One point that is missing is that this is not simply a mixdown or multiplication process.
For a 100 Hz output cycle for example there are 100 times more
clock edges involved in comparison to a 10 KHz cycle, and thusly much more averaging.
No, I have no interest to do the math.
Now back to my 4 & 8 GHz cavity resonators made from 3/4" copper pipe
end caps from the hardware store. After a string of frequency multipliers
that high, the phase noise starts to be annoying.
Gerhard
For a 100 Hz output cycle for example there are 100 times more
clock edges involved in comparison to a 10 KHz cycle, and thusly much more averaging.
No, I have no interest to do the math.
Now back to my 4 & 8 GHz cavity resonators made from 3/4" copper pipe
end caps from the hardware store. After a string of frequency multipliers
that high, the phase noise starts to be annoying.
Gerhard
This kind of crap is demeaning to the Deaf community, yet you’re the one posting from a high horse. Incredible.
True, but I think that's covered by: "As the shift in time is the same for all components, the shift expressed in radians is larger for higher frequencies." It's also why I calculate such low levels for the sidebands around audio frequencies, especially low audio frequencies - which doesn't match Joseph's measurements at all.
The Hammond box is anodized so the connectors are virtually isolated.
The ground of J1, the oscillator output, should be connected to the aluminum case so you should scratch the front panel around J1 to get electrical contact.
I just can't stop but ask..
What is the actual output frequency in a DS dac?
Audio signal is definitely not, it's just another modulation.
Last year, after I had repaired our E4407B, I had it home a while. Had done a series of screenshots of the 4499 output, just after I/V, so only a first order bland filter.
In this shot it's in PCM mode, probably (bad documentation) just 44,1k or 48k native.
The center frequency component is 6,15MHz, the marker is on the 12,30MHz harmonic
What is the actual output frequency in a DS dac?
Audio signal is definitely not, it's just another modulation.
Last year, after I had repaired our E4407B, I had it home a while. Had done a series of screenshots of the 4499 output, just after I/V, so only a first order bland filter.
In this shot it's in PCM mode, probably (bad documentation) just 44,1k or 48k native.
The center frequency component is 6,15MHz, the marker is on the 12,30MHz harmonic
Attachments
Last edited:
This is why I'm asking. in this DS dac this ratio is fixed for all audio signal frequencies..?
Also, it is much smaller than if it were the actual audio freq. So higher modulation factor from the clock phase noise? And constant..
This is why John talked about basically no attenuation for the DS clock phase noise?
Chris,
you can ask for it another 1000 times, sorry but we will not change our priorities, I was clear but let me repeat our priorities:
1) support to the members who got our designs (oscillators, frequency doublers, battery supply system, regulators, sine to square converters, clock switches, I2S over HDMI Tx/Rx, old oscillators design and so on)
2) complete current projects like the FIFO Lite and the DAC Lite
I would thank George and Doede for the tests they are doing, but the measurements you are asking are not our priority.
Moreover, as I have already said, as soon as the FIFO and the DAC Lite are ready we will make some measurements at DAC output, but following our approach (up conversion) and not yours.
BTW, if someone would provide a configurable image reject upconverter is welcome, he will save us a lot of work so the measurements will be anticipated.
Finally, having read the last 50 pages of this thread, what about this?
The Well synchronized asynchronous FIFO buffer - Slaved I2S reclocker
If a better clock is useless to improve the performance of a DAC even an asynchronous FIFO buffer is useless since the key of its performance is just a better clock than the one of the source.
So I expect another 50 pages of criticism since we are building another useless device.
Lastly, it's clear we don't understand the topic, and we sure do not understand how to run an experiment or design a project.
Sorry, we are hobbyists and not audio professionals so we can't do more.
I thought there was room for even beginners like us on a diy audio forum, but maybe I was wrong.
Andrea
Thank you Andrea for answer! So J1 (oscillator ) must be grounded to case and J2 (power) no.
I'd say you can still see the rather irregular waveform coming out of the sigma-delta modulator as a superposition of many frequency components that all get phase modulated, the higher frequency components more than the lower frequency ones as the shift in seconds is the same for all. With close-in noise, the desired audio frequency component is then the only part that gets sidebands in the audio frequency range. Maybe I'm wrong, but I don't see how.
I think there is no difference, in any case I don't think it would create a ground loop.
BTW I wouldn't scratch the panel around the power connector.
Hello,
I am happy i got a so called stomp box from Hammond or one of its almost one one copies on the internet.
There is no anodizing and no need to scrape of that surface which is rather '' difficult '' to do it right. I mean if you want the best contact possible.
As you can see there are two parts and the way they are connected will make it function much more like a '' perfect screen ''
The material is a big thicker so you can even add some threaded holes ( M3 or M4 is very well possible with this thickness)
You can go over the top with screening these circuit it seems as i read how critical everything is.
On top of you can create an almost airtight box ( if you add a rubber grommet around the power input connector)which is good for airborn vibrations.
Greetings, Eduard
I am happy i got a so called stomp box from Hammond or one of its almost one one copies on the internet.
There is no anodizing and no need to scrape of that surface which is rather '' difficult '' to do it right. I mean if you want the best contact possible.
As you can see there are two parts and the way they are connected will make it function much more like a '' perfect screen ''
The material is a big thicker so you can even add some threaded holes ( M3 or M4 is very well possible with this thickness)
You can go over the top with screening these circuit it seems as i read how critical everything is.
On top of you can create an almost airtight box ( if you add a rubber grommet around the power input connector)which is good for airborn vibrations.
Greetings, Eduard
Attachments
Hello,
By the way i think it is rather strange to tell people to go for a lifepo4/supercap configuration in order to exclude groundloops.
Most people will use ONE single supply where the output( in my case a supercap but connected to the mains to hold its charge) is connected by 6 individual dc output cables to each of the 6 boxes. The six boxes make no physical contact but of course they '' share '' the output terminal of the supercap.
This SHOULD work perfectly. If it creates a ground loop that can only be avoided by disconnecting the mains my idea then is there is something WRONG elsewhere.
Greetings, eduard
By the way i think it is rather strange to tell people to go for a lifepo4/supercap configuration in order to exclude groundloops.
Most people will use ONE single supply where the output( in my case a supercap but connected to the mains to hold its charge) is connected by 6 individual dc output cables to each of the 6 boxes. The six boxes make no physical contact but of course they '' share '' the output terminal of the supercap.
This SHOULD work perfectly. If it creates a ground loop that can only be avoided by disconnecting the mains my idea then is there is something WRONG elsewhere.
Greetings, eduard
The star washer digs into the metal and all is good no need to scrape I found.
Hello Simon,
Depends upon the quality of the anodization.
There are many '' types'' I have made objects for the military that would need more than the washer to be removed.
AND usually these small connectors cannot stand the torque necessary to grind to the surface. SO better use a tool to do it.
I have been working in metal for several decades!
greetings, eduard
Hello Simon,
Depends upon the quality of the anodization.
There are many '' types'' I have made objects for the military that would need more than the washer to be removed.
AND usually these small connectors cannot stand the torque necessary to grind to the surface. SO better use a tool to do it.
I have been working in metal for several decades!
greetings, eduard
Wow, what a disappointingly vindictive & unsavoury post!! Obviously your last post was disingenuous - "I have said everything and tried to help you" which you have contradicted here on both counts!
Maybe this time you will follow through with your announcement in this post - "I am outta here"?
Maybe this time you will follow through with your announcement in this post - "I am outta here"?
- Status
- Not open for further replies.