cd player. Why make the opamps in the audio ourput of the cd player deal with that?
Later BZ
Member
Joined 2002
Little story my dad showed me how to use my o-scope. : O ) i was having problems with one of my amplifiers. i didn't know that the cap was bad in the psu. he showed me that my amplifier was running on. the positive side showed no ripple but the negative did. showing me the psu ripple showed why the amplifier sounded like crap compared to the other mono block. till this day i still use the scop to look at the ripple on my psu's if there is any ripple i rebuild the psu.
It is very important to have a clean ripple free psu.
It is very important to have a clean ripple free psu.
till said:
I think this is simplificating the matter a tad too much. Nico Thevissen is more a salesman than a technician as far as I know. It surprises me that they still don't offer a PCB for the supply
A friend of mine feeds his CDPro 2 with batteries and he can switch between batteries and transformer. It was a big difference.
The quality of the 5V supply should be high as you hear it immediately when it has its flaws. A 7805 is the worst solution for this device. It is sensitive on the 5V line as most cdplayers do. Changing caps in the 5V line can be heard immediately too ( as usual ).
wazzup said:thanks for the posts.
But how does it really influence the digital part? Is it worse because it will oscillate with a bad supply?
My collegue says this "do the parts work at 4.9 v?"
me: "yes"
He "does it affect the digital bitstream?"
me "Maybe, most likely not ideal"
he "If it works, it works"
me "ermmm"
I must kick his ***
Hi,
I have some questions (to start with) for your genius colleaugue, please pass them on, as well as his answers to us:
1 - How does a multibit (for example) converter work, and how does clock jitter affect the linearity hence affects the sound quality
2 - How does induced noise in the clock increase the clock jitter
If he answer these 2 questions correctly, he will understand why supply noise is important. At that stage tell him the power supply for my clock has noise below 3 nV/Sqrrt Hz.
all the best
Member
Joined 2002
Member
Joined 2002
nV / (Hz ^ .5)
It's a noise figure I've seen in a lot of op amp spec sheets and I think it is a way of relating the amount of absolute noise as a function of the frequency of the signal. For instance, the spec says that noise tends to go up as frequency goes down.
nanoVolt = .000000001 volt
Seems like a small amount except when roughly converted to dB. If we set 1 volt = 0 dB (not always true, but bear with me), then at -3 dB per half, 1 nV = -90 dB.
ensen.
It's a noise figure I've seen in a lot of op amp spec sheets and I think it is a way of relating the amount of absolute noise as a function of the frequency of the signal. For instance, the spec says that noise tends to go up as frequency goes down.
nanoVolt = .000000001 volt
Seems like a small amount except when roughly converted to dB. If we set 1 volt = 0 dB (not always true, but bear with me), then at -3 dB per half, 1 nV = -90 dB.
ensen.
You Can Tell Your Friend....
The data wavefom coming off the disc is an analog waveform.
This is converted to a squarewave (rectangular) data stream by a 'data slicing' circuit stage.
If the power supply to this stage is modulated (noisy), the threshold level varies and timing jitter is introduced into the data stream, and this jitter is very audible.
The analog audio stages do not exhibit perfect PSRR (power supply rejection ratio), and noise on the supplies will appear as valid audio.
Providing seperate transformers (and regulated supplies) for the servo, digital and audio output stages effects an immediate and very worthwhile improvement in cleanliness of the audio output.
Eric / - Never, Repeat Never Allow Electricians Near Electronic Equipment
The data wavefom coming off the disc is an analog waveform.
This is converted to a squarewave (rectangular) data stream by a 'data slicing' circuit stage.
If the power supply to this stage is modulated (noisy), the threshold level varies and timing jitter is introduced into the data stream, and this jitter is very audible.
The analog audio stages do not exhibit perfect PSRR (power supply rejection ratio), and noise on the supplies will appear as valid audio.
Providing seperate transformers (and regulated supplies) for the servo, digital and audio output stages effects an immediate and very worthwhile improvement in cleanliness of the audio output.
Eric / - Never, Repeat Never Allow Electricians Near Electronic Equipment
purplepeople said:nV / (Hz ^ .5)
It's a noise figure I've seen in a lot of op amp spec sheets and I think it is a way of relating the amount of absolute noise as a function of the frequency of the signal. For instance, the spec says that noise tends to go up as frequency goes down.
nanoVolt = .000000001 volt
Seems like a small amount except when roughly converted to dB. If we set 1 volt = 0 dB (not always true, but bear with me), then at -3 dB per half, 1 nV = -90 dB.
Hi
It is NOT noise as function of frequency (the unit of specification does not support that)
When measuring wider band signals, like noise, in the time domain, you either have to
- Express the bandwidth
- Measure it per Hz
If you take white noise, that has flat spectral distribution, with a level of 3 nV/Sqrrt Hz, you can, for example, express that in the time domain. According above, you have to state the bandwidth. Let us say we integrate that noise over 1 MHz, then there is a ratio of 1 milion between 1 MHz and the unit measurement per Hz.
60 dB then has to be added to come to 3 uV in the time domain, over 1 MHz.
so far
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.