SE Amp with NO HUM

[markw51]

Several years ago I build a few SE tube amps. They all sounded pretty good but they all had at least SOME hum, even if they had DC filaments. I'm in the mood to build another tube amp but don't wany any audible hum this time since I now sit very close to my speakers. The guy who runs Angela Instruments says if you can't take a little hum you should build a push-pull amp.

This leads me to question of whether commerically built SE amps, like Cary Audio, also have hum. If they don't, how do they do it?

Mark

[Sheldon]

I built this and it has no hum: http://www.diyaudio.com/forums/showt...558#post883558

And I mean NO hum. I can put headphones directly on the outputs and hear nothing. True, it's a low powered amp but the same methods can be applied to something with more power.

You will need a quiet supply, either regulated or with enough L's and C's to lower the ripple to an acceptable level. You can figure the level you need by knowing the AC voltage division between the power supply and output. Then, if you are doing an unregulated supply, PSUD is your friend.

I used Guido Tent's DC supplies on the DHT filaments. They work well. If you have a few excess volts to work with on your filament supply, you can easily make a constant current supply for the filaments with three terminal voltage reg. Some just use ordinary voltage regulation and like that too.

Not described in the thread, but later on I also used DC on the driver tube. If you lift the filament reference about 20-30 V, and are careful with your wiring, you may find that AC here is perfectly fine. DC reduced the hum with my amp, but the reduction was minimal compared to DC on the power tube or the inductive coupling mentioned below. Only reason I did it was to use the amp with headphones too. Even then, more of an obsessive thing.

You will see some pictures of the amp in the thread. One refinement also done later that is not shown are two thick aluminum plates between the PS transformer and the OPT next to it. I was trying to get rid of any last traces of hum and found that I had some inductive coupling here, and the plates eliminated it. Make the space bigger between the transformers and you won't have this issue.

You definately can make a silent SE amp. I'm no expert, so if I can do it? And I made life harder than it needs to be by stuffing it all into a small package.

Sheldon

[mwiebe]

I would guess many folks here do not have an issue with hum in the amps they have built, and probably an equal amount wish they could get their amps quieter. So a good part of excessive hum has nothing to do with the amp.

Using AC filaments everywhere, I expect you could build an amp with about 2 to 4 mVAC on the speaker outputs. Will this be quite enough for you? Depends on the efficiency of your speakers, how much background noise there is, what kind of music you listen to and so on.

Outside the amp outside the house/apt, hum can be running on your wall power supply, dimmer switches, refrigerators, neighbor’s welding or futzing around with a linear amp, a host of issues.

Inside the house a power cord could be lying along side an interconnect, hum could be riding on a source signal, an unshielded motor nearby, solid state hash floating around between your system.

You may have a ground issue, sources could have ground issues, your amp could have ground issues and everything could have ground issues together. That third pin on your outlet plug could have a ground problem and not be a ground at all. A simple input transformer could give you bliss.

Sometimes objectionable hum can be attributed to excessive gain in the amplification chain, given your source output and speaker sensitivity. This takes a little head scratching to solve. Getting gain correct can greatly reduce hum and other background noise. I think excessive gain is a bigger issue then many realize which also amplifies grounding problems.

B+ could have more ripple then you think, and spending more attention to a power supply will not only net lower noise but probably a better sounding system.

There are driver topologies which are better at rejecting power-supply noise then others, John Broskie covers this well in his Tube Cad Journal and some of his suggested circuits can even quiet what the welder next door is doing. A change in topology may give you the silence you are looking for.

Steve Bench has many interesting hum nulling suggestions on his site which offer more tricks.

In addition to trying DC filaments, you could also use indirectly heated triodes for output. This may violate some purity law somewhere, but an indirectly heated single-end amp may give you musical nirvana, especially if you don’t have the patience to chase hum and ground issues.

This is a stab at noise issues in amps. You can see many issues of noise have nothing to do with the amp at all, and even a switch to push-pull will not solve them. A hum free amp in a bad enviorment will not make you happy. Good luck.
Matt

[Eli Duttman]

Mark,

Pick a 2A3 based circuit you like. Build with triode wired 6AV5Gs, using DC on all signal tube heaters. PSU ripple should be the dominant factor in the residual hum level you get.

I make no claims about how good any 6AV5G sounds.

[dsavitsk]

Originally posted by markw51
The guy who runs Angela Instruments says if you can't take a little hum you should build a push-pull amp.

The guy at Angela needs to try a little harder. This is a single ended headphone amp with AC filaments all around and is dead silent. The PS is a RCLCRCRC filter followed by a regulator, so ripple is almost nonexistant. You just need to get the layout correct.

[Fuling]

I have a four channel 807 SE amp that is dead quiet on my 94dB speakers.
It uses DC for all the heaters (100 000 + uF and 1,1mH choke) and a capacitance multiplier for the HV.
The PSU is in a separate box located two feet away from the amp.

[phn]

Originally posted by dsavitsk


The guy at Angela needs to try a little harder. This is a single ended headphone amp with AC filaments all around and is dead silent. The PS is a RCLCRCRC filter followed by a regulator, so ripple is almost nonexistant. You just need to get the layout correct.

I second that.

No need for DC if the PSU is properly designed--coupling, coupling, coupling. And regulated PSU is the way to go--insignificant impedance.

[ray_moth]

Hum picked up from tube heaters and other AC bearing conductors is a different issue from PS ripple. The two don't even sound the same - hum from AC conductors is half the frequency of PS ripple (unless you use half-wave rectification). The cure for each is different, too.

Avoiding hum emanating from AC (50Hz or 60Hz) can be challenging, because it requires good practice in terms of layout, grounding scheme, wiring and shielding. Even a poor connection or soldering joint can give rise to this type of hum.

Avoiding PS hum/ripple (100Hz or 120Hz) is easier because you can solve it by using more filtering stages or an effective voltage regulator.

[SHiFTY]

1) Star grounding with multiple stars for power supply, signals, grids and speaker terminals will get rid of most induced hum

2) Proper alignment of transformers and filament/power supply cables

3) The last bits can be got by using shielded cable on low level circuits, DC filaments on small signal tubes and DH output tubes.

[Geek]

I've built a couple of even downright sloppy SE amps when I was younger and I've never had hum.

Sure SE amps are susceptible, especially on the supply rail, but nothing a 10H choke can't cure...