Looking for advice on mechanical hum

[rayl1234]

I am again going to pitch the regen solution. PS Audio has generous return policy.

 

 

[extracampine]

Ray - sorry, I had somehow missed your earlier post. I was considering a P5 at one point. I have heard however that the P5 itself has transformers and is liable to hum. Also, it is a rather expensive solution to solve a problem of hum which shouldn't be there in the first place! I would be keen to try a P5 or equivalent though, just to see if that fixed the problem.

[sandyk]

On ‎5‎/‎01‎/‎2018 at 7:15 PM, One and a half said:

I would attach a power scope to the AC and analyse the voltage waveform with intent on a harmonics up to the 31st should be enough.

 

I don't have access to gear like this that Garry mentioned, but if it happened to me I would try to get a rough idea by looking at the mains waveform on the secondary windings of a transformer that wasn't connected to anything else, with my old C.R.O.

To distort the mains waveform so badly it would normally need to come from some heavy current drawing device such as Peter mentioned  previously.

 If you are able to verify that it is from some external source, your electricity supplier should be able to investigate further.

[One and a half]

5 minutes ago, sandyk said:

 

I don't have access to gear like this that Garry mentioned, but if it happened to me I would try to get a rough idea by looking at the mains waveform on the secondary windings of a transformer that wasn't connected to anything else, with my old C.R.O.

To distort the mains waveform so badly it would normally need to come from some heavy current drawing device such as Peter mentioned  previously.

 If you are able to verify that it is from some external source, your electricity supplier should be able to investigate further.

That's also a good option and safer. The transformer doesn't need to be anything special even the secondary can be anything small like 6V 12V.

[PeterSt]

8 hours ago, extracampine said:

Peter - interesting that you still reckon that it is DC offset, despite my 2 DC blockers not working.

 

I am merely building on the fact that all your transformers mechanically hum while they will do that because of DC offset. So I was figuring that your DC blocker is not effective, somehow. Logically from this could follow that the offset is created (with)in your own audio chain, unless you had a blocker for each individual device (which doesn't seem likely to me).

 

What I wanted to add to my previous post is that you should measure the offset between your ring and an other with separate earth (pin) but since you don't own a multi meter (as I recall it) it could be a bit of a far out thing for you. Besides, if you don't have the two rings with separate earth, you won't come far with this.

 

8 hours ago, extracampine said:

I do not have any PCs that run off SMPSs.

 

So all around the house you only have tables and smartphones ? This *is* possible but not likely ...

 

8 hours ago, extracampine said:

Regarding switching things off around the house - I can open the circuit breaker for all circuits in the house except the ring circuit that the hifi is on, and disconnect everything from that circuit apart from the DAC, to make sure that it is only the DAC that is drawing power in the house - is that what you mean?

 

Yes, if you can arrange for it like this, then sure.

 

8 hours ago, extracampine said:

The only thing is, I could switch everything off in the house, but couln't it be a neighbours appliance that is affecting our mains?

 

To be honest, I never gave this a serious thought because when I run into such a problem I always find the cause and the culprit is always in my own house. Also, for over 10 years I have a DC blocker laying around (you can make them in 5 minutes with the components available) at a few feet from the stereo, but I really never used it because it would be the wrong solution. Removing the cause is the correct path to travel.

But could neighbors cause DC offset in your mains rings ? If they are connected to the same earth (pin), I guess so.

Anyway, I don't expect the source of the issue to be external to your audio gear or else that ONE DC blocker would have helped. But excluding all we might think can't be the source anyway, is always good practice (that's why shutting off all else is a good starting point).

 

Don't forget to pull your PC's monitor (brick) from its power source; Monitors are the worst for everything - not implying they easily cause DC offset, but it it easily overlooked when you switch "all off". Thus, that brick (when present) and not only the monitor.

 

Also don't underestimate what battery powered devices can do/destroy when in the audio chain, electrically connected. They run on an other potential and "skew" the balance in an unnatural way (hard to explain). I don't really see how DC offset can emerge easily from it, but currents flowing where you don't want it, for sure yes. So in this case, never think that your laptop, running from batteries, can be left alone while it is USB connected to the audio chain. Disconnect the USB cable ... (you won't have this situation, but just trying to explain what all could be wrong without real notice and this is a LOT).

 

8 hours ago, extracampine said:

a NAS which connects to the audio PC in the listening room

 

This all runs from the HDPlex, right ?

... not. So there you might have your unexpected SMPS. 

 

8 hours ago, extracampine said:

What do you mean by USB isolator?

 

Devices like Intona, Uptone ISO Regen. They galvanically isolate USB so the electrical connection from audio PC to DAC is cut (which usually is a good thing, but which may come with its own issues including masking the source of error).

[extracampine]

Well, I just opened the circuit breakers for all circuits in the house except the one the DAC is on, and turned off all items except for the DAC (including the NAS, good point Peter). And no change to the hum. So perhaps that rules out something else in the house generating noise on the line.

 

I do own a multimeter - or if I didn't, they're cheap enough to buy. I am going to test the mains with it to see what the AC voltage is doing; hopefully report back with this in the next few days.

 

As far as I know there is only one earth here Peter.

 

I am also going to take the DAC to my work to see if it hums there. And if it does, I will take it maybe to an audio store that is a bit further away to see if it hums there.

[mansr]

6 minutes ago, extracampine said:

I do own a multimeter - or if I didn't, they're cheap enough to buy. I am going to test the mains with it to see what the AC voltage is doing; hopefully report back with this in the next few days.

Some multimeters can measure DC offset of an AC voltage. The procedure varies, so check the manual for yours.

[Don Hills]

As I said over on HA, you could borrow or rent an isolating transformer. That will prove whether or not your problem is due to DC offset.

[extracampine]

Could you post a link to the kind of transformer you mean?

[Don Hills]

https://www.google.com/search?&q=isolation+transformer

 

Plug the input into a wall socket. Plug a piece of humming equipment onto the output socket. If it stops humming, the cause of the hum is DC offset.

[STC]

I still think the best way is to measure the DC offset. It should be less than 10mV.

[PeterSt]

25 minutes ago, STC said:

I still think the best way is to measure the DC offset. It should be less than 10mV.

 

Haha, and with what multimeter are you going to measure *that* ?

So you must be talking about DC offset on the DAC output (or preamp output). That should be that kind of figure (most take 20mV for maximum) and this can easily be measured on the max 5V or so this output comprises ... but this is about DC offset on the mains voltage (~240V or ~130V) and the offending offset we talk about should be 30V or more (the 30V is a made up figure and it will depend on the transformer).

 

16 hours ago, extracampine said:

I just opened the circuit breakers for all circuits in the house except the one the DAC is on, and turned off all items except for the DAC (including the NAS, good point Peter). And no change to the hum.

 

And you did not try the power amp on its own, or the HDPlex on its own, etc. ?

Anyway, I recall that you already tried the DAC elsewhere (repair shop ?) and that it was silent over there. So it wouldn't be that you suffer from a general problem in your neighborhood, right ? Not that you didn't suggest this already, but if so it would be a very bad thing.

 

Of course your DAC will turn out to be silent at work. Of course your next step is visiting a neighbor with it ...

(but don't forget to test each of your audio devices which hum, individually)

 

Strange ...

 

[extracampine]

Don - thanks for the specific suggestion  Yes I will look into that. Won't such a transformer dampen dynamics and limit peak draw?

 

Peter - you're saying that measuring DC offset on the mains using a multimeter is not a viable option? Perhaps a more powerful oscilloscope is required for such measurements? I did not try the power amp or HDPLEX on their own, no - I'm mainly concerned about the DAC as that has the most prominent hum. And no, I have not yet tried the DAC elsewhere - what I did do was send it back to the manufacturers (in Poland) where there was no hum found.

 

What I will do in the next few days is try the DAC elsewhere in the UK, first at my work and if it still hums there (it is only 5 minutes drive away) then I will try it further away (e.g. a music shop 30 minutes drive away). If it does turn out to be silent at work I will be pleased as it will be a clue as to the problem!

[One and a half]

Tip 2: Measuring an AC Signal with a DC Offset

An ac-coupled true RMS measurement is desirable in situations where you are measuring small ac signals in the presence of large dc offsets. For example, this situation is common when measuring ac ripple present on dc power supplies. There are situations, however, where you might want to know the ac+dc true RMS value. You can determine this value through a manual calculation, shown below, add the square of the dc measurement and the square of ac measurement and take the square root. You should perform the dc measurement using at least 10 power line cycles of integration to reject the ac signal.

True RMS ac + dc =SQRT( ac^2 + dc^2)

Even if your multimeter contains a built-in function to measure the True RMS ac+dc value. You may want to manually determine the value using a dc measurement, an ac only measurement and apply the formula. The manual method will yield a better result, particularly when the dc and ac values are of different magnitudes.

 

DC Offset on AC is a crossover distortion when the sine wave passes through zero. To measure use a scope. What to do? Lets get the DAC to the workplace and measure some volts to start with. Jan 8th deadline!

 

 

[STC]

1 hour ago, PeterSt said:

DAC output (or preamp output). That should be that kind of figure (most take 20mV for maximum) a

 

Yes, my mistake.

 

I was referring to DC offset of the amplifier which actually made the amplifier hum. 

[PeterSt]

31 minutes ago, extracampine said:

Peter - you're saying that measuring DC offset on the mains using a multimeter is not a viable option?

 

No, that is not what I'm saying. I only reacted to the idea of someone that 10mV of DC offset is max - but this counts for DAC or preamp (voltage) output - not for mains voltage.

 

This is how DC offset looks like; in this case the offset is 3 Volt.

The offset is always relative to a ground reference, usually (but not always) denoted as 0 Volt.

We talk about AC here (that's why we see a wave form in the first place, otherwise it would be a straight horizontal line).

Because it is AC, we can not measure DC as such. However, we are able to measure the difference in AC voltage. So like you see here, the difference of 3V emerges everywhere. And thus we can state that the 5V AC Peak-positive (not from peak to peak) we would measure in this case, will show 8V peak-positive in the offset situation.

But although it looks easy, it is not easy at all to measure this. This is because all is relative to the ground (gnd) reference point and when your whole mains is offset 3V, no device would care about it. It just doesn't know that the 0V point (which is supposed to be the gnd point) is not relative to 0V. In reality : what the earth depicts. So suppose that your earth pin is 1000ft away, and at the point it sticks in the ground there's 0V indeed. Now, when it arrives at your house it has become 3V for whatever reason. But, it still is the ground reference, and with no other earth pin right next to you (see more below !), you are not able to measure this 3V by any means. Your devices won't "sense" it either. It is just the ground reference point ...

 

If we set up 5 Volt with your earth pin as the reference, it just will show as the blue line on an oscilloscope. Why is that ? because the oscilloscope is connected to the same earth (pin) reference, and thus sees the 0V point (which, remember, actually is 3V) just the same as the device which set up the 5V. There's just no DC offset to be seen anywhere ... It's really not there ...

 

The problem emerges when a second ground reference comes into play. For example that second earth pin right next to you, which carries a 0V for real. So now you have two references - one of 3V and one of 0V. If you now plug your oscilloscope in the 0V reference and measure the device with the 5V, you will see the 3V of offset.

 

Watch out : although the situation above can be incurred for with two mains rings, the one with a heavy load (shifting the 0V reference upwards) and the other with a light load, this won't be your situation (hey, if your situation is about DC offset at all). But look at the green line in above plot BUT keep on seeing the 0V as its reference. You now have 8V on the positive side of 0V and 2V on the negative side of 0V. And THAT is killing. Now try to envision that you have a device which is capable of "pulling" the positive voltage (let's say that it consumes more of negative than of positive voltage) and there you have the other reason for DC offset.

Thus, my example of the two ground references can exist too but then you first must have two mains rings (and careful, because an isolated power transformer already forms this own 0V reference point and when you plug in a monitor from the normal mains ring in the PC which is hooked up to the isolation transformer, you're gone (the two references will fight each other and probably everywhere you have the misformed wave form)).

 

The subject is way large and very complicated already without humming transformers  and usually is not seen / noticed.

 

So can you measure DC Offset of your mains ring with a multi meter ?

Yes you can. You need a clean second earth reference not connected to the other, and measure DC between the two earth's (the earth pins in the outlets). If you know your mains makes use of an earth pin somewhere, then you may be able to use a pipe of the water supply in the house. Of course that should not connect to the earth pin, which is common to do).

 

Can you, in your situation, measure the DC offset your transformers may suffer from with one mains ring and one earth pin ?

I'd say Yes. The oscilloscope (you can buy second hand for less than GBP 100) should show the offset if it's there. So I'd say that (despite my long story) if the transformers suffer fro DC offset which really is there (misformed wave form), the oscilloscope will show the same (it will show the green line of my example on average above the 0V reference point of the scope).

 

[mansr]

DC offset has nothing to do with earth. An AC voltage between two terminals has a DC offset if the arithmetic mean over one period is non-zero. It can be measured with a multimeter or an oscilloscope.

 

With a multimeter, set it to DC voltage. Disable auto-ranging, if available, and select a range (typically 500 V) suitable for the AC voltage. Attach the probes to the live and neutral terminals of a power socket. The meter should now display the DC offset. Some multimeters, e.g. Fluke 287/289, can display both AC and DC components simultaneously.

 

Using a scope requires a little more care. First, make sure the scope input supports mains level voltages. Mine says 300 Vrms, so it's fine. Next, you must verify which terminal is the neutral, or you might get a short circuit as the "ground" on the scope input is typically connected to safety ground. A multimeter can be used for this. Connect the scope to the power socket observing the correct polarity. Adjust the time base and vertical scale such that a few periods of are visible and use the scope's measurement features.

 

Always use the utmost caution when working on mains power. A misplaced finger can kill you. Use only equipment rated CAT II or higher. If you feel unsure or uncomfortable about anything, do not proceed.

[Summit]

DC offset is the result of an electric asymmetry that do not equally load the AC positive and negative sine wave crating a DC offset. Big electrical motors and switching PSU are often the worst offenders.

 

http://www.jenving.com/products/view/dc-blocker-md01-16-eu-3024000378

[PeterSt]

4 hours ago, mansr said:

DC offset has nothing to do with earth.

 

Except that earth is the only absolute reference (and then still FWIW).

[mansr]

1 hour ago, PeterSt said:

Except that earth is the only absolute reference (and then still FWIW).

We're talking about the voltage between two wires. Their voltage relative to earth (or the moon, for that matter) is irrelevant to the functioning of a transformer connected to them.

[rayl1234]

16 minutes ago, mansr said:

We're talking about the voltage between two wires. Their voltage relative to earth (or the moon, for that matter) is irrelevant to the functioning of a transformer connected to them.

 

What PeterSt was trying to say is -- there is no such concept as an absolute voltage, only a relative voltage (as voltage is a potential, like height....).  i.e. 0 to 240 vs 5000 to 5240 vs -120 to +120....  those absolute numbers only have meaning if you define what 0 means, e.g. relative to earth.... or relative to the average.... And from there, DC offset only has meaning relative to an external reference (DC offset relative to XXXX).

 

 

[mansr]

8 minutes ago, rayl1234 said:

What PeterSt was trying to say is -- there is no such concept as an absolute voltage, only a relative voltage (as voltage is a potential, like height....).  i.e. 0 to 240 vs 5000 to 5240 vs -120 to +120....  those absolute numbers only have meaning if you define what 0 means, e.g. relative to earth.... or relative to the average.... And from there, DC offset only has meaning relative to an external reference (DC offset relative to XXXX).

We have two wires. The relevant voltage is that of one relative to the other. A DC offset exists if the average voltage between the wires is non-zero. An offset of a million volts relative to ground doesn't matter to something, such as a transformer, connected only to these two wires.

[rayl1234]

1 minute ago, mansr said:

We have two wires. The relevant voltage is that of one relative to the other. A DC offset exists if the average voltage between the wires is non-zero. An offset of a million volts relative to ground doesn't matter to something, such as a transformer, connected only to these two wires.

 

HUH?

 

The "average voltage between the wires" is never going to be zero.... Rather than repeat the equations here, here is a quick tutorial:

 

http://www.electronics-tutorials.ws/accircuits/average-voltage.html

[mansr]

34 minutes ago, rayl1234 said:

HUH?

 

The "average voltage between the wires" is never going to be zero.... Rather than repeat the equations here, here is a quick tutorial:

 

http://www.electronics-tutorials.ws/accircuits/average-voltage.html

Did you even read that page?

 

"For a periodic waveform, the area above the horizontal axis is positive while the area below the horizontal axis is negative. The result is that the average or mean value of a symmetrical alternating quantity is zero because the area above the horizontal axis (the positive half cycle) is the same as the area below the axis (the negative half cycle) and cancel each other out in the sum of the two areas as a negative cancels a positive producing zero average voltage."

[rayl1234]

20 minutes ago, mansr said:

Did you even read that page?

 

"For a periodic waveform, the area above the horizontal axis is positive while the area below the horizontal axis is negative. The result is that the average or mean value of a symmetrical alternating quantity is zero because the area above the horizontal axis (the positive half cycle) is the same as the area below the axis (the negative half cycle) and cancel each other out in the sum of the two areas as a negative cancels a positive producing zero average voltage."

 

Yes, but that is not the definition, which is stated further down:

 

"The average value of a whole sinusoidal waveform over one complete cycle is zero as the two halves cancel each other out, so the average value is taken over half a cycle"

 

and illustrated:

 

 

 

PS: To be clear, I did not want to derail the thread or challenge whether DC offset is or is not the issue.  Only to say that a DMM's reading of average voltage, if it offers such, is not the right number. Need a meter that reads out DC offset, which Fluke true RMS DMMs can do.....