Tube amp hum/buzz

[John Warren]

I've been asked about tube amp hum/buzz and how to get rid of it.  It's noticeable buzz when the amp gets to heat and doesn't change when the volume pots are adjusted, it's just there.   This is what it looks like on an oscilloscope, at 20mV/division.  The amp is idling and the load is an 8 Ohm power resistor.  The scope probe is across the resistor over a 100ms time window.     

 

 

Below is with the time window reduced to 10ms.  It's not a sinewave, it's a 60Hz sawtooth.  And with a peak to peak amplitude of 40mW, ignoring high frequency glitches, it's audible especially when you're speakers have high sensitivity tweeters.      

 

 

The 6.3VAC heater secondary tap is prone to picking up wave characteristic (periodic) noise from rectification, line voltage and other sources.  The noise is coupled to the tube cathodes.  Heater supplies configured below (a Fisher design) are problematic.  

 

    

 

There's a few ways to improve the situation (see artificial center-tap, Humdinger potentiometer).  There's other approaches but these are common.  

 

Here I use the artificial center tap concept but replace the "typical" 100 Ohm, 2W resistors with .022uF, 400VAC Orange Drops thus filtering the rails of periodic harmonic garbage and the result (before and after) is below.  RMS value of noise is reduced three-fold and, most important, the amp is silent even when the ear is against the grille cloth.  Note that the peak to peak readings report glitches which are momentary impulses that are of no significance.

 

 

So why did the engineers at Fisher, Scott, Pilot and a few others not add a couple of resistors or capacitors to their designs?  I would propose that the hum, though noticeable with no program material is of such low magnitude that it made little difference to the sound quality. 

 

 

 

[Curious_George]

BOM. The bean counters use to drive me crazy trying to save $.25, which equates to $1 at retail. 

[MeloManiac]

It's an audible 'on' indicator, compare it to an on switch with an led. 

😉

My Leben CS300 also has that hum, which is also mentioned by reviewers. Personally, I'm not bothered by it, as it is unnoticeable when music is playing, and I listen to music all the time. 

As for the why... I think it is part of the original design and the engineers simply didn't want to stray away from it. Out of respect and tradition, or perhaps they feared that taking away that hum would create another problem (bringing down sound quality in another frequency, or heat issues reducing the life of the tubes etc) 

Finally, I read elsewhere that that hum is caused by the layout - design of the Amp. The position of all the parts and the lack of shielding, and bringing it all together in a small box that is surprisingly heavy, but doesn't take much shelf space comes at the price of a hum and a buz... 

 

As for Leben amps, the hum is only present in the original CS300 model. In later models that hum was gone. 

[Langston]

On 8/29/2021 at 6:00 PM, John Warren said:

Here I use the artificial center tap concept but replace...

 

Nice and simple. Haven't seen that before. : ) Another thought is to make sure the big tubes that consume the largest heater currents are closest to the transformer (source) in the heater supply circuit, followed by the lesser current users toward the end of the circuit. This minimizes a potential noise source across the small tubes. The "humdinger" adjustment pot you referred to helped a lot with the Fender Twin, especially after increasing the twist rate of the downstream heater circuit wires.

 

 

It's been awhile since I've been inside tube amps. Most of my experience was with guitar amps I rented out as backline in concert work. Crazy simple and amazingly reliable stuff.

 

God bless you and your precious family - Langston

[erik2A3]

This is ancient.  I must have written about filament hum reduction in valve amps hundreds of times here in the early 2000s.  There are MANY sources of noise in tube amps.  They traditionally have both inherently high input sensitivity and high input impedances - both of which make the potential for noise worse.

 

The schematic above this post represents a filament circuit, with the filament winding (usually green as also shown above) connected to a hum-null potentiometer (aka: humdinger).  There reason a hum-null pot - which simply balances current across the filaments WHERE AC IS USED TO HEAT THE FILAMENTS and helps reduce 60 cycle hum - is because, as is also evident in the schematic above, there is no GROUNDED center-tap present in the filament winding.  The same is the case with the OP's reference to two 100 ohm resistors on either side of the filament winding to ground.  In that case, the two resistors form a simulated center-tapped filament winding, which can similarly be used as a sort of FIXED hum-null device.  However, the resistors naturally do not have the benefit of a potentiometer, which can be adjusted in a manner that 60 Hz hum across the output of the amplifier into a speaker voice coil can be both heard and measured, and thereby reduced to the lowest level possible.  There are also other ways the center tap can be used to reduce hum, one of which involves connecting it to source of 40-60 VDC (or so), such as the top of the cathode of a cathode-biased output stage - in other words, raising filaments above ground.  In choosing between the use of a grounded, center-tapped filament winding and the two-resistor-pseudo-center tap, I prefer the more robust and ultimately safer grounded center tap.  Cost savings was often a priority, however, and two small resistors don't cost very much.

 

In older amplifiers, AC current was used to heat the cathodes of tubes, and this increased the potential for its (the AC's) related noise components, due to proximity between cathodes and their associated heaters, for audible hum at the output stage.  This can be a problem for indirectly heated cathodes, but is of much greater concern with directly heated single-ended (push-pull topologies are different) output devices like the big triodes (45, 2a3, 300B, etc.).  These filamentary cathode tubes present to the builder/user/listener an even greater challenge in terms of residual noise than smaller signal tubes and/or indirectly cathode heated output stage valves.  A 2a3's or 300b's cathode IS its filament, and attention must be given to the 60Hz filament hum that almost invariably becomes combined with audio output.  

 

Almost all of the instances above can be quite effectively solved by using well-designed and filtered DC current to heat a tube's filament rather than alternating current.  What we often run into in the case of DC filaments, however, has to do with the subjective and often debated efficacy of one type of current vs the other.  Which SOUNDS better, AC or DC?  And that is something I'm not going to address.

 

But filament related noise is only one example of many, many possible sources of hum and buzz in tube equipment.  Poor grounding, which results in ground loops; inadequately filtered B+ (for plates/anodes); EMI and RFI.  All are possible causes of unwanted noise that must be dealt with individually and differently.  Simply adding a simulated center tap on a filament transformer with two 100 ohm resistors, for example, is going to do absolutely nothing for 120 Hz hum that's coming from a lack of filtering in the high voltage plate supply, or poorly handled lead dressing, or the angry buzz coming from the dimmer switch in one's listening room.

 

Have fun, erik (who on these pages many years ago used the avatar Erik Mandaville - which also happens to be my name)

 

edit: And so.  If you are somewhat new to tubes and are having problems with bothersome or excessive noise with your preamp or amplifier, please DO NOT turn the poor thing upside down, take the bottom plate off, and start digging around inside for the filament wiring.  To only consider a component's heater supply as the source of noise is a gross over simplification.  Find an experienced technician who is familiar with the high voltages associated with tube equipment to help you.  Chances are, in fact, that whatever residual noise you are experiencing is NOT coming from tube heaters (which is a fundamental aspect of designing tube amplifiers, and something for which a competent designer will have already considered).  Noise can exist in lots of different forms:  high pitched oscillation, cracks/pops, 60Hz hum, 120 Hz hum, fine buzzing, course loud buzzing, extremely strong hum (as in faulty input ground connection or heater/cathode short), and a knowledgeable technician can most often diagnose and repair problems without the use of signal tracers and scopes.  I once rebuilt a large amplifier for a friend who decided to poke around inside his amp to move wires around in the hopes of solving a noise problem.  The amplifier was energized, and he both got severely shocked and blew up the power supply .....quite literally.  In another case, I knew someone who was similarly unfamiliar with tube electronics, but made the decision to change coupling capacitors (to see if he could hear a difference) in his 22 tube (yes, a total of 22 tubes) OTL amplifier -- again, while the amp was energized and upside down on a pillow. This involved soldering an unsoldering inter-stage coupling capacitors in situ in a live amplifier.  Same situation, and I completely rebuilt a very complicated power supply board.  Everything was taken out, rectifiers, resistors, filter caps, all of it.  Be safe.

 

There are lethal voltages in tube equipment.  If you have tried such things as, possibly, a new pair of interconnects (sometimes ICs can have a faulty ground return that can cause noise; or changing tubes; or mains filter -- simple things like that -- and the problem persists, please take it to someone who knows how to safely work on the more involved and potentially dangerous aspects.

 

 

[Langston]

10 hours ago, erik2A3 said:

There are lethal voltages in tube equipment.

 

Interesting post, thanks for taking the time to write it Erik. : )

 

The voltages are amazing in tube amps, but when you need a given amount of power and can't do current you're stuck with it.

 

This reminds me of an event from my concert production days - 2017 and I'm providing sound for a Warner Bros. event in Ft. Walton Beach FL called Bugs Bunny at the Symphony II with full orchestra and a few muso's that toured with the show. One guy brought the original slide guitar from the 30's they found in a Hollywood warehouse that was used by the original orchestra for the Bugs Bunny opening theme. They spec'd a Fender Twin (the source of that schematic I posted above) for the slide, which I brought along with a backup.

 

So the amp starts buzzing during rehearsal after the cocky engineer with the tour connects it to the slide. I offer one of my cables in place of the ragged thing he brought and he says "nope! it's the amp!" I offer the backup and he says "nope! I got this, this isn't my first rodeo." So he reaches down to the underhung power tubes, amp still on and hot, and then proceeds to fly about 5ft of so backwards before landing. A bit reminiscent of a Loony Tunes episode. I help him up and see his hand and say nothing. Obviously. He's the expert. Turns out it really was the amp, probably a cracked tube glass that fell apart when he touched it. It wasn't the heat that got him - it was the plate voltage!

 

We used the backup.

 

Here's a clip of the opening where you can hear the slide guitar start off the theme that modulates from about 250Hz up to 550Hz:

Bugs Bunny Opening Theme.wav

 

Here's the rider I was given for the event for anyone that's bored today:

Sinfonia - Bugs Bunny Symphony 2017.pdf

 

God bless you and your precious family - Langston

[erik2A3]

Langston,

 

What a fascinating story, and thank you for sharing your professional experience in this thread!  Totally true, watch out for plate voltages!  I'm glad he was ok.  It's happened to me too, and every single person I know who has spent any significant amount of time working on tube equipment.  

 

I will definitely look forward to listening to those sound files later today.  Thanks so much for including them!

 

erik

[Curious_George]

I ashamedly admit getting shocked from my fore finger to pinky finger (on the same hand) when I was building the power supply for my 211 amp... 1200VDC. It got my attention real quick. That never happened again or since. When I take the bottom cover off of the 211 amp, I make sure I am fully rested and wide awake. No tinkering with that beast if I'm not wide awake and alert.  

[erik2A3]

Curious George,

 

You are braver than I for sure!  I'm well aware of the 211 requirements, and I think I'm probably just too much of a wimp to work with those.  Half that voltage was bad enough when, I also must admit with shame, I was soldering in a plate load resistor, thinking I had unplugged and discharged filter caps.  That's another important thing to remember! Even caps that have sat for a long time can, if there is no bleeder resistor on the output of the PSU, still have enough charge to give a super bad shock.  Years ago I had an amp (another big OTL) sent to me to look at, and there was still a monster charge in the big reservoir caps after a few days shipping.  I shorted them to ground (which actually isn't the best way to bleed off a charge -- that I also admit -- but I just didn't think there would be much of anything still in those things.  POW!!!