difference between RF and ground loop buzz? - diyAudio
Go Back   Home > Forums > Amplifiers > Chip Amps

Chip Amps Amplifiers based on integrated circuits

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 16th December 2012, 04:42 PM   #1
diyAudio Member
 
Join Date: Oct 2012
Default difference between RF and ground loop buzz?

Are there ways to identify if a buzz is RFi or a ground loop? For example, does DC offset mean ground loop? Is a ground loop independent of volume?

I have a basically symmetric stereo set up on a lm3875 but the buzzing channel input passes across the V+/- wires and the silent channel doesn't have this feature. I'd like to be sure I've diagnosed the problem before I start doing surgery.
  Reply With Quote
Old 17th December 2012, 12:43 PM   #2
diyAudio Member
 
Join Date: May 2008
RF is radio frequency. You cannot hear it directly, because it is above the human hearing range.

Buzz can come from a ground loop, from incorrect grounding or from inductive coupling.

If your diagnosis were correct, you would be talking about inductive coupling. In that case you would only have to remove the wires a bit or lift the PCB a bit. A few cm are enough to reduce that effect, because the magnetic force involved decreases by the square of the distance. It would however mean that the smoothing is not working well, because the V+/- wires should be passing DC, which is not audible. The AC ripple on top of that should normally be too small to induce audible hum. That means the cause for the buzz is probably something else. A grounding schematic and/or photos could help to solve the issue.
__________________
If you've always done it like that, then it's probably wrong. (Henry Ford)
  Reply With Quote
Old 17th December 2012, 10:59 PM   #3
diyAudio Member
 
Join Date: Oct 2012
Thanks PB. I'm just learning what these things are one step at a time. I fixed my buzz, and you are right that it was something else.

I used two cat6 strands for the signal and signal ground lines that go from the pot to the board. They aren't really that long---one channel not more than 3 inches, but the other is more like 4-5. The dc offsets on the short channel are quite nice at around 19mV, but the other is closer to 45. I guess this isn't totally unacceptable, but I wonder that it is that is making this difference.

Also, there seemed to be a little rise in a buzz about 3/4 through the reach of the pot. Maybe it was the test (junk) speakers I was using to check things out?
  Reply With Quote
Old 17th December 2012, 11:39 PM   #4
diyAudio Member
 
Join Date: Nov 2012
RF can cause offsets due to RF rectification and it can also cause audible interference. RF rectification can also create a peak detector in your circuit that can pick out an AM signal.

Unless your input in truly balanced and accepts a differential signal, CAT6 won't help you. It's twisted pair, which is designed to enhance noise rejection for differential signaling, while also reducing crosstalk. If you input is single ended, you need a single conductor wire with a good shield.
  Reply With Quote
Old 18th December 2012, 12:23 AM   #5
diyAudio Member
 
Join Date: Oct 2012
Oh man JJ. I wish I knew what you mean by "If your input is single ended." I am using a pair of cat6 wires of course, but I'm sure this isn't what you mean.

I'm guessing the suggestion would be something like coax. I think that involves binding the shield to the signal ground--- right? Or can it be sent to the chassis?

Anyway, I think I've taken care of the major issue which was almost certainly grounding, and I've routed the signal just a little differently, which I think helped too, but I'm not going to rip it apart just to see which was the true cause.

Except for the (relatively) high offset in one channel things seem to be OK. That extra 20mV of DC offset is still there I guess, but if less than 100mV is acceptable, I wonder if I'll be chasing a specter..

...Just checking things out a little closer now, I need to accept that this machine isn't going in to service in a great listening room. I cranked it up put my ear close to the speaker and I think I heard the train was shorted out and stuck (not circling the tree.)

My last GC had perfect offsets, and I tried a different internal setup and this 40mV must be the cost?
  Reply With Quote
Old 18th December 2012, 02:33 AM   #6
diyAudio Member
 
Join Date: Nov 2012
No worries.. A differential input has two inputs, sometimes called positive and negative, sometimes called hot and cold. The idea behind it is that it subtracts one from the other. Because of this, any noise that gets introduced on both lines will get subtracted and cancelled out. The impedance on both of these inputs needs to be matched, such that the external noise will get introduced on the lines the same so they subtract out in the end.

If you input is single ended, however, you have one high impedance input that carries the signal, and one reference, which is usually ground. The reference is a low impedance to the common of you circuitry, which, like it said is usually ground. The signal input is high impedance, though. So, the idea here with a shielded cable is that the shield "shunts out" any noise pickup directly into the ground, shielding the high impedance side from the noise. While it is true the signal input is reference to the common, the mismatch in impedances makes it unbalanced and susceptible to noise if unshielded.

It's hard to know where the offset is coming from without seeing the schematics. You can try shorting the input directly at the input and then measuring the output to eliminate any weird noise pickup.
  Reply With Quote
Old 18th December 2012, 03:13 AM   #7
DUG is offline DUG  Canada
diyAudio Member
 
Join Date: Jan 2003
Location: mississauga ontario canada
"Also, there seemed to be a little rise in a buzz about 3/4 through the reach of the pot."

Is the case of the pot connected to gnd?

If it is not then this might be your problem.
__________________
Doug We are all learning...we can all help
"You can't stop the signal, Mal. Everything goes somewhere..."
  Reply With Quote
Old 19th December 2012, 07:33 AM   #8
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
You need to post some detailed photos of the circuit.

If your input signal conductor and your input signal ground reference conductor are separated from each other, anywhere, then that creates "enclosed loop area", which makes an antenna for your AC mains (and everything else). [See "Faraday's Law", or Maxwell's Equations.]

The signal input and its ground should be tightly twisted together, ALL the way from the input jack to the input resistor at the actual amp chip input pin (or to where they go onto a PCB).

You could use coax but it would be best to use shielded twisted pair. If you are using shielded twisted pair, then the shield does NOT get connected to the input signal ground or the input jack.

The shield (of shielded twisted pair) should be connected to the chassis, only, at the input end only, and should be connected to nothing at the other end.

The input signal ground should NOT be connected to the chassis. i.e. Your input jack must be insulated from the chassis. The ground end of the input resistor should have its own dedicated connection back to the star ground point, probably somewhere after the ground of the last power supply reservoir capacitor.

Similarly, your AC mains and secondary and rectifier wire pairs must each be tightly twisted together, ALL the way to each end. Do not separate the incoming AC wires (for example, to run only one of them to a power switch). If any space is left between the two wires in any of those pairs, it will act as a transmitting antenna, which will send hum into all of the receiving antennas you might have inadvertently created, especially where input signal and input signal ground (or other) pairs have space between their conductors.

Your DC power and ground rails should similarly not have any enclosed loop area. Speaker output wire pairs should be the same way. You get the idea. Tightly twist together (at least three to four turns per inch) ALL wire pairs.

Keep all small-signal conductors as far away as possible from, and at a right angle to, any conductor with large or fast-changing currents.

On PCBS, be sure to use a ground plane under or over the entire small-signal input area, from input connection to chip input pin, so there is no (or minimal) space between the signal and its ground conductor. Otherwise (on single-sided PCBs), keep the signal and signal ground traces RIGHT next to each other, everywhere, with a minimal gap between them. And wherever a component is inserted into the signal trace, extend the ground trace under the component.

And keep the small-signal ground separate from all other grounds, all the way back to the star ground point.

All components that connect to chipamp pins should be connected as close to the pins as possible.

There must be a small-size, small-value bypass capacitor connected right at each chipamp power pin, and to power ground (0.1uF X7R ceramic type is good, there; film types and NPO or C0G ceramic are potentially risky, there, but can be better as long as they don't cause high-frequency instability). There should be larger electrolytic decoupling capacitors (largest value that can possibly fit; more than one of smaller values is even better) connected right at each chipamp power pin, and to power ground. Larger values of decoupling capacitance should be connected as close as possible to the chipamp's power pins. Multiple smaller electrolytics in parallel are much better than one large one. Power and ground planes on opposite sides of thin PCB (1 mm thick or less), with multiple parallel electrolytic decoupling caps, are MUCH better than one large cap per rail and single traces to power and ground pins.

How much reservoir capacitance is being used after the power supply rectifiers? What kind of rectifier bridge or rectifier diodes are being used?

Last edited by gootee; 19th December 2012 at 07:54 AM.
  Reply With Quote
Old 20th December 2012, 02:03 AM   #9
diyAudio Member
 
Join Date: Oct 2012
Hey thanks, all this is helpful. I'm just learning. I'm using the diyaudio pcbs. This was my second go, and the first one was layed out with exceptionally short connections and was dead silent.

I regret that I don't have a camera in order, but I'm also pretty happy with where things are.

For my 2nd go I tried something that solved the costs to the end-user I built in with my first very-short-connection design. (The volume was in an odd spot given the shape of the chasis.)

I think I've got the buzz sorted. The only thing that is less than optimal is that the offset is just a little higher than 14ish.

I'm really just learning one cm at a time, and what you've written gets me just a little closer to understanding shielded connectors. I think I'll throw a lifetime supply into my next mouser order. looks like a reasonable choice would be

8451 010100 Belden Wire & Cable | Mouser


So, there's one thing I still don't know. Other than audible buzzing, how do you test the quality of the system? My sense is that a ground loop would manifest as DC offset, but are there other tests I could do to evaluate the job I've done?
  Reply With Quote
Old 20th December 2012, 03:26 AM   #10
gootee is offline gootee  United States
diyAudio Member
 
Join Date: Nov 2006
Location: Indiana
Blog Entries: 1
That shielded twisted pair will be fine.

A ground loop would typically make itself known by causing audible hum, since, as a loop, it would be one of the antennas I mentioned, and probably a big one.

Any loop that is in the presence of a time-varying magnetic field will have a time-varying current induced in it. That's Faraday's Law. (The converse is also true such that a loop with a time-varying current will induce a time-varying magnetic field in the air.)

Conductors have resistance and self-inductance. So a (hopefully small) voltage will be induced, by that induced current, between any two points on such a loop of wire. In a "ground loop", you would probably see a lot of 60 Hz AC current (or 50 Hz, elsewhere), since there's a lot of that around in the air, from the AC power system.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
ground loop buzz problem SCOTT123 Solid State 3 24th March 2008 10:53 AM
ground loop Jamh Chip Amps 10 2nd November 2007 08:18 PM
Ground Buzz Ipanema Chip Amps 14 26th August 2006 07:55 PM
hum,ground loop,earth loop problem with your answers please frank2395 Pass Labs 2 17th May 2005 07:46 PM
does connecting mains ground with circuit ground create a ground loop? jarthel Everything Else 0 25th June 2003 01:55 PM


New To Site? Need Help?

All times are GMT. The time now is 08:36 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2