Hunting down buzzes in an audio system
Since I wrote the Foreplay page (which, Holy Cow! is now linked from several other sites) I occaisionally get an eMail requesting advice on reducing hum in their preamp. Man, build a page, become an expert...
Anyway, I've noticed that while there are sveral good ages on grounding schemes for your DIY component designs, most of the good advice for solving real-world system ground loop problems is buried in maillists and discussion boards. So I've hunted down some of that advice and built this page. I hope the information helps you as it has me (several times).
Good Listening, pRC
(Yup, Again!) Tubes and AC power mean big voltages. Big voltages can stop your heart and kill you. Be careful! Follow all the suggestions of folks who are wiser than me and don't even try this if you don't know what you are attempting. Take responsibility for your own actions and don't sue me, OK?
In this page:
- What is a Ground Loop, anyway?
- a Grounding schematic with a star ground
- Basic Debugging
- Useful posts from Audio Asylum on Ground and Hum solutions
- Chasing Internal Ground Loops (DIY tube builders only...)
- More Advanced Steps
- Old Links to Useful Web Stuff on Grounding
A bigger question is "What is a ground in an amplifier?"
A perfect ground is zero noise and zero resistance between any of the resistors, capacitors, transistors, ICs that connect to the ground. But we don't live in a perfect world, and we don't have perfect grounds.
The physicists working on early electrical amplifiers found a loophole - we can shunt a LOT of noise onto ground and signals would simply ride on top of them, like a boat on top of ocean waves. IF you can create a low resistance ground between the parts of the amplifier.
Star grounds, single-point grounds, ground planes, ribbon grounds, chassis grounds - all are approaches to minimizing resistance between parts. When the parts of an amplifier see different resistances, noise from the ground can be shunted into the signal path - we call that a leak.
Cabling components together exacerbates the number of variables. Now we can have ground noise coming from another component travelling over the ground of an interconnect cable - not a great connection.
But there is a potential fix: Electricity follows the lowest resistance path; if you add a bigger, better ground wire between the components the ground potential and noise will follow the better conductor, noise will be equalized between the boxes and output noise be reduced.
Unfortunately, I am no stranger to hum since I also have a home theater setup. Every time I turned on the TV in the home theater system in the 1990s low-level hum went up about 5 dB. Seems that I had a ground loop/leak with my TV Cable connection.
This is just an illustration, but every high-resolution system I have built has had, well, unique issues with grounding. If you don't get these system ground loops minimized you cannot even hear the real hum level from the preamp. Ground loops are caused when the interconnect cable offers a better path to ground than the chassis/power cable ground of the component itself (hope that makes sense).
Even if you lift the ground on the preamp or receiver you are still in the system ground loop. Interconnect grounds (input jacks) are all on the same ground bus you can get leakage from one source component to another *through* the connections; worse, you can also get leakage between the amps and source/inputs via their cables since the ground bus also connects to the output RCAs.
Here is, in order, what I would look for:
1) Does the Hum change with Volume change?
Independent of volume = preamp output loop
If the hum goes up and down with volume changes, then the issue is probably between your preamp and your amplifier... and/or between your amplifier and your subwoofer amp...
Dependent of Volume = source component loop
If the hum stays the same level when you make volume changes, then the issue is probably between your sources (turntable, phono stage, CD/DVD player, TV/HDTV, streaming box) and the inputs on your preamp... or receiver...
2) Noisy Tubes (no tubes? skip to next step...)
Some tubes are just noisy, or hummy. Have you tried swapping tubes to see if the problem moves from one channel to the other? Have you tried alternative versions of your tubes? Doc Bottlehead chose an inexpensive tube, and it's interesting to hear the differences between tubes anyway...
3) System Ground Loop
To find them, hook up your multimeter probeas and set the thing to AC voltage, then start bridging the probes between chassis contact points for various components. Check all of the possible combinations and, believe it or not, make a little chart and write the values down before you start making changes.
Now comes the fun/tweaky part - make a few ground cables from, oh, 16 ga stranded copper with crimped eyelets on the ends - now start trying all possible combinations of added grounding between your components by removing a screw from each and adding/removing the new cables between the boxes. Write down the results of each trial, because this gets confusing in a hurry.
Remember, the only way to see the ground loop voltage (or current) is with the system up and running. If there are any components that you only turn on 'some of the time', then map the voltages in both states just to be sure.
4) Cheating the Digital Devil
Digital preamps, HDTVs, DVD and Blu-ray players, Streaming internet boxes, TV cable boxes - all of these can create digital hash on grounds. A quick way to eliminate this is to switch digital connections from wire (RCA-type digital cables) to glass fiber (TOSLINK has no electrical connection).
5) Blaming your TV Cable Company
Noise and hum can also come in from your TV cable. It's a huge system connected across miles of local distribution only with the ground shield in Co-Axial cable - there will be noise, ground loops and leaks that can all sneak into your system through their cable. Isolation transformers can beat this but they can also degrade your signal - avoid cheap trannies!
Bonus potentially dangerous activity
You are responsible if you decide to do this - Get several extension-cord 'cheater' plugs and file down the tall (polarized) conductor on the plug (you do live in the US? Don't even think about doing this with voltages higher than 115). You can use these adapters to 'lift' the ground, or change the polarity of individual components. Again - make a change, measure the values between the components, write them down.
The goal is to get all of the AC voltages down as far as possible. Now the bad news - you can probably never get it down to zero (unless you have a broken meter), and some components contribute more noise than others. In general I like to get the voltages between the source components (CD, phono preamp) and the preamp down as far as practical, even if that means that other boxes have slightly higher voltages.
Posted by mikech on April 13, 2000 at 08:33:54:
In Reply to: Re: noise on the ground posted by steve on April 12, 2000 at 16:04:04
Noise, hum and who knows what else. Do you know if your 2-prong plugs are correctly connected in the wall socket? This is a long standing tweak... For each component in your system, measure the AC voltage on the chassis (touch one lead to your grounded wall outlet and the other to the components chassis). Next, flip the AC plug orientation in the wall outlet and measure the voltage on the chassis again. You'll find that the orientation will effect the amount of voltage read. Orient each AC plug so that there is the least amount of voltage measured on the chassis. Be sure to do this with your interconnect cables disconnected to that each piece of equipment isn't grounded to one another. The difference can be quite amazing!
Oh, if all of your equipment is ungrounded, just run a single wire from your AC outlet ground (the center screw) to one of your components (the preamp).
Posted by mike hamilton on May 01, 2000 at 14:11:55:
In Reply to: Hum through speakers from preamp.... posted by BobY on April 30, 2000 at 21:06:30
In my experience hum is not "coming" from the preamp but rather from a ground loop between your amp and preamp OR noisy ground. I like to think of this noise as riding on your signal. If you have your amp on all the time and the preamp in standby and you notice the hum goes away, thats because your preamp has no output current and there is nothing for the ground noise to "ride" on.
The ground noise happens because you have different lengths of ground wire between your preamp and earth and the amp and earth. The difference in lenght can cause a difference in potential and therefore current can flow (using the interconnects as part of the path).
I have a long run from my preamp to amp, they are plugged into different outlets and different circuits. I tried using an extension cord so the amp and pre would have the same power source. This did not do the trick. Still hummmmmmm. The ground loop thru the interconnects and the power cord still exists.
So I installed Isolated Ground outlets and ran a dedicated ground to both. The dedicated ground consists of a new ground rod on a different side of the house than the main AC grounding rod. The new ground wires (12AWG or better) are run separately and are equal length from the ground rod to each outlet.
No more humm.
Posted by VoltSecond on July 13, 2000 at 22:58:57
In Reply to: Re: F'play hum: ideas anyone? posted by Ben on July 12, 2000 at 16:07:50
Unplug all of your inputs to the Foreplay. Temporarily wrap the sides and bottom of the Foreplay with one piece of aluminum foil and attach the foil to the ground of both of your unused inputs. Make sure you don't short to anything inside the Foreplay. Put plastic tape over the bottom first if you are nervous. This will give you some shielding. Do not run the preamp long with the air blocked from the bottom. If this works, build a shield inside the base using ¼ inch galvanized steel screen.
If the hum went away, turn the power off, plug one input back in and check to see if the hum comes back.
Leave everything hooked up to the Foreplay and unplug everything from wall (115V/TV etc) but the Foreplay and the amp. Unplug the subwoofer if ya got one too! Turn the amp and Foreplay on. If the hum goes away, try plugging the other equipment in one piece at a time.
Same as Idea 1, but this time unplug all the left channel connections. You may want to plug the amp's left channel input with a 10K resistor.
Have you checked the "polarity" of your equipment? On each piece of equipment, unplug everything but the 115V. Measure the chassis voltage to safety ground with a DVM with the equipment on. Reverse the polarity of the wall plug (use a cheater if you have to) and recheck the chassis voltage. Adjust the wall plug for each piece so that the chassis voltage reads the smallest voltage.
Move the Foreplay, amp and one speaker to a different room to see if something in the room or that particular wall socket is bothering it.
Get a wall socket tester and check to see if the socket is wired right.
Comment on 10K shorting plug:
At Radio Shack they have cheap RCA connector plugs and 10k resistors. In the RCA plug, solder one end of the 10K to the center post, solder the other end of the resistor to the outer ground connection, clean all the flux off and slip the plastic sleeve over the entire mess. I usually put the body of the resistor half way inside the center connector pin. Make a few of these, I find them handy.
Make one with a 75-ohm resistor and plug the unused digital out of your CD player with it.
Posted by Jon Risch (i) on July 03, 2000 at 12:26:22:
In Reply to: Do expensive power cords sound better? posted by henry the eigth on July 03, 2000 at 00:57:51:
I have posted this before, but the search engine seems down, so: Connect a RS #15-1140 and a #15-1253 (both about $3 each) back to back using the 300 ohm leads, and place this in line with your cable TV feed right as it comes into the system.
This should lift any ground loops, and provide isolation from the Cable TV line.
Some have claimed that these inexpensive RS devices will loose signal on the top RF ranges, but I don't see any real problems unless you are trying to do this on a satellite type signal, regular Cable TV should not be a problem. If you are already using multiple signal splitter to send the signal to more than one device at a time, then the signal strength loss due to this may push you over the edge with these back to back baluns. Then a booster amp is in order, also available at RS.
Posted by Steve Hornsby on January 18, 2000 at 14:02:53:
With interest I read some of the posts on hum problems. I was experiencing some hum on my Foreplay and managed to get rid of a bit with a 3 pronger - but still had some hum. Weird thing was, the hum was low with my Heathkit AA-151 (2 pronger) and seriously there with the Celeste 4070SE (3 pronger). I tried some of the fixes (mostly grounding) to no avail.
Fluked out however when I moved the output cables on the Foreplay. The simple movement of these cables towards and away from the transformer increased and decreased the hum respectively. No small change! A whopping change! Enough so that with the cables coming off the sides of the FP, the hum is down to a point where it is not apparent unless I am quite close to the speaker. I don't even know if the hum I have now is normal, but it doesn't intrude on quiet passages.
Question 1. The cables are cheaper AudioQuest types that will soon be on the second system. I am looking for new ones. Are there any specific properties of cables that I should look for that make them less sensitive to the ???whatever it is?? coming off the transformer.
Question 2. Is there a simple way to sheild the cables from the transformer. Some amps seem to have nice square enclosures on their transformers. Are these enclosures aesthetic or do they serve to sheild?
Simply moving those cables made my audio day. Maybe some other simple ideas on shielding would make this (my) beauty of a preamp damn near silent!
Thanks to all.
Posted by Rick McInnis on January 20, 2000 at 09:32:16:
In Reply to: Output cables source of hum! Any ideas? posted by Steve Hornsby on January 18, 2000 at 14:02:53:
I was having a horrendous hum problem from my turntable mounted moving coil transformer to the phono pre-amp. I have found complete and utter silence by using Allen Wright's silver foil wrapped around the cable. (Alpha Core "Micro Purl Silver", very trick and inexpensive)
I had this stuff already to use for speaker cables. I would try copper foil to save money. But, nonetheless, I now enjoy perfect silence from hum and residual RF.
It is interesting to me that there seems to be no need to sheild the wires coming from the cartridge. But critical sheilding was needed from the output. This is not what I would have expected!
It seems to me it would be easier to shield the cable than to shield the transformers. Yet, there is the question of sonic penalties from cable sheilding... Why is compromise always required? Oh, well.
I've not heard any discussion of how sheilding affects transformer sonics. But, there must be some difference, I would think.
I know this isn't the same problem you are having but it is similar and I would think in an even more noise sensitive position.
Posted by Paul Joppa on January 18, 2000 at 19:06:01:
In Reply to: Output cables source of hum! Any ideas?
posted by Steve Hornsby on January 18, 2000 at 14:02:53:
Check the Jensen transformer website for some good information on hum and buzz. Sure, they are trying to sell transformers that cost more than a whole Foreplay, but their "white papers" are full of very good information.
With two devices using 3-prong line cords, there is a ground loop set up which will pick up ANY magnetic field. Moving the wires changes what parts of the field are picked up, but the sensitivity remains. Any electrical device which has a magnetic field will radiate into the loop. Look carefully at John Camille's articles and you'll see the optional ground switch, which connects the signal ground to the chassis through a small ceramic cap - grounding at RF but not at 60Hz. I think he specs a 0.01uF cap, but you might benefir from one 10 or even 100 times smaller.
Just like a system ground loop only smaller, a preamp can have sections with higher ground potential that use unexpected parts of the circuit as returns, and that can cause noise.
I don't know anything about your particular kit so I can only ask questions here - Does it use the Bottlehead standard ground ccheme? Do you have any other upgrades besides the attenuators? Did you use the stock 2-conductor power cord? Would highly recommend the braided 12 ga grounded power cord upgrade (see the Bottlehead Forum), and try connecting the ground in 3 different ways: 1) directly to the main ground tab, 2) with an RC network connected to the same tab, 3) with a yin/yang diode pair connected to the same tab. Again - make a change, measure the values between the components, write them down.
OK, it's not tubes anyway!
Inside the Foreplay, you might try moving the resistor-capacitor (RC) filter, or adding another, between the power inlet ground and the chassis ground. Also, do you have your power transformer isolated with a separate ground connection? If you add some rubber grommets to the XFMR mount holes on the chassis and a strip of isolation rubber under the transformer frame, you can then tie in an external ground tab from one of the mounting bolts. Try different gnd points for that (to chassis, or to the power inlet ground, with and without RC or yin/yang diode filters).
Anyway, I'd look at the system ground loops first, and then try the ground mods on the F'play.
You can measure the ground potentials within the preamp in the same way as the system problems I described before - be careful, you'll have to have the preamp powered up and connected to your system to see the true condition! Measure AC volts between different sections of the ground bus and see if you can minimize whatever stray leakage is occurring by altering the PS grounds first; if that doesn't work look at sheilding issues (apply copper tape to the outside of the filament AC line and ground that to the chassis or power gnd) or altering your ground bus configuration.
Wow, this sounds like a lotta work, so that's why I would start with the easy stuff like the system loops, and then work my way up to a tweaking lather on the Foreplay.
First off, another disclaimer: the Foreplay I built was decidedly non-stock, with a totally different ground bus and power-supply filtering.
Posted by Quest on January 03, 2000 at 11:57:55:
In Reply to: Hmmm--Foreplay Hum posted by J. W. Hall on January 03, 2000 at 10:34:14
Hello folks, The hum problem you guys experienced could be 1 of the following problems:
(1)AC wall plug polarity reversed. Try plugging it the other way round and see if that helps.
(2) You have to "twist" the black(primary) wires, the red pairs (secondary B+) and the green pairs (filament). This is especially true on the green wires because you are heating the filaments with AC. Any stray magnetic fields caused by not twisting the filament wires together will be induced and picked up by nearby components or wiring and it will be amplified.
(3) ground loop. Have you double check to make sure your RCA jack's grounding tabs are all soldered properly and tightened down securely? If not there will always be a possible ground loop inducing hum into the system.
(4)Interconnect wires too close to the unshielded power transformer. Try relocating the interconnect wires and see how it works.
(5) Transformer shielding problem. Stray magnetic lines is cauing the trouble. You may try using some copper/iron sheets to shield it.
To me point #1 2 and 4 the highest candidate of all. However I've seen guys who had trouble in point #3 and once the problem is remedied the hum is pretty much gone.
Posted by J Epstein on July 15, 2000 at 10:11:18:
In Reply to: grid series resistor posted by Bob S. on July 15, 2000 at 00:22:08:
Common terminology for this is a 'grid stopper.' In combination with the inherent interelectrode capacitance of the tube itself (this is the capacitor made up of the tube's internal elements, not an external capacitor) it forms a low pass filter with a very high frequency cutoff. This stabilizes the circuit with respect to high freqency parasitic oscillations, which can be a big pain in the ass to diagnose and cure - so it's usually easier to assume they will be there and prevent them at the outset.
This part should NOT be mounted on a tagstrip, terminal strip, etc.: it should be soldered DIRECTLY to the tube socket pin, with the minimum possible lead length between the resistor and the socket pin. This prevents the lead between the tube and the resistor from acting as an antenna for high-frequency pickup.
I usually slip some heat shrink over the pin, resistor, and solder joint bewtwen the resistor and the incoming lead to keep the whole thing neat and sturdy. There is very little current flowing through this resistor so it can typically be a 1/4W or 1/2W part.
These precautions are most necessary in high-gain circuits but parasitic oscillations can strike in the oddest places. And as I said before, they can take a variety of forms and prove very hard to diagnose - I have had them show up as un-tameable hum, seemingly microphonic tubes, etc. What happens is, the parasitic oscillation rides on any noise that might already be there (but at harmless levels) and masquerades as a louder version, sort of like an AM radio signal riding on its carrier wave. A nice fast oscilloscope is a good tool for spotting them - you'll see a sawtooth "fuzz" on whatever pure tone you are looking at.
Posted by Paul Joppa on July 15, 2000 at 20:40:15:
In Reply to: Re: grid series resistor posted by J Epstein on July 15, 2000 at 10:11:18:
I had a "hum" problem once, that was actually a parasitic. While poking around, I noticed that the LEDs on my soldering iron (which indicate the temperature) were flickering, and they went steady if I turned the amp off. The @#$% amp was acting like a radio transmitter, and the soldering iron was picking it up!
A few grid stoppers, and all was well 30 minutes later.
Posted by Lukasz F. (i) on June 28, 2000 at 15:43:12:
In Reply to: tube gear hum posted by piezo on June 28, 2000 at 14:06:53:
Just guessing :
1.try your interconnects for broken ground wire. It can be broken 'not completely' and this problem shows just sometimes. You still get the music but with hum. Open the plugs and inspect, also try with a meter (while moving the cable where it enters the plug).
2. open the AC outlets and tighten screws.
3. Use a piece of copper wire and touch the chassis of cd, pre and amps connecting one to another. Listen if buzz goes away.
4. some gear uses floating input winding in the transformer. It may have symmertical winding with earthed center tap. If there are any AC capacitors in the neighborhood or in any one of your gear, and some ground problems occur, the capacitor can conduct 50 (60) Hz to the chassis of one equipment. Ground the chassis individually to a water pipe and listen for improvements.
This should cover most cases of hum. Otherwise you may have a particular problem with AC filaments that requires a balancing anti-hum pot in the filament circuit.
Posted by olen on January 03, 2000 at 20:39:26:
In Reply to: Hmmm--Foreplay Hum posted by J. W. Hall on January 03, 2000 at 10:34:14:
I had the same problem, so I grabbed one of the little aligator clip jumpers out of my tool box and clipped one end to the foreplay's chassis and I tried clipping the other end to the amp, then the cd player, and the hum was still there. So I tried clipping it to my DSS receiver which is grounded via the coax, and the hum all but disappeared. I suppose you could do the same thing by using a three prong plug and connect the chassis to the gound, or run a jumper to your nearest gas or water pipe. olen.
Posted by Doc B. on January 05, 2000 at 08:45:49:
In Reply to: Re: Hmmm--Foreplay Hum...grounding? posted by olen on January 04, 2000 at 19:00:48:
[discussing 3 conductor braided AC Cable mod for Foreplay]
I connected at terminal 13, the one that ties the ground buss to the chassis plate. You might want to go hardcore, isolate the power transformer from the chassis, attach the transformer frame to the third wire, and attach the third wire to terminal 13 through a pair of opposed diodes in parallel.
Posted by Quest on January 06, 2000 at 15:17:59:
In Reply to: Re: Hmmm--Foreplay Hum...grounding? posted by olen on January 05, 2000 at 11:37:47:
(1) go to hardware store and get some small rubber/vinyl (I personally prefer anything that is soft and supple) grommets. Over-bore the transformer mount hole a tad bigger and slip the 2 grommets into there. Now with a bit of care you can suspend the transformer from vibration and noise.
(2) more? take a long narrow strip of copper foil (I bought mine from hobby shop that sells HO scaled model train) about 2cm in width. Wrap the transformer tightly following exactly the transformer winding direction (hard to explain in words) and solder the end to secure it. Now use a small insulated/bare wire to "drain" it to that transformer mount and from transformer mount connect to the tab #13 (the common buss joint).
(3) Bet you want more, more, more?? Here it goes: according to Doc's secret KGB stuff (no it's actually from 'Buddha') you can use a bi-directional diode arrangement to 'drain' the eddy current from the transformer mounting frame to the common ground. As a result you get improved S/N ratio and largely reduced noise level...... Try it out.
p.s. Take it to the MAX?? how about potting the power transformer with wax or commercial potting compound? This is going to be my next mod/tweak!! (Don't forget to ground the case properly)
Posted by Paul Joppa on July 14, 2000 at 19:46:25:
In Reply to: F'play hum: ideas anyone? posted by mingc on July 12, 2000 at 15:20:17:
Oh ... MILDLY volume dependent? That's different, I thought you meant it varied with the segnal level in direct proportion.
If there is hum with the volume turned down all the way then it is probably not coming from the signal lines. Are the filaments properly biased up to 30v (or whatever it is)? It might be electric field pickup - try VoltSecond's shielding suggestion.
Incidentally, if you need shielding you can buy from Hammond a 2" high aluminum box, 6"x10", which will just fit the preamp. Doesn't look as nice, but it will provide some shielding.
Here's another check for the ground loop type of problem. Remove the cables between CD and preamp, and connect your meter between the outer (ground) RCA connector of the CD and preamp. Measure the AC current with both turned on.
Try it with and without the output connected to the amp.
Try it measuring the current between preamp and amp, with and without the CD connected.
I measure 20uA (20 microamps) this way between my preamp (old Sony reciever) and amp (modified S.E.X. kit). If you measure a lot more, the hum may come from that current making a voltage drop across the resistance of your cables.
Posted by rcrump (i) on July 09, 2000 at 10:41:28:
In Reply to: capacitors in line filters posted by dude on July 07, 2000 at 14:49:22
Victor is correct... Only use caps that are rated for AC such as the oil bath PPN types that are internally fused motor run caps....These are normally housed in a metal case so measure the AC potential on the case and go with the lowest with AC hooked up both ways... Put a drain resistor of 200K-250K across the cap for safety. I suspect that you have a problem with the cap you used or the cap is hooked up backwards with respect to AC potential if you encounter hum.
Useful in the early 2000s anyway!
- Audio Asylum See the geeks playing nicely - Great search engine
- Preamp Grounding Tutorial Great page!
- Jon Risch's Web Pages Wiring, DIY Cables and DIY Power Filters
- Steve Bench's Vacuum Tube Site Way cool technical stuff
- Quest's Foreplay Pages
- Duncan Amps Tube Specs and Cool Technical Assistance
- James Melhush' Tubes and Speakers and Links
- Bonavolt's Audio Electronics
- Experiments in Electrocution DIY tube amp stuff