What else can I do to make SEX 2.1 quieter?

[Todd R]

The amp is sounding great, and installing the impedance boards helped a bit in quieting the amp because I could match the impedance more closely to the phone in use, but I need to eliminate the remaining noise.

The issue is when I use my JH Audio 16 Pro monitors (Input Sensitivity: 118dB @ 1mW Impedance: 18 Ohms). I still have a hum/buzz that I can hear. It still occurs when disconnected from the source, it does not get louder or quieter when I change the volume, floating the ground does not change it, and as soon as I hit the power switch the noise goes away instantly. I have tried it plugged directly into the wall as well as into my PS Audio PPP regenerator.

These are the only headphones I can hear the noise with. It is not an issue with any other headphone I have tried it with.

Any suggestions for what I can do to get it quieter?

 

[Grainger49]

Sounds like heater induced hum.  Pin 8 of the tubes should be grounded.  Check this.  The heaters are DC, which should be quiet.  But I'm pointing a finger at the heaters.  Someone else might correct me.

 

[Todd R]

According to the assembly instructions, the cable shield is grounded, but not pin 8.

 

[Grainger49]

I was looking at the schematic from the old manual.  There is an earth/ground sign on the lead to pin 8 on the old schematic I have.  Actually on one leg of the 6.3V transformer.

 

[Todd R]

I see that too. I'm looking through the instructions and am not figuring out where that connection was ever made unless it was one of the mounting bolts to the safety ground?
The only other one I see is from terminal 10 to terminal 3L.
That's all of them, correct?

 

[Paul Joppa]

The new power transformer (PT-7) has a center tap on the filament winding, which is taken to chassis ground to kill the capacitively coupled noise from other windings. It's transformer terminal 10.

Current production SEX 2.1 no longer has the 120 ohm resistor in the headphone line. That resistor, while problematical with some specific headphones, did reduce the voltage sensitivity of low impedance headphones, bringing them in line with the high impedance phones. (For the OP's 18 ohm phones, the difference would be 18dB.) However, the special switching jack is no longer available so that feature had to drop out. You can get the same effect by switching to balanced operation, since you have the impedance switch boards. What value resistor is installed?

Making the amp itself even quieter than it is would take some experimenting, in order to find out exactly where the noise originates.

 

[Todd R]

Paul,
I have the 60.4 ohm resistor installed.
I have tried the headphones in balanced and it didn't make a difference.

 

[Paul Joppa]

That's odd. It doesn't make a huge difference, but you should see some attenuation - 8 or 9dB. Only in the lower two impedance settings though. Don't match the "impedances", use the 4 ohm (LO) setting to get the lowest voltage out.

 

[Todd R]

Any other thoughts?
Maybe better components somewhere or a different grounding scheme?

 

[fullheadofnothing]

I've noticed that some 6DN7s are hummier than others. My coin bases are unlistenable, for example. I have another pair that's ok for speakers, but way too hummy for headphone use. The tubes are cheap, so get some more and eXperiment if you haven't already.

 

[Paul Joppa]

Any other thoughts?
Maybe better components somewhere or a different grounding scheme?

Attenuating the headphones works for all noises, but actually reducing the noise depends on why it is noisy. In the development phase, the first thing we did was to add DC heating on the tubes (you might want to check that the center-tap is actually connected to the chassis ground). The next thing we did was to add a de-coupling stage in the power supply for the driver section. Those were the two biggest sources of hum. So now you are down to diagnosing what is the next-biggest one. Putting larger or better caps in the power supply (for instance) won't help unless the ones that are there are too small. Same for replacing a resistor with a choke.

 

[Todd R]

I've noticed that some 6DN7s are hummier than others. My coin bases are unlistenable, for example. I have another pair that's ok for speakers, but way too hummy for headphone use. The tubes are cheap, so get some more and eXperiment if you haven't already.

Which ones did you find that were quiet?

 

[Todd R]

Any other thoughts?
Maybe better components somewhere or a different grounding scheme?

(you might want to check that the center-tap is actually connected to the chassis ground).

It is

 

[Doc B]

This is assuming that you are hearing 60Hz hum - if I had to guess it might be that the one set of headphones that is picking up hum is the only set sensitive enough to pick up hum induced by a bit of magnetic coupling from the power trans to the signal wiring. I've been experimenting lately with some different cables and I have determined that a very symmetrical coax cable such as Belden 9259 or 89259 seems to do an excellent job of rejecting that kind of noise. So it might be worth experimenting with replacing some of the output wiring with coax.

 

[Grainger49]

Dan,

I've been using 9259 since the late 80s.  I didn't' know anyone remembered it.  Really GOOD coax.  I even bought a double 9259 back about 1988. 

Todd,

Reading Dan's post above, could you shield your headset wiring?  Just as an experiment, wrap it in aluminum foil (just nearest to the SEX amp) and touch the foil to the top plate.  This should help if the headset wiring is picking up some hum from the power transformer.

Ok, I have a SEX 2.0 manual, and looking at the completed picture...  Hanging on the statement that the hum disappears when it is turned off...

SWAGGING here, try getting the extra wires of the output transformer (maybe not in the 2.1, I don't know) away from the power transformer. 

Try STP to the speaker outputs and the headphone outputs.  Of course grounded at one end but not both.  In the case of the headphones maybe Shielded Twisted Triplet? 

Wrap a piece of foil around the outside of the Solen output caps and ground that. 

Doc and PJ have done so much to eliminate noise that is all I have.

Maybe VoltSecond can help.  He knows more about killing noise than I do.

 

[Doc B]

If you like 9259, try 89259, the foamed Teflon version. We've been comparing the two and 89259 is even better. Unfortunately it seems to be about 10X as much money by the foot.

 

[Grainger49]

Yeah, I got some surplus 9259 when I specified 500' for an antenna job in 1992.  (Surplus got me my Tektronics oscilloscope and my 2.5kVA isolation transformer {both pulled from a dumpster})  I also soldered the F- Connectors for the job (not a union company so Cliff let me do the soldering).  I'll have to get some 89259.

I'm not sure, maybe the double 9259 was numbered 89259, or not.  I'll look.  It is on the cable rack behind my stereo.

 

[fullheadofnothing]

To answer your question Todd, my quietest tubes are GE, but I would guess that noise is more variable tube to tube than brand to brand.

One other thing I was thinking about, have you put shorting plugs on the input to rule out noises from outside the amp (source, cabling, et cetera)?

One question I have for the group, the headphones in question are 118dB/mW. Assuming my math is correct (huge assumption) that means a nanowatt would produce noise of about 60 dB. Is this asking too much of the circuit?

In regards to the 120Ω resistors, would it be feasible to put them into an inline headphone adapter? This especially would make sense if they were terminated with a 1/8" plug and you needed an adapter anyway, but even a 1/4"->1/4" that you only use with the phones in question might make sense.

 

[Paul Joppa]

...
One question I have for the group, the headphones in question are 118dB/mW. Assuming my math is correct (huge assumption) that means a nanowatt would produce noise of about 60 dB. Is this asking too much of the circuit?

In regards to the 120Ω resistors, would it be feasible to put them into an inline headphone adapter? This especially would make sense if they were terminated with a 1/8" plug and you needed an adapter anyway, but even a 1/4"->1/4" that you only use with the phones in question might make sense.

Yes, it is asking too much. The residual noise is about 9 nanowatts (at least, it was with the earlier version; I don't think we've measured it with the newest revision but it should be very similar. That gives 69dB. The 120 ohm resistor will take that down about 18dB, to 51dB. Using the 4 ohm tap will pull it down another 6dB, to 45dB. In a typical room, background noise would be about 40dB(A) - the A scale de-emphasizes the hum but is an accurate reflection of buzz.

I would consider adding both the 120 ohm series resistor and a parallel resistor of around 6 ohms. This will give another 12dB (taking the noise down to 33dB) and will remove any possible underdamping of the headphones. It also reduces the potential power to only 2mW, but that would be 121dB at the 118dB/mw rating, which is still serious overkill. These resistors could be easily built into a suitable homemade adapter.

These are very sensitive phones; the range of sensitivities I have on hand is 84 to 120dB, average about 98dB/mW.

 

[jrebman]

Todd,

Wonder if a copper foil lining of the inside of the wood frame and then grounded to the earth terminal on the AC inlet will do anything?  Plan to try this myself.

-- Jim