Hi Gino,
In reply to your questions:
The AD7760 Evaluation board. A 2.5 Msps ADC with quite low noise level.
No, my apology - it should have been "top-up" (I think ;-)). It's basically a very low noise power supply which only just keeps the batteries at their float voltage. For SLAs it would be 13.3 to 13.8 VDC (12VDC SLA type), for LiFePO4s it would be about 3.425 VDC, for Li-ion it would be about 4.050 VDC.
For PSU I may suggest something like the TWTMC D&D PSU ... See page 3 in the attached document. If you want the lower cut off frequency to go down (it's about 0.3 Hz with 470 uF) you can increase the 470 uF capacitor to e.g. 2200 uF. That would give 0.07 Hz cut-off frequency. I would parallel the 2200 uF/470 uF with an 100 nF C0G capacitor.
You may also replace the output transistor pair Q3 & Q4 with a double emitter follower e.g. with a 2SC2240 followed by a 2SC4793 where the quiescent current for the 2SC2240 is about 10 mA and 15 mA for the 2SC4793.
Remember, though, that there's no current limiting in the TWTMC PSU (nor in the 2SC2240 & 2SC4793 variant) so if the batteries draw a lot of current the transistors may be destroyed. Current limiting can be made e.g. by replacing the LD1085 with a current limited LM317T - see p. 20 in the attached LM317T datasheet.
Hope this helps.
Regards,
Jesper
BTW.: If you're into a well-defined soundstage may I then suggest that you consider a very good oscillator? The TWTMC may be such an oscillator but the GB is just closed - or a simpler solution could be the NZ2520SDs available e.g. from diyinhk:
NDK NZ2520SD 3.3V 80Mhz 49.152Mhz 45.1584Mhz 24.576Mhz 22.5792Mhz 12Mhz Ultra low phase noise oscillator - DIYINHK
In reply to your questions:
The AD7760 Evaluation board. A 2.5 Msps ADC with quite low noise level.
No, my apology - it should have been "top-up" (I think ;-)). It's basically a very low noise power supply which only just keeps the batteries at their float voltage. For SLAs it would be 13.3 to 13.8 VDC (12VDC SLA type), for LiFePO4s it would be about 3.425 VDC, for Li-ion it would be about 4.050 VDC.
For PSU I may suggest something like the TWTMC D&D PSU ... See page 3 in the attached document. If you want the lower cut off frequency to go down (it's about 0.3 Hz with 470 uF) you can increase the 470 uF capacitor to e.g. 2200 uF. That would give 0.07 Hz cut-off frequency. I would parallel the 2200 uF/470 uF with an 100 nF C0G capacitor.
You may also replace the output transistor pair Q3 & Q4 with a double emitter follower e.g. with a 2SC2240 followed by a 2SC4793 where the quiescent current for the 2SC2240 is about 10 mA and 15 mA for the 2SC4793.
Remember, though, that there's no current limiting in the TWTMC PSU (nor in the 2SC2240 & 2SC4793 variant) so if the batteries draw a lot of current the transistors may be destroyed. Current limiting can be made e.g. by replacing the LD1085 with a current limited LM317T - see p. 20 in the attached LM317T datasheet.
Hope this helps.
Regards,
Jesper
BTW.: If you're into a well-defined soundstage may I then suggest that you consider a very good oscillator? The TWTMC may be such an oscillator but the GB is just closed - or a simpler solution could be the NZ2520SDs available e.g. from diyinhk:
NDK NZ2520SD 3.3V 80Mhz 49.152Mhz 45.1584Mhz 24.576Mhz 22.5792Mhz 12Mhz Ultra low phase noise oscillator - DIYINHK
Hi Jesper, thanks a lot for the very kind reply.
Actually i meant the instrument you used to get the measurement.
Is it a scope ? which model ?
wow ... i do not know if i will be able to understand something.
I will try in the next weekend for sure. But this is difficult.
I have another question ...
if i start from a battery and assuming that it has a low noise
do you think that a normal regulator (even a fixed one or a LM317) can introduce much additional noise ?
because this was the idea ... SLAB followed by a simple regulator. I have to reduce the voltage to 5V in one case and 8-9V for another device.
Both are usb to spdif converters but require different voltage.
now i am completely lost ... what is an oscillator ? you mean a quartz ?
as i said my almost only interest now it is the usb to spdif conversion.
Sometimes ago i have decided to use the pc as only source but i would like to realized a good connection with the dac.
I have this silly idea that if the usb to spdif converter is very good i could connect any pc around, cheap or less cheap, and always be able to get a great input signal for my dac.
Maybe i am wrong. I will try to understand a little of the material you kindly provided.
May i ask you if you use the pc for music ?
Thanks a lot again.
Good evening, gino
Last edited:
Hi gino,
Yes, in my experience it can be quite audible depending on the rest of the PSU design. If you want to use an LM317T may I again suggest the figure named "Current Limited 6V Charger" on p. 20 in the previously attached LM317 datasheet ... A suggestion could be to leave the 1 ohm resistor as it is (willl current-limit at appr. 0.6A) and adjust the voltages to your needs (there's a formula for calculating this in the datasheet). If you also break the connection between the output and the node where the 240 ohm resistor connects - and place e.g. a 3.3 ohm/5 watt resistor here instead - you will add some output impedance to the LM317T. This - combined with some capacitors in parallel with the battery in the figure on p. 20 will create a low cut-off frequency of e.g. 5 Hz if you use a 10000 uF capacitor. The cut-off frequency will scale linearly relative to the 10000 uF so you can decide where you want to place the cut-off frequency by changing the capacitance. Remember adequate heatsinking.
Good luck with your endeavours,
Jesper
Yes, in my experience it can be quite audible depending on the rest of the PSU design. If you want to use an LM317T may I again suggest the figure named "Current Limited 6V Charger" on p. 20 in the previously attached LM317 datasheet ... A suggestion could be to leave the 1 ohm resistor as it is (willl current-limit at appr. 0.6A) and adjust the voltages to your needs (there's a formula for calculating this in the datasheet). If you also break the connection between the output and the node where the 240 ohm resistor connects - and place e.g. a 3.3 ohm/5 watt resistor here instead - you will add some output impedance to the LM317T. This - combined with some capacitors in parallel with the battery in the figure on p. 20 will create a low cut-off frequency of e.g. 5 Hz if you use a 10000 uF capacitor. The cut-off frequency will scale linearly relative to the 10000 uF so you can decide where you want to place the cut-off frequency by changing the capacitance. Remember adequate heatsinking.
I do but with re-clocking etc. ... can be superb sound but also technically challenging ...
Good luck with your endeavours,
Jesper
Last edited:
Good morning Jesper,
this is very bad. I thought that most of noise leaking through a regulation stage was not filtered noise coming from upstream the regulator.
Now i understand that the self-generated noise of a regulator stage can be substantial.
I have to study much more. I am an uneducated beginner.
well this is not mandatory. I would much prefer a single chip regulator design to keep circuit as simple as possible.
I do not like too complex projects at all.
I will study the pdf your provide me with for sure.
I just need some evenings free. Thanks again for the documentation.
i will leace the reclocking for the next phase.
Sorry to ask again but i have been impressed by the noise graphs.
Which scope are you using ?
i am trying to understand how much could cost a rig to measure noise in power supply. This is by far the most interesting topic for me.
And fundamental for any audio device.
Thanks a lot again.
Regards, gino
Hi again ...
It's not a scope ... It's an analog-to-digital converter evaluation board. This one actually:
http://www.analog.com/en/design-cen...tion-boards-kits/eval-ad7760.html#eb-overview
The advantage is that it'll allow me to look at noise into the low MHz region (i.e. appr. 1.25 MHz) with a sort of 24 bit resolution. However, it requires a bit of care to make sure that the input isn't damaged by accidental over-voltage.
If you are looking for a scope which has an FFT function I personally have good experience with the Rigol scopes. IMHO reasonably priced and can be bought on ebay. Please note though that such an FFT most often is only 8 or 10 bit resolution (256 or 1024 "steps") - thus with a limited level of detail. For FFTs many people here on diyaudio use e.g. an AK5394 based sound card and combine it with e.g. the software called ARTA (free I think).
If you read into this thread (search for the comments by 1audio) there are some advice on which sound card to buy. Some of them can be very low noise and with the AK5394.
http://www.diyaudio.com/forums/equipment-tools/205304-low-distortion-audio-range-oscillator.html
Also if you have some patience this thread may be interesting:
http://www.diyaudio.com/forums/equipment-tools/277808-diy-audio-analyzer-ak5397-ak5394a-ak4490.html
There's also the QA401 which appears to have quite a fine reputation.
https://www.quantasylum.com/content/Products/QA401.aspx
So, in terms of a "scope" capable of FFT in general the oscilloscopes as such have low resolution FFTs whereas the soundcards may have high resolution FFTs, however, their bandwidth usually is limited to a 100 kHz or 200 kHz. As I also make digital equipment operating at e.g. 50 MHz I have a scope (Rigol 10**) which I combine with the AD7760 evaluation board which is higher resolution (and comes with its own software).
So ... Cheers,
Jesper
It's not a scope ... It's an analog-to-digital converter evaluation board. This one actually:
http://www.analog.com/en/design-cen...tion-boards-kits/eval-ad7760.html#eb-overview
The advantage is that it'll allow me to look at noise into the low MHz region (i.e. appr. 1.25 MHz) with a sort of 24 bit resolution. However, it requires a bit of care to make sure that the input isn't damaged by accidental over-voltage.
If you are looking for a scope which has an FFT function I personally have good experience with the Rigol scopes. IMHO reasonably priced and can be bought on ebay. Please note though that such an FFT most often is only 8 or 10 bit resolution (256 or 1024 "steps") - thus with a limited level of detail. For FFTs many people here on diyaudio use e.g. an AK5394 based sound card and combine it with e.g. the software called ARTA (free I think).
If you read into this thread (search for the comments by 1audio) there are some advice on which sound card to buy. Some of them can be very low noise and with the AK5394.
http://www.diyaudio.com/forums/equipment-tools/205304-low-distortion-audio-range-oscillator.html
Also if you have some patience this thread may be interesting:
http://www.diyaudio.com/forums/equipment-tools/277808-diy-audio-analyzer-ak5397-ak5394a-ak4490.html
There's also the QA401 which appears to have quite a fine reputation.
https://www.quantasylum.com/content/Products/QA401.aspx
So, in terms of a "scope" capable of FFT in general the oscilloscopes as such have low resolution FFTs whereas the soundcards may have high resolution FFTs, however, their bandwidth usually is limited to a 100 kHz or 200 kHz. As I also make digital equipment operating at e.g. 50 MHz I have a scope (Rigol 10**) which I combine with the AD7760 evaluation board which is higher resolution (and comes with its own software).
So ... Cheers,
Jesper
Hi ! thanks a lot for the kind reply.
It sounds like a solution for very expert people. The look of the graphs is magnificient. Very nice indeed.
Thanks a lot again Jesper ! now i have stuff to read and try to understand for the next 6 months ! I love these noise measurements.
I have seen a very simple but telling lesson on Youtube about noise and its impact on the playback. When the noise is high the music details are covered by it and cannot be heard. All the microinformation are lost.
I liked the efficacy of the explanation.
I will try to understand something of the post you have mentioned.
Thanks a lot again.
Have a nice weekend, gino
Hi Jesper !
sorry to bother you again.
But are you saying that if i take a usb/PCI audio interface with the right SW i can use the pc as a scope for noise measurements ?
this is very interesting.
I have already one cheap usb AD/DA interface to try out. Just to practise.
In case which would be a decent SW scope for win 7 just to start ?Thanks a lot again
gino
sorry to bother you again.
But are you saying that if i take a usb/PCI audio interface with the right SW i can use the pc as a scope for noise measurements ?
this is very interesting.
I have already one cheap usb AD/DA interface to try out. Just to practise.
In case which would be a decent SW scope for win 7 just to start ?Thanks a lot again
gino
Last edited:
Hi ! thanks a lot for the very interesting advice.
Sorry for the questions that can sound trivial but i am ignorant as much as curious about the topic.
Could you give me please some more info about you set-up ? like which specific interface are you using and how it is connected to the batteries (and which batteries) ?
That would help me a lot.
Usb audio interfaces are my main concern these days in general.
Thanks a lot again.
Kind regards, gino
One test rig uses a $1 USB soundcard modified for single channel use. It plugs in to a ADuM4160 based isolator board, which is powered from a 12V VRLA 7Ah. That generally keeps mains earth loop noise pretty low for a normal mains powered PC. Of course a laptop is even better, as it is battery powered too!
http://dalmura.com.au/projects/Soundcard%20mods.pdf
I use a EMU 0404 for reference measurements - higher bandwidth to about 50kHz, and even lower noise floor - it uses ASIO driver rather than standard USB audio driver. I haven't yet made up a 12V battery to 5V regulated supply to that device, but that would also remove one mains AC connection.
The battery supply is being used to provide a voltage source to run equipment that has no AC mains connection - that is its merit.
http://dalmura.com.au/projects/Soundcard%20mods.pdf
I use a EMU 0404 for reference measurements - higher bandwidth to about 50kHz, and even lower noise floor - it uses ASIO driver rather than standard USB audio driver. I haven't yet made up a 12V battery to 5V regulated supply to that device, but that would also remove one mains AC connection.
The battery supply is being used to provide a voltage source to run equipment that has no AC mains connection - that is its merit.
Last edited:
Hi thanks a lot again for the very valuable reply.
i have also a a ADuM4160 based isolator myself.
I was worrying a little that it could introduce some sort of jitter but now i think i can leave it in the middle of the link with the usb interface.
Nice to know that the EMU 0404 is very low noise because sometimes it pops up in the 2nd hand market here.
I feel like i am close to define something. Usb interfaces are my almost only interest.
Especially after reading on another thread that with the right one i could control nicely my amplified speakers that i like.
I understand you are a batteries fan ... me too i am intrigued by batteries.
For the same reason you stress .... no connection with the bloody AC mains, the origin of many bad things in audio.
I will look for the EMU card you mention then.
Thanks a lot indeed.
gino
Last edited:
Hi
Thanks for the helpful advice. I am collecting information on this difficult topic.
from what i understand SLABs have indeed problems to provide istantaneous peaks of currents. i do not know if this is related to their impedance but it could be. But maybe this could be cured with some fast uF in parallel ?
Something like a 10mF low ESR cap ?
I am thinking to the car audio systems that use big capacities in parallel with the battery.
I like SLAB because very intelligent chargers are readily available
I could leave them on charge when not used and switch off the charger during the listening sessions.
The complete isolation from the mains is a big plus imhe.
Thanks again, gino.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.