Hi !
sorry to jump in with a weird question.
More a curiosity.
Where can i find information about the noise of a SLA battery in MHz band ?
I have this feeling that should be quite low and difficult to match with a traditional power supply. Am i wrong ?
It is to power a clock generator circuit.
Thank you very much indeed.
gino
sorry to jump in with a weird question.
More a curiosity.
Where can i find information about the noise of a SLA battery in MHz band ?
I have this feeling that should be quite low and difficult to match with a traditional power supply. Am i wrong ?
It is to power a clock generator circuit.
Thank you very much indeed.
gino
In almost all circuits most MHz noise arrives from outside, so little point in measuring what a battery does. You could try reading the literature on low noise oscillator design, but I doubt if people use batteries for this purpose. Note that for low oscillator noise the main requirement is low PSU noise at baseband (i.e. LF) frequencies, not at the oscillator frequency. Osc frequency noise from the PSU can be easily filtered away. And, of course, you need a little osc frequency noise to get the thing started!
Hi and thanks a lot for the very helpful advice.
Sorry for the question but all these discussions about noise and how detrimental it is for performance have obsessed me.
This i did not know and it is indeed very interesting. I will search info on filters then.
Just another question ... in general it is preferable to put the filter on the AC or in the DC after the regulation ?
i mean, better to use a AC mains filter or a DC filter after the power supply to deal with high Hz noise ?
Thanks a lot again for the very useful explanation.
Regards, gino
You will always need filters in the DC supply. You may need filters on the AC side too, but that is much less important. Some audiophiles obsess about AC mains noise, but I guess that is because they don't know enough about circuits to think about the real issues.
Be aware that low noise oscillator design is a very specialist area of electronics and a DIYer would need to do lots of reading and thinking in order to even approach what an expert can do. For most audio purposes there is no need to worry, as simply using a better quality crystal will do.
Be aware that low noise oscillator design is a very specialist area of electronics and a DIYer would need to do lots of reading and thinking in order to even approach what an expert can do. For most audio purposes there is no need to worry, as simply using a better quality crystal will do.
Thanks a lot indeed again.
I have decided to study only the power supply side. I like power supplies a lot.
Unfortunately when the noise is low measurements are not trivial at all.
But i think i got the message. DC filters are key.
I will start reading about them.
thanks a lot again, gino
Gino, you still seem to be looking for simple recipe-like answers to complex questions. That is not the way to learn electronics.
Don't take away "DC filters are key". Instead, think about
- what are we trying to achieve? (low noise in an oscillator)
- where does oscillator noise come from? (resonator, circuit, PSU)
- how did that noise get there, and how can we reduce it? (thermal, shot, radiated interference, conducted interference)
- the key issue is: how does an oscillator work? (hint: an oscillator does not really oscillate, but is a very narrow band amplifier which amplifies its own noise via positive feedback - so you can't have a noiseless oscillator)
Don't take away "DC filters are key". Instead, think about
- what are we trying to achieve? (low noise in an oscillator)
- where does oscillator noise come from? (resonator, circuit, PSU)
- how did that noise get there, and how can we reduce it? (thermal, shot, radiated interference, conducted interference)
- the key issue is: how does an oscillator work? (hint: an oscillator does not really oscillate, but is a very narrow band amplifier which amplifies its own noise via positive feedback - so you can't have a noiseless oscillator)
Hi i know that is a very complex issue.
To be more specific i have to power this converter here
New Version CM6631A USB to SPDIF Coaxial Optical 24bit 192kHz Sound Card DAC | eBay
i need 5 clean VDC (i have a special USB cable that disconnect the board from the dirty pc usb power lines and has a power socket to use an external PS).
The real question is ... a very low noise 5VDC power supply can offer better performance that a cellphone charger (to extreme the concept) ?
I am using a SMPS now of low quality.
Can a better PS give me some benefits ?
Thanks again, gino
Last edited:
I new I saw it here sometime ago. SLA noise.
http://www.diyaudio.com/forums/chip...red-supplies-21.html?postid=187556#post187556
Although his graph only goes out to 40 KHz.
http://www.diyaudio.com/forums/chip...red-supplies-21.html?postid=187556#post187556
Although his graph only goes out to 40 KHz.
Last edited:
Hi ! thank you very much indeed ! very useful link.
And quite conclusive against the batteries option
considering also that batteries are not convenient.
Better search some good regulated power supply.
The one of the graph looks quite impressive in terms of noise. Quite spectacular. I will look at the thread in this section.
No batteries.
Thanks a lot again, gino
Last edited:
Hi and thanks a lot for the very helpful advice.
When i started to be interested in noise i found a video on Youtube showing a testing rig and the procedure of measurements.
Its is a very challenging test from a technical point of view. So i gave up with the idea of trying to measure noise.
As i said above i have become obsessed with noise because when i listen to my system late in the night the sound is remarkably better.
I guess this is due to lower noise in the electrical grid and not soo good noise filtering in my system.
I will study it. I have indeed seen batteries used in some application.
A friend of mine a radioamateur told me once about some important advantages with batteries that i did not understand, as usual.
Thanks a lot again, gino
If you suspect that interference is a problem then don't guess and run off into wild goose chases. Instead, measure and think. If you are careful then you can look at mains noise. You can buy a cheap SDR receiver to look at radiated noise. Does the noise actually reduce at night? Are you in a different frame of mind at night? You need data.
Once you have data, you can then consider solutions. Solve the problem, not something else. Be an engineer, not a politician!
Once you have data, you can then consider solutions. Solve the problem, not something else. Be an engineer, not a politician!
Hi well it is very expensive and difficult to measure very low level wideband noise.
I have seen on Youtube ... the price of the rig is 3 times the value of my car.
I agree with you that measurements are the basis. They are just not easy to carry-out. I love lab tests.
I do not think that interference is actually the problem ... more noise from mains.
Usually during the night there should be less noise on the electrical grid ?
even without measuring i think mains noise is the problem and therefore the power supply. The battery idea was intriguing also for this.
Mains are dirty. In some case very.
However i understand that a good power supply is fine enough.
Issue close.
I will look at the threads on +5VDC power supply projects.
Thanks a lot again to All.
Regards, gino
OK. You must be one of these 'intuitive' people who can detect electrical phenomena without the inconvenience of using measuring instruments such as 'scopes and meters.
If it is very low level then it can't be a problem. If it is a problem then it is not low level.
If the problem is conducted noise/interference then use a mains filter. If it is radiated then use better screening/shielding and add RF filters. If the equipment is poorly designed then use better equipment.
The classic audiophile response is to obsess over some relatively unimportant detail while completely ignoring major technical problems (or, in some cases, deliberately introducing such problems e.g. by using silly DIY or 'hi-end' cables).
I was just looking for +5VDC low noise power suply options.
I have no chance to set up a decent measurement rig. Too expensive and i do not even have the space for it. I have to start at least from a kit (not smd).
I am not so naive to think that design, build and optimize a high quality power supply is a trivial task. I know that it is a very challenging task.
The batteries can be easier for some reason. I have just to deal with the regulation stage.
So in the end i would just need a good kit of a regulator (not smd).
Actually i do not have a clue which is the level of noise that can be considered acceptable for the device i have. Really do not have a clue.
The problem with kits is that very seldom have a set of measurements.
They say very low noise but how low is not clear.
I guess that in order to be able to perform these tests i need instruments of thousands of dollars. But i agree about the not scientific approach of many DIYers.
For instance i was following a thread about the suppression of noise on a commercial power supply used in a usb to spdif converter, similar to the one i have in mind.
They were carrying out mods and praising the improvements obtained.
I asked how they were doing the tests.
The answer was by listening on the stereo the resulting sound.
This is a very audiophile approach i agree completely.
Because as you say an audiophile is more willing to spend on fancy parts (also quite expensive) instead of investing in some nice measurement device.
About the actual device i can only read some reviews. I do not know if it is basically good or bad. Everyone say that his projects are the best.
Not easy to spot the very good ones.
I read a comment of a guy saying that it is of decent quality and a good power supply can be beneficial. From this my question on power supplies.
As it gets the power from the usb port i have already a +5VDC SMPS (i have no idea about its noise but i guess it is quite noisy. Like a usb charger)
A battery followed by a very good regulator should be a better option.
Maybe it will not be beneficial. Surely not detrimental.
If the sound will not be satisfactory i think i will have to look for another converter.
Again ... there are many of them on the market.
To build one is out of discussion because they are all smd ... no smd tools here.
I can only buy one and try and hope that is good.
As i said it is like ask the landlord if his wine is good ... of course ... the best in town.
However i will also look in the digital section of this forum before run out and buying something.
Thanks again, gino
Last edited:
Hi all,
I've just briefly read into this thread so please excuse me if there's something I missed ...
Some time ago I did some measurements on SLA, LiPO4, and Li-ion batteries into my AD7760 based evaluation board. The software here provides for FFT analyses, THD etc. and in general the noise floor is ~ -160 dBs referenced to about 3.6 VAC (with 2.5 Msps and 256 times decimation).
I tried to load the different batteries with various currents - to my memory up to 0.8A/1A - and all battery types had no discernible noise when directly connected to the AD7760 EVB via a capacitor.
In terms of measurements made with a precision oscillator my experience is that powering the oscillator (actually the entire chain of electronics - in some cases including the computer) with a/more batteries (so that all supplies are floating relative to eachother) makes for a (much) lower noise floor especially in the lower frequencies.
Best regards,
Jesper
I've just briefly read into this thread so please excuse me if there's something I missed ...
Some time ago I did some measurements on SLA, LiPO4, and Li-ion batteries into my AD7760 based evaluation board. The software here provides for FFT analyses, THD etc. and in general the noise floor is ~ -160 dBs referenced to about 3.6 VAC (with 2.5 Msps and 256 times decimation).
I tried to load the different batteries with various currents - to my memory up to 0.8A/1A - and all battery types had no discernible noise when directly connected to the AD7760 EVB via a capacitor.
In terms of measurements made with a precision oscillator my experience is that powering the oscillator (actually the entire chain of electronics - in some cases including the computer) with a/more batteries (so that all supplies are floating relative to eachother) makes for a (much) lower noise floor especially in the lower frequencies.
Best regards,
Jesper
Is that because they were batteries, or because they were floating and so avoided ground issues?gentlevoice said:
Low noise oscillators are important in some test equipment and communications; audio is less demanding. Audiophiles fiddle with whatever their knowledge and skills allow them to fiddle with: for some this is just cables, others swap PSUs or capacitors, some bypass filters (which were present for good reasons). This is all fairly harmless provided they don't imagine they are really changing anything for the better.
I think both ... The floating approach made for less ground issues per se and the batteries additionally were low noise.
There are more FFTs of my battery measurements here:
http://www.diyaudio.com/forums/powe...listened-li-ion-batteries-12.html#post4117260
Best regards,
Jesper
Hi ! thanks a lot for the very interesting and very nice graphs. May i ask you which instrument have you used to get them ?
I have also another question ...
i do not understand what means top-off. Which low noise PSU are you using and how it is connected to the batteries ?
Your words like
excite me.
I drool for soundstage. It is for me pretty much everything.
Also because i never heard a system with a great soundstage sounding bad overall.
Thanks a lot indeed.
Regards, gino
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.