ripple & hum HPF?

[syn]

Found this ripple and hum reduction circuit, but i am not quiet sure where to insert it into (linear) PSU. Also would it be OK to use a single inductor=L1+L2 instead of two inductors in series? Tried to Google some more info on this circuit, no luck. Any info appreciated.

http://electronicdesign.com/Articles/ArticleID/14911/14911.html

 

[gyraf]

Dosn't go inside the psu, but removes dirt from signal line.

Very different beast than what you seem to expect.

Jakob E.

 

[syn]

gyraf

Thank you.
Are there any reasons not use this after the power transformer secondary?
I see that caps rating is 50V so it is not for the power line (220Vac)?

 

[gyraf]

Not for the ac portion of your psu - it'd simply remove your power. nor for the dc portion - it's ac-coupled...

this circuit is a notch filter for removing hum from a signal line, not at all for the power circuit, imo..

 

[syn]

 

[syn]

thanks again. any known ways how to reduce ripple and hum from the PSU beside caps and regulators? whatever i do can't get rid of this:

 

[beatpoet]

What have you tried? What device are you powering?

It looks like another pi filter wouldn't hurt.

 

[syn]

beatpoet
thank you. Various devices, no tubes, all transistors, ICs, (line stuff gssl, some racked ch strips). I don't use mic pre's much almost at all (to amp. mics).PSU is 317 data sheet/JLM powerstationt based, bipolar, same-similar in all units. Toroids dual secondary.The shown diagram is for 1.227Vac
signal, unbalanced.All that i made is having this problem...Cap filter brand/size seems to have little influence on to this (~none).Tried various
grounding schemos as well, no progress...

 

[thermionic]

Syn,

I built a 2-channel line preamp this week, powered from an LM317/337 supply. The highest peak of mains harmonics is @ -128dB

The reason I mention the above is to say that there is nothing wrong with the 317/337 combination, and I employed a basic circuit that's very similar to the one you see in Mackie 1604 desks (I don't like their sound, but the PSU is cheap and nice!)

I really think something is very wrong with your grounding / general construction / positioning / case shielding - I wonder if you are 'barking up the wrong tree' with the topology of the PSU.

Can you post a screenshot of your power rails under load, i.e. while they're powering the G-SSL etc?

Can you compare the PSU’s DC line scope waveform unloaded (disconnected from main PCB) and loaded?

Can you post a schematic of your grounding layout? Some internal photos?

I am really bad with PCs, so I cannot post a JPG here, but I can send anyone a screenshot of an FFT from my preamp to give you an idea of the difference. My PSU is outboard though, so it should be several dB quieter with regards to mains harmonics.

I suspect you could have a simple error in methodology here - don't despair!

Justin

 

[syn]

Thermionic
-128dB!!!Great!
Yeah i suspect me as well, not saying 317/337 are bad
or topology...
Some more info:
Lifting gnd on the output makes things worse.
I stick to RANE “sound system interconnection” when it comes to gnd.ing,
connectors and cables.

[quote author="thermionic"]
Can you post a screenshot of your power rails under load, i.e. while they're powering the G-SSL etc?
[/quote]

The above screen shot is taken while powering the unit (two fet bloaks line amp).Regardless of the equipment to power, the performance is
similar to the above...

[quote author="thermionic"]
Can you compare the PSU’s DC line scope waveform unloaded (disconnected from main PCB) and loaded?
[/quote]

No unfortunately no scope here.

[quote author="thermionic"]
Can you post a schematic of your grounding layout? Some internal photos?
[/quote]

I use one or another grounding layout:

thank you

 

[Andy Peters]

[quote author="syn"]thanks again. any known ways how to reduce ripple and hum from the PSU beside caps and regulators? whatever i do can't get rid of this:[/quote]

Can you post a picture of your power supply's layout?

-a

 

[syn]

sure, this is on the bench now and the above RMAA is for it powering two DOAs .

schematic.1646 part removed.

 

[thermionic]

Syn - I got an excuse to post some FFTs from my preamp here: http://recforums.prosoundweb.com/index.php/t/21433/998/

If I can be really honest, considering what you can find an old 'scope on Ebay for these days, you really are making your life a lot more difficult by not using a 'scope. The 'scope will enable you to trace through all stages of your power and signal wiring so you that can narrow down the vacinity of where the problem is coming in. It may not be a PSU fault, and could be inductive pick-up somewhere in your audio wiring... It could be a lot of things, and a 'scope will give you a 'nose' to sniff with!

BTW - does it make a difference if you bypass the 10R resistor going to Case Earth?

Justin

 

[syn]

thermionic
thnx for pointing me in to the right direction.It is good to know that those 317/337 can perform so well...

 

[mikep]

[quote author="thermionic"]you really are making your life a lot more difficult by not using a 'scope.
[/quote]

he's right. and Ill add that the circuit might be oscillating at high frequency. that can cause "hum". you'll never see it on your PC because of that annoying nyquist guy. what are you powering with it? that could be the culprit. did you try loading the PSU with a resistor? cap couple into an amplifier and listen to the noise.

mike

 

[pstamler]

Right now you have the adjust terminal of the regulator grounded. Try this: disconnect the adjust terminal from main circuit ground, and instead run a wire from the adjust terminal to the grounded terminal of the load. Then of course the load's ground goes back to the main ground point. Sometimes this gives a good deal less noise.

Peace,
Paul

 

[thermionic]

I suspect that Paul may have pointed out the elephant in the room!

BTW - that's a great suggestion from Mike to 'listen' to the supply via a coupling cap, but it's probably safest to do it via a pot wired as a divider, starting with maximum resistance. Mike could also be onto something with the oscillation angle - that would certainly explain an FFT such as the one you've presented.

Justin

 

[syn]

mikep
great idea, tomorrow first thing to do. this kind of PSU noise is present on all my builds more or less. this time i'm playing with tamas- bigbloke. Single DOA unbalanced in, unbalanced out (2200uF@the output) API like circuit, line level.
thank you

pstamler
well another interesting thing to try. tomorrow second thing to do.
thanks a lot

thermionic
what is the meaning of that (r7/r8 (2k7)) resistor in mackie psu?
thank you

I am here at the moment :

and audio part @ +4dBu

thank you

 

[Guest]

Syn;
pay attention to wires.
You may calculate current peaks on wires that go from rectifier to filtering caps. They are much higher than a steady current consumed by your gear!
Now, if you solder a wire that goes to your regulator to a rectifier instead of to a filtering cap you may calculate a voltage drop on that wire that has very small resistance, but it has some!

 

[thermionic]

Syn - R7 / R11 as well as R8 / R12 are used in parallel to achieve +/-16v. You can replace them with a pair of 5K trimmers to adjust voltage exactly.

Justin