Very Small Hum

[ruffrecords]

I have been doing some detailed tests on the Classic Solo. With the gain turned up to full, there is almost exactly 50dB of gain from the DI input to the balanced output. With the DI input shorted, the output noise is about -56dBu which makes the EIN around -106dBu. If I could add a noiseless 1:10 transformer to that I could have a mic input that would have an EIN of -126dBu which would make me very happy. When I do this though, the output noise rises to -51.5dBu which means the EIN is only just -121.5dBu.  Not  a million miles away from what I normally measure but not as good as it could be. (It is about 1dB worse if I take the top off the enclosure. ) When I listen to the noise I notice it has a small mains hum/buzz in it. This is completely absent when the DI input is shorted so it would appear to be entering either at the mic input or the high impedance secondary of the transformer.

With the top off, I probed about with my finger and discovered the orange drop input capacitor is very hum sensitive and the hum/buzz increases significantly as my finger approaches it. There is no hum from my finger if my other hand is touching the chassis.  I also notice the hum disappears if I disconnect the AC input to the HT supply leaving the mains transformer running and supplying the heaters. The noise also drops about 3dB.

I suspected it might be E field from rectifier spikes so I added a 100 ohm in series with the bridge and the reservoir capacitor. No change.

So no I am left wondering if this is in fact just some general E field pick up by the Orange drop cap.

Any thoughts?

Cheers

Ian

 

[mskeete]

Can you shield the cap?

 

[JohnRoberts]

I am not familiar with what an orange drop cap is, if film there may be an inner and outer wrap. Connecting the outer wrap to the low impedance circuit node could provide some self shielding.

If you touch with your finger while your body is grounded does it get quieter?

It's always something... also is hum voltage significant in context? Hiss may have more noise energy.

JR

 

[sasarist]

I didn't understand  clearly , but if you have OT, in 99% of situation,  rotation of output transformer or PSU transformer will solve hum problem.
Also I am sure that Orange cap is not problem , I have same situation with 1176 rev A ...changed couple of input capacitors and they all react the same....I dont know, probably high impedance input put first capacitor on supersensitive side.

 

[Deepdark]

RFI suppression capacitor??

http://www.vishay.com/docs/26033/gentechinfofilm.pdf

 

[Brian Roth]

Decades ago, I had a similar problem with an Orange Drop in a High-Z circuit.  I solved the problem by wrapping a thin piece of copper around the cap's body, and tying that copper to chassis ground.

Bri

 

[CJ]

Wima parallel plate caps pick up a lot, ceramics same thing,

maybe someting with a metal case,

these are supposed to sound good>

 

[abbey road d enfer]

I have been doing some detailed tests on the Classic Solo.

What is a Classic Solo?

When I do this though, the output noise rises to -51.5dBu which means the EIN is only just -121.5dBu. 

Is it valves or SS? If SS, rthe increase in source impedance produces a noise increase due to input noise current. If valves, that tends to indicate the noise is induced either magnetically or electrostatically.

With the top off, I probed about with my finger and discovered the orange drop input capacitor is very hum sensitive and the hum/buzz increases significantly as my finger approaches it.

This is utterly predictable, as it's a high-Z node and the physical size makes it sensible.

There is no hum from my finger if my other hand is touching the chassis. 

I wouldn't be so sure... I would suggest you do the same test with a small metallic plate connected to chassis via a wire.

I also notice the hum disappears if I disconnect the AC input to the HT supply leaving the mains transformer running and supplying the heaters.
So no I am left wondering if this is in fact just some general E field pick up by the Orange drop cap.

That's what it seems to me. In that case, shielding the high-Z should better things. But what happens now if the DI input is unloaded. The noise should be there... Is it? And does closing the lid change anything? And what happens with the primary shorted?

 

[ruffrecords]

What is a Classic Solo?  Is it valves or SS? If SS, the increase in source impedance produces a noise increase due to input noise current. If valves, that tends to indicate the noise is induced either magnetically or electrostatically.

It is my stand alone version of my two tube mic pre./DI box:

http://groupdiy.com/index.php?topic=60239.msg764154#msg764154

So no I am left wondering if this is in fact just some general E field pick up by the Orange drop cap.

That's what it seems to me. In that case, shielding the high-Z should better things. But what happens now if the DI input is unloaded. The noise should be there... Is it? And does closing the lid change anything? And what happens with the primary shorted?

Thanks for the input Abbey. I did try the test with a small plate connected to chassis but it made no difference. I also tried a grounded plate in various positions with no effect. The buzz drops by about 1dB with the top on. The mic input tests were done with the transformer primary shorted at the XLR input or did you mean at the transformer itself?

I have not tried the unloaded DI test. I will do that today.

I just realised the capacitor is only there to block DC from the input so I will also try with it shorted.

Cheers

Ian

 

[moamps]

.... I also notice the hum disappears if I disconnect the AC input to the HT supply leaving the mains transformer running and supplying the heaters. ..

IMO, if the inside of the DI looks like this on the attached picture, you should rearrange the placement of the PCB board. The 240VAC wires are now placed to close to the input tube and its circuitry. Rotate the pcb 90 deg CW, move output transformer to the left side of the box and use some shield on 240VAC wires. Or/and split this PCB to two and place power supply PCB farther.The HV power supply inlet is too close to the input tube.

 

[abbey road d enfer]

The mic input tests were done with the transformer primary shorted at the XLR input

Ok. Now, is tehre a big difference if the primary is left open? And does hum change with orientation?

I have not tried the unloaded DI test. I will do that today.

Yes, that would give us an evaluation of the contribution of electrostatic noise. 

I just realised the capacitor is only there to block DC from the input so I will also try with it shorted.

Shorted won't change much, because the stray capacitance would be the same. You need to do the test with the cap removed totally, both sides, and a second test with a wire in place of the cap. That would show how much is picked up by the connection from PCB to jack.

 

[ruffrecords]

.... I also notice the hum disappears if I disconnect the AC input to the HT supply leaving the mains transformer running and supplying the heaters. ..

IMO, if the inside of the DI looks like this on the attached picture, you should rearrange the placement of the PCB board. The 240VAC wires are now placed to close to the input tube and its circuitry. Rotate the pcb 90 deg CW, move output transformer to the left side of the box and use some shield on 240VAC wires. Or/and split this PCB to two and place power supply PCB farther.The HV power supply inlet is too close to the input tube.

The enclosure in this case is a different one but the PCB and general arrangement of parts is similar. In the one I am testing I have deliberately routed the AC inputs to the power supply behind the smoothing caps to keep them as far away from the input tube as possible. Having said that, there is no change in the hum when moving the AC cables. I will definitely try screening the AC.

Cheers

Ian

 

[Deepdark]

Did you tried twisting teh ac cables from the power transformer to the board? I always tightly twist them with a drill.

 

[warpie]

Did you tried twisting teh ac cables from the power transformer to the board? I always tightly twist them with a drill.

You mean the secondaries?

 

[Deepdark]

I this case, the secondaries, since the primaris looks already twisted. In everyday situation, I do both. I always twist each pair of wire as tight as possible. Of course, shielding them will works, too

 

[dmp]

I think a few other posts point this out - but some caps with an outer foil will have lower hum if installed with the outer foil attached to the side closer to ground.
If you flip the cap does the hum decrease?
There is a thread on the hoffman amp forum talking about this and testing caps for the best orientation (foil to low impedance side).
Orange drop caps are shown as having this effect and they need to be measured if I recall correctly - there is no indication from the pachage or label which lead is connected to the high impedance side.

Found it:
http://el34world.com/Forum/index.php?topic=11427.msg158386#msg158386

 

[JohnRoberts]

I think a few other posts point this out - but some caps with an outer foil will have lower hum if installed with the outer foil attached to the side closer to ground.
If you flip the cap does the hum decrease?
There is a thread on the hoffman amp forum talking about this and testing caps for the best orientation (foil to low impedance side).
Orange drop caps are shown as having this effect and they need to be measured if I recall correctly - there is no indication from the pachage or label which lead is connected to the high impedance side.

Found it:
http://el34world.com/Forum/index.php?topic=11427.msg158386#msg158386

+1  this was common with old polystyrene film caps. As I recall they used to have a line marking the end connected to the outer wrap.

On pictures I find of modern orange drop caps I don't see a mark.

JR

 

[abbey road d enfer]

I think a few other posts point this out - but some caps with an outer foil will have lower hum if installed with the outer foil attached to the side closer to ground.
If you flip the cap does the hum decrease?
There is a thread on the hoffman amp forum talking about this and testing caps for the best orientation (foil to low impedance side).
Orange drop caps are shown as having this effect and they need to be measured if I recall correctly - there is no indication from the pachage or label which lead is connected to the high impedance side.

That would be significant if it was a decoupling capacitor, but here, as it's a coupling cap, both armatures are at the same potential AC-wise, except at ultra-low frequencies.

 

[JohnRoberts]

I think a few other posts point this out - but some caps with an outer foil will have lower hum if installed with the outer foil attached to the side closer to ground.
If you flip the cap does the hum decrease?
There is a thread on the hoffman amp forum talking about this and testing caps for the best orientation (foil to low impedance side).
Orange drop caps are shown as having this effect and they need to be measured if I recall correctly - there is no indication from the pachage or label which lead is connected to the high impedance side.

That would be significant if it was a decoupling capacitor, but here, as it's a coupling cap, both armatures are at the same potential AC-wise, except at ultra-low frequencies.

I don't know the circuit but the issue is termination impedance on both ends of the capacitor. Of course this is only a concern at LF where the cap itself is NOT low impedance.

I have a hard time imagining any application except for balanced gain trims, where both ends of the cap see similar termination impedance.

Of course I am just speculating since I do not know the application being discussed, but back with my polystyrenes it did make a difference.

JR

 

[abbey road d enfer]

I think a few other posts point this out - but some caps with an outer foil will have lower hum if installed with the outer foil attached to the side closer to ground.
If you flip the cap does the hum decrease?
There is a thread on the hoffman amp forum talking about this and testing caps for the best orientation (foil to low impedance side).
Orange drop caps are shown as having this effect and they need to be measured if I recall correctly - there is no indication from the pachage or label which lead is connected to the high impedance side.

That would be significant if it was a decoupling capacitor, but here, as it's a coupling cap, both armatures are at the same potential AC-wise, except at ultra-low frequencies.

I don't know the circuit but the issue is termination impedance on both ends of the capacitor. Of course this is only a concern at LF where the cap itself is NOT low impedance.

I have a hard time imagining any application except for balanced gain trims, where both ends of the cap see similar termination impedance.

Of course I am just speculating since I do not know the application being discussed, but back with my polystyrenes it did make a difference.

JR

It's the input cap, that comes from the input jack and goes to the grid. It's a 100nF. Whatever the direction, 100nF is >> than the stray capacitance, so the difference would be negligible. Indeed, with a shorted input, there would be a difference.