Well this is frustrating ...

[Deke609]

If the attenuator is all the way down, is the hum nearly inaudible? 

Yes, at about -27dB it is barely audible. If I wasn't looking for it, I wouldn't notice it.

Is everything plugged into the same outlet?

Yes. And I've now tried 4 different outlets. 1 off the main panel, 3 off the sub-panel, including the 20A dedicated. All give me hum.

When you figure out what's actually wrong, it's likely to be something incredibly simple. 

That would be nice. I need to take a fresh look at the problem.

Something I've had happen that could cause this would be one of the two connetions on an RCA jack not being connected.  Or, for instance, if somehow you swapped ground and signal hot, that could make a weird problem like this.

I will look. But for that to be the case, I'd have had to make the same error at least twice and at least 1 year apart. B/c I got the same hum with BeePre (built in the last year and tinkered with) and Kaiju (built recently and tickered with), and BeePre and SII-45 (haven't done anything to it in a while - no snubbers, no CLC filters, etc.). And I get it with the Nickel Wonder (built from schematic a while ago; made no mods) w/ Kaiju, SII-45 and a retail-bought SS HP amp.  In mid Sept I had no hum (other than reg dropout in the Beepre) and had never experienced hum.  So I think something has changed ... I just have no idea what.

I'm stumped. THe only thing I can think of is the water leak. But I don't know how that would lead to what I assume is a ground loop somewhere. Or is not a ground loop?

Please keep the ideas coming!!!

thanks, Derek

 

[Jamier]

Derek, have you tried different interconnects/ cables, including power cables?

Jamie

 

[Deke609]

Derek, have you tried different interconnects/ cables, including power cables?

Jamie

Yup - four different interconnects, including 2 store bought to rule out me having made the same mistake in building the cables. And different power cords - diy and stock/commercial ones. No difference in hum.

 

[Deke609]

Does anyone see any clues in the jagged 60Hz waveform that I posted on the first page of this thread?

many thanks, Derek

 

[Doc B]

It's not just 60. Looks like 120 and maybe 180 on there too. Just hook a bare wire to your scope input hot and see if it shows the same kind of waveform. Since you seem to hear hum with every piece of gear, that will tell you if every piece of gear is picking up something in the environment.

Also, I don't see anywhere that you have shorted the inputs and looked at the waveform. Turning down the level pot is almost the same thing, but just almost. That will help to determine if the root problem is within the DUT.

 

[Deke609]

... hook a bare wire to your scope input hot and see if it shows the same kind of waveform. Since you seem to hear hum with every piece of gear, that will tell you if every piece of gear is picking up something in the environment.

Many thanks Doc. Will do.

Also, I don't see anywhere that you have shorted the inputs and looked at the waveform. Turning down the level pot is almost the same thing, but just almost. That will help to determine if the root problem is within the DUT.

So short the preamp inputs and connect to amp and test? Or short inputs of unconnected preamp/amp and test?  Or all of the foregoing?

Many thanks, Derek

 

[Doc B]

Start by looking at one piece of gear at a time in order to narrow down where the problem lies. Disconnect the preamp from the amp, short the preamp inputs and look at the output of the preamp. Then do the same with the amp, short the inputs of the amp and look at the amp outputs.

Note that this applies only to Bottlehead products. I don't take any responsibility for the stability or functionality of anyone else's amps with the inputs shorted

 

[Deke609]

Ah hah! Should have tried this earlier: DAC straight to the SII-45, no preamp. NO HUM! ZIP, ZILCH, ZERO. PERFECT SILENCE!  (And as an intesrting side note, it sounds much better by itself than with the Nickel Wonder in front of it ... which, come to think of it, I think is the reason why I put it in a corner to collect dust in the first place).

So there's something about the interaction of the preamps and the amps. Maybe  I reallly did make the same mistake multiple times. So I think I need to look for possible sources of a difference in ground potentials between preamps and amps. I wonder if the "balanced outputs" mod (+ve and -ve of both speaker out referenced to ground via resistor) might do this? Or interact with the diode chassis ground elevation mod?  I am going to enjoy listening to the SII-45 by itself for a bit and look over schematics and photos of my builds to see if I can spot anything.

Feeling more relaxed now and confident that the riddle will eventually be solved.

Start by looking at one piece of gear at a time in order to narrow down where the problem lies. Disconnect the preamp from the amp, short the preamp inputs and look at the output of the preamp. Then do the same with the amp, short the inputs of the amp and look at the amp outputs.

Many thanks Doc. I will try this tomorrow morning. I'm pretty tired today - at 2 am yesterday I scrubbing/filing heavy gauge ground bus wire in an effort to improve earth ground

Note that this applies only to Bottlehead products. I don't take any responsibility for the stability or functionality of anyone else's amps with the inputs shorted

Hah! I don't care about the other ones. And I'm neither litigious nor the type to blame people for trying to help me (for which help I am thankful and appreciative)!   

I hereby absolve and agree to hold harmless all who post or have posted advice, opinions, suggestions etc., of whatsoever kind, on this thread or other any other on this forum; and for greater clarity, declare that I am solely responsible in law, equity or otherwise for any actions or inactions I have taken or may in the future take, whether in reliance up on or influenced by such advice, opinions or suggestions. Seriously. Not joking.

cheers and thanks, Derek

 

[Deke609]

That's a new piece of information.  If the attenuator is all the way down, is the hum nearly inaudible? 

Yes, at about -27dB it is barely audible. If I wasn't looking for it, I wouldn't notice it.

@PB: I should clarify that only the attenuator on the SII has this effect. Not the volume pots on the BeePre (pre-dismantling) or the Nickel Wonder. Changing the coarse attenuator on the BP had little effect on the hum: slightly changed the tone, but not the volume. With the Bp attenuators set to minimum volume, I got pretty much the same hum loudness. Same thing with the Nickel Wonder: same hum with volume all the way down as nearly fully open - except that turning it all the way up seemed to shift the hum to one side and change the tone a bit - but I figured this was just weird bahavior of a $10 pot turned too far (literally forcing it to point it wouldn't turn any more). 

And I need to correct a previous misstatement: the hum becomes nearly inaudible at -18dB, not -27dB.

Does the fact that the SII attenuators, and only the SII attenuators, affect the volume of the hum point to anything?

many thanks, Derek

 

[Deke609]

Last observation for the day: no hum with the Quickie as preamp in front of the SII-45. I assume the same will hold true with the Kaiju. I will try that tomorrow.

Off-topic tangent: the Quickie w/ PJCCS is my son's and I'd never listened to it except connected to my son's Quicksand. It sounds insanely good in front of the SII-45. It beats the pants off the Nickel Wonder - which cost me about 2 X as much in parts (the tiny Hammond nickel transformers alone were $100 or more). Wow! It's no BeePre, but I could understand someone hearing both and deciding they could live with Quickie. Crazy good. And really fun to listen to.

cheers, Derek

 

[Paul Joppa]

It sounds to me like the problem is capacitive coupling from a power transformer winding to the signal ground. The Quickie has no power transformer.

This is described in one of the Jensen Transformer white papers. These papers of course reflect the reasons you should use transformers, but it is perfectly true that isolation transformers will break such a current better than pretty much anything else.

It's too late (and I had too nice a wine with Friday dinner) to follow up in much detail, but a ground lift in one device and/or the other should make a difference, if my analysis is correct.

Usually we look at a preamp / amp, and because the power amp transformer is larger, we expect it to be the "source," but I imagine both contribute in the real world. That may be why reversing the phase with a cheater sometimes works and sometimes doesn't.

 

[Deke609]

Many thanks PJ. I will track down the Jensen white paper and see if I can make sense of it tonight.

cheers and thanks, Derek

 

[Deke609]

Here is Bill Whitlock's (Jensen) White Paper that PJ referenced. Since hum is a common problem this may be of help to others now and in the future.  It is pretty accessible - no calculus, just a bit of simple algebra based on Ohm's Law and the formulas for voltage dividers, parallel resistors and capacitive reactance. And it's a really cool example of how you can use these simple formulas/concepts to work through a pretty complex problem.

Many thanks PJ for flagging it.  I am still digesting it (as fully as I can) and will probably have a couple of follow-up questions (a) to make sure I am at least getting the basic idea, and (b) about the practical solution: which looks to be a diy interconnect box with a wideband, low distortion 1:1 isolation transformer.

cheers, Derek

 

[Paul Joppa]

For what it's worth, you don't have to DIY; Jensen makes isolator boxes of many kinds; here's an example:

https://www.jensen-transformers.com/product/ci-2rr/

That's a little pricey, and some listeners don't like the resulting sound quality. I see it as a convenient diagnostic tool. If you like the sound, it's also a solution, but as I said a ground lift (back to back diodes shunted by a 10-ohm resistor and a small capacitor to bypass it at radio frequencies)  in the preamp and/or in the power amp will often do the job. That's because the source impedance of the capacitive coupling is very high, so the 10 ohm resistor almost completely shorts that current. Merlin Blencowe's article on grounding has a treatment of the ground lift:

http://www.valvewizard.co.uk/Grounding.pdf

 

[Deke609]

Excellent! Thanks PJ. Yes, the Jensen "IsoMax" stuff is pricey - which is why I thought I'd diy, plus it's fun to make one's own stuff.

I will try the diodes/resistor/cap approach. If it completely kills all audible hum I will just go with that. Otherwise, I will look into getting a good 1:1 transformer.

cheers and thanks, Derek

 

[Jamier]

So, this is a slightly OT, but I’m a little confused about PJ’s comment that earth grounding is not  important when it comes to noise reduction. I completely believe him, he has more experience than, well, maybe anyone, but a recent build that PB helped me on, led me to believe that earthing is important to noise reduction so I’m trying to understand. PB advised me to ground the signal input ground at the point where the input entered the audio circuit ( in this case it is a PCB). I ran a wire from the signal input ground connection to a star ground that I located next to the IEC. The star ground connects directly to the ground pin on the IEC. When I first powered up the amp(s), they are monos, this wire was not connected and I had some audible hum. When I connected this wire to the star ground that hum ceased. The amps are push/pull K-501s. Also, connected to the star ground are the signal input cable shield ( from the load end) and two wires from aluminum plates that the binding posts and input RCAs (these are on opposite sides of the amps from each other) are mounted in, they have insulated shoulder washers. If the earthing isn’t the reason for the hum reduction, what is? Can you help me understand? I haven’t used the amps in a few weeks so I will experiment with this connection and see if the hum repeats when disconnected but I remember that this connection reduced the hum considerably, I’m pretty sure.

Jamie

 

[Paul Joppa]

Jamie - the noise reduction was because the signal ground connected to the chassis, not because the chassis happened to be connected to earth ground for safety reasons.

 

[Jamier]

PJ,

    Thank you for correcting my misunderstanding. If I had made the chassis from something metallic that might have occurred to me, but I used wood and created the star ground from a screw terminal block, which became, I guess, the conductive part of the chassis. At least I had the good sense to attach everything metal to it. Well, at any rate, I’m learning, albeit slowly. Apparently, I ‘m not afraid to ask stupid questions, so for those of you who are, stay tuned and I will probably ask every one of them and probably some twice.

Thanks again, Jamie

 

[Deke609]

I scoped the voltage of mains neutral to mains ground a few minutes ago.  It looks a lot like what I saw coming out of the SII-45 when it was fed by the Nickel Wonder.

Top attachment is mains neutral:ground. Bottom is the SII-45 output I posted before.

Any thoughts?

I tried adding one diodes/resistor/cap combo to the SII-45 and it did nothing. I will try adding one to the nickel wonder tomorrow (ran out of time today)

cheers and many thanks, Derek

[Edit: COuple of additional points. (1) If you look toward the bottom of the images, there are some basic stats about the waveform. (2) The hum has gotten worse over the last month. It started as a barely audible hum coming out of the Kaiju a month ago. It is much louder now.]

 

[Deke609]

No change with with the diodes/resistor/cap ground elevators installed in both SII-45 and the Nickel Wonder. 

Don't know how to make sense of this.  No hum with DAC directly connected to the SII-45. And no hum with DAC -> Quickie -> SII-45.  But connect the Nickel Wonder (which when originally built caused no hum) to the SII-45, or a retail purchased and unmodified SS headamp, and I get hum.  Had the same issue (but not as loud) when the Beepre was still put together (but didn't test it with other amps - just the Kaiju -- b/c disconnecting the BP made the hum go away and so I assumed that it was the source of the problem).  So something has changed in the environment - maybe corrosion in a receptacle or at the mains sub-panel caused by the August upstairs water leak; maybe the City's new line on the street; or a host of any other things -- Arghh!)

In Bill Whitlock's White Paper, he suggests that the power transformer in each connected amp produces a parasitic capacitive coupling with the amps's chassis ground, and that the interconnects then connect the two chassis, and that this opens the door to a ground loop that becomes part of the signal going to the second amp (b/c somewhere in the amps the signal wire is referenced to ground). 

If the above interpretation of the white paper is correct, then doesn't it follow that elevating the chassis ground from earth ground in both amps leaves what I'll call the "Whitlock problem" untouched - can't a loop still circulate between the two chassis by way of the interconnects?  If so, then maybe I need to try isolation transformers.  The specs of the Jensen CI-2RR look good on paper: frequency range of <10Hz to >40kHz and low distortion - but is $200 USD. [Edit: and I'd really like to diy them]. Does anyone know of equal or better isolation transformers (wideband, low distortion) that would be suited to the task? I came across a couple pcb-mount transformers made a company called Audac that look even better to me on paper (models TR106 and TR266: http://audac.eu/products/c/transformers/line-transformers, but I can't find any reviews of them and haven't figured out where/how I could buy them.

cheers and many thanks, Derek