Hum Bug Fixed

Noise at left speaker terminal (1V scale)

My kitchen system has had an increasingly awful "hum" (mostly 60 Hz AC line frequency and harmonics) for the last 3 years or so.  I finally had to do something about this, ironically, on Christmas Day 2024.

I used to pay very close attention to ground loops and other hum causing issues in the kitchen audio system ever since it first started coming together around 2005.  It's always been a very complicated system including VCR, digital video recorder, multiple monitors, inside and outside antennas, Tivo, big network switch, computer with loads of peripherals, etc.  I took great care when adding each new device to ensure a ground loop was not being added, and if it was, I found some way to eliminate it.

But what happened around 2022 is that the DEQ 2496 that was handling my subwoofer EQ, and which I was also using as an RTA to verify that the hum at 60 Hz was near zero (typically -110dB), summarily died after about 10 years of use as they have tended to do.

I then struggled to get the subwoofer working just as it had been before, but using the EQ built into my HTR receiver (a Yamaha HTR-5790).  But unlike the Behringer, the receiver lacked the balanced outputs to drive the small chifi preamp which gives me a subwoofer level control.  So my carefully constructed grounding scheme, reliant on a balanced connection in the middle, was no more.

And lots more stuff happened too, I just wasn't keeping track of things anymore because I didn't have a handy RTA to keep checking that things met the same hum free standard.  So I didn't much bother, and as might have been predicted, it just kept getting worse and worse as I added new features and/or devices.

Then last week I discovered to my horror that I had been connecting the Hafler 9300 power amp to the preamp output of the HTR-5790 using a 2000's vintage 12 foot Radio Shack standard interconnect.  Betting that was a problem, I ordered a new 13 foot Blue Jeans LC-2.  These are RCA terminated flexible coaxial cables with the best shield coverage you can get, along with the lowest capacitance, and they are as robust as it gets.

But as I've been rather busy with parties and preparations prior to Christmas Day, I didn't get around to installing the cables which had arrived last Friday until after my big Christmas Day dinner (a buffet with friends).

I wanted to make sure I had good before and after measurements, so I got out my Meguro MN-455B noise meter, which is a very useful thing to have (and I needed to have a real electronic "noise" meter since the 1970's, there has been hardly any other way to get "A" weighted electronic noise measurements, but I only got this wonderful Meguro in the 2000's on ebay).

Because the Revel M20 speakers have widely spaced connectors, I couldn't just plug in my coax-to-double-banana adapter into the 5 way binding posts.  I had to rig up an extended connection for the minus terminal using two more double bananas and a short wire.

When it was all together, I was shocked to discover, that with the Yamaha actually turned off, but the Hafler on, I was getting 0.88V "A" weighted hum and noise at the speaker terminals.  No wonder it was beginning to get annoying.  It made no difference whether I had the Yamaha powered on or off.

I removed the input connections to the Hafler, and the audible hum vanished.  I could not hear anything from the normal listening position, and there was just a trace of hum at the speakers themselves.  But the A weighted noise level had only gone down to 0.4V.  That was because the Hafler 9300 has a fairly high 47k ohm input impedance, and a 500 kHz bandwidth, and the "A" weighting isn't even defined above 20kHz (but is often just extrapolated, not brick-wall filtered at 20kHz because it's inaudible).  I'm not exactly sure what the Meguro meter does on the "JIS A" setting.

But open inputs, a high input impedance, and high bandwidth ensures there's going to be a lot of high frequency noise, even if there weren't much anyway.  That's what I think is going on...the hum went away but the HF noise was even greater because of the now open inputs.

I could have measured with shorted inputs, as John Curl himself told me I must do in a message after I complained about the noise level in my HCA-1000A varying with input level position.  But I didn't bother this time, because I know the amplifier is near perfect anyway.  I didn't need to test the amplifier hum, what I wanted to do is find a way to have it hooked up without hum.

So I rolled out the new 13 foot cables, only to discover that 13 feet was actually way more than needed.  It left me with an extra 3 feet curled up on top of the amplifier (until I found a better idea).  What this revealed was that the original 12 foot cable, and I pulled it out to be sure it was a 12 foot cable, was twisted around so much stuff that it was only barely long enough, even though it should have had 3 feet to spare.

Even with the excess cable piled on top of the amplifier, but plugged into it, the audible hum was gone and the noise level was now 0.19V.

New cable with excess piled on amp, first time

But finding the Blue Jeans cable was too long left me in a quandry.  Do I send the cable back, and get one made to the correct length?  I wondered if Blue Jeans had cut the cable too long, so I removed it.  Only to discover that they had just barely cut the cable by an inch or so longer than the 13 feet, not including the jacks.  Radio Shack cables barely meet the advertised length WITH the jacks.  Now it was clear how much the Radio Shack cable had been wrapped around lots of stuff, probably mostly AC line cords on the floor.

I still couldn't make up my mind, so I reconnected the LC-2 cables just to have something working.  But now the hum level had gone up, to 0.6V.

WTF!!!

I quickly discovered that even with the best-in-class shielding of the LC-2, it makes a hell of a difference how the cable is laid with respect to AC line cords and other stuff.  Particularly bad, it seems, is the back panel of the UPS that powers most of my system.  If I run the cable next to that back panel, and it tends to fall that way if I am not careful to avoid it, hum goes way up.

Hum came back when cables ran this way

Rather than experimenting with each possible variation, I decided to just do-it-right from the start, by arranging the cable to be as as far away from everything, and especially AC power cords and the UPS, as possible.  I used my AV stapler to staple the LC-2 cables to the wall (the staples are just barely big enough, and a lot of them failed to attach properly and had to be removed, but it seems to be holding now) way above everything, then over to the table leg where they are held on with velcro ties, then down to the amplifier.

This got me to my lowest measured level, but not actually much better than my lucky first hookup, at 0.18V.

Noise after re-routing new cables

It sounds hum and noise free even ear to speaker.

I used to ignore hum and noise.  I started taking it seriously around 2002 when I found my incredibly well built 1978 Pioneer Series 20 D23 4 way electronic crossover had developed a ripple-hum (around 120 or 180 Hz).  That was what ultimately led to my using digital crossovers, and finally DSP based crossovers.

The kitchen Hum had become very irritating when there was no music playing.  But I never noticed it when music was playing, and even with it present I still felt I could make very fine audio judgements.