• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Do you need AC Noise Filtering For Your Audio Devices?

Short video on the fact that most audio gear already has noise and surge protection in them obviating the need to have them externally.


Amazon basics amplifier teardown: https://www.audiosciencereview.com/...azon-basics-80-watt-amplifier-teardown.20945/

Hello Amir, I just remembered that in 1984 the German Test Institute tested amplifiers with a refrigerator. It was switched on and off to simulate spikes in the power grid. If I remember correctly, only two out of a dozen amplifiers did not give a cracking sound. That would be a nice additional test. In certain situations, EMI line filters make perfect sense.
 
Hello Amir, I just remembered that in 1984 the German Test Institute tested amplifiers with a refrigerator. It was switched on and off to simulate spikes in the power grid. If I remember correctly, only two out of a dozen amplifiers did not give a cracking sound. That would be a nice additional test. In certain situations, EMI line filters make perfect sense.

Theory still predicts the bulk of the filtering against such noise lies in the main power supply circuit. Actual experience, the EMI filters I have can't even totally block out the spike from a lamp or an SMPS switching on, and I'm not even sure if the reduced level is due to luck i.e. which part of the 50Hz cycle I switched it on during testing.

Yes, a pebble certainly reduces the flowrate of a river. But how relevant is that when dealing with rivers and dams.
 
Theory still predicts the bulk of the filtering against such noise lies in the main power supply circuit. Actual experience, the EMI filters I have can't even totally block out the spike from a lamp or an SMPS switching on, and I'm not even sure if the reduced level is due to luck i.e. which part of the 50Hz cycle I switched it on during testing.

Yes, a pebble certainly reduces the flowrate of a river. But how relevant is that when dealing with rivers and dams.

In the case that the noise origin is genuinely conducted in on power lines, a series filter could help (if you have a real problem to begin with). But in my experience with RF immunity, you will often just squeeze a balloon (the same problem still exists, entering equipment elsewhere) unless you treat all interfaces at the same time.

The field strengths generated by things like broadband esd discharges and narrowband carriers in powerful walk-in talkies are unlikely to be suppressed by anything except * hardened equipment. Hearing a tick or a noise coincident with such events should not worry anyone about their sq.

* Protection can even extend to nuclear EMP, if you are willing to shroud your equipment with thick honeycomb copper shielding.
 
Wow, great thread. So refreshing for us non-engineers who have dealt with non-measurable claims of improvements in cables, power conditioners, etc for so many years to see clear explanations and measurements. And timely, too. I am in the middle of planning a remodel and am planning on installing a dedicated 20 amp circuit to my home theater. One thing I've been struggling with is how to do power distribution from the 20 amp socket to all of my home theater devices. I want something high quality, non-restricting (in terms of current flow), plenty of plugs, etc, but most of the home theater units make you pay for their power conditioning features, which you have demonstrated is unnecessary. After some research I'm leaning towards something like the Tripp Lite 20 amp PDU.

https://www.amazon.com/gp/product/B00VET26IU/ref=ox_sc_act_title_2?smid=ATVPDKIKX0DER&psc=1

I realize that it is intended for data centers, but it appears to be the best option for quality distribution of power without paying many many times more the price for power conditioner / surge protection features. I like Amir's idea of using a whole home surge protector, so I'm not looking for that in a PDU either. Just robust, quality power distribution. Wondering if any others have considered or used something like this? Or have come across other products that might be better? Thanks!
 
I don't get the attention on "non-restricting to current flow" when, you realize there is no current flowing thru the mains cable >50% of the time? Yet you don't even realize or hear that. Something must be doing the magic. If that magic can allow the equipment to work with 5ms of no-current, surely it can handle shorter durations.

https://electronics.stackexchange.com/questions/229420/what-is-a-capacitor-ripple-current

Red line is capacitor current

wScw6.png
 
In general, current ratings are the current that the device is able to deliver without overheating or throwing a circuit breaker or wasting a fuse. If one of these things happens, then current will be limited as a consequence, but otherwise the idea of current limiting is somewhat curious. The only way in general that current will be limited by any device is if the device is overloaded with current and one of these catastrophic things happens, or else if the voltage output by the device is reduced, such that current is reduced in accordance with Ohm's law and in accordance with the impedance of the device to which current is being supplied. In the particular case of current supplied by an amplifier, the amplifier may employ circuitry that will automatically reduce the gain in the event that current is excessive, and possibly in the event that voltage peaks come near to the clipping threshold. When/if an amplifier automatically reduces gain in response to one of these circumstances, there will be a reduction in current commensurate with the reduction in output voltage. Simple devices that deliver AC power, i.e., power cords and similar, do not have gain and do not generally have a smart method for reducing the voltage they output. As such, if simple devices such as power cords or power strips limit or reduce current, it is either because one of the aforementioned catastrophic events has occurred, or else because the voltage they output is reduced in accordance with the voltage dividing effect. The voltage dividing effect occurs when/if the internal impedance of the device supplying current is great enough to be significant in comparison with the impedance of the device to which current is being supplied, such that a significant portion of the voltage nominally output by the device supplying current is taken up within that device.

A perfect illustration of this is what happens with batteries generally, as they age. As they age, the internal resistance increases, such that a greater portion of the voltage nominally generated within the battery is taken up within the battery. The increase in total series impedance means that current is reduced. The reduction in current requires, again in accordance with Ohm's law, that there be a reduction in voltage seen by the device being powered by the battery, i.e., a reduction in voltage across the battery terminals. From the standpoint of the voltage dividing effect, the increase in the battery's internal resistance implies that the impedance of the device powered by the battery decreases proportionally, i.e., as a proportion of the total series impedance. This decrease in the powered device's proportional share of the total series impedance implies a corresponding decrease in its proportional share of the nominal voltage generated within the battery. Thus, the battery supplies less current than it did when it was young and vigorous, but this reduction in current occurs in perfect conjunction with a reduction in voltage output by the battery, and the reduction in current could not possibly occur except in conjunction with the reduction in output voltage.

(I should probably note, at the risk of complicating this more than I already have, that if you measure the voltage supplied by a battery when the circuit is open, that no matter how great the internal resistance of the battery, it will be very small in comparison to the near-infinite impedance across the terminals, which means that essentially all of the battery's internally generated voltage will appear across the terminals, giving the false impression of a good battery even for a battery that has already kicked the bucket in practical terms. To check a battery, you can check either current or voltage, but either way, you have to do it while the battery is being asked to supply as much current as it needs to be able to supply when in use. Since voltage readings are usually simpler than current readings with an ammeter - which generally involves a shunt - the easier way to check a battery is by checking to see whether the voltage it supplies to the normal load for the battery is the same or very nearly the same as the voltage supplied to that same normal load by a new or fully charged battery. Of course this has very little relevance to what I wanted to say about the concept of current limiting, but after I had said what I wanted to say, I decided that I had better say this as well, to tie up the loose end.)
 
In the case that the noise origin is genuinely conducted in on power lines, a series filter could help (if you have a real problem to begin with). But in my experience with RF immunity, you will often just squeeze a balloon (the same problem still exists, entering equipment elsewhere) unless you treat all interfaces at the same time.


Yup, the assumption is also that the noise from whatever switched on is purely transmitted through the line/neutral connection which is not necessarily the case. The high current on L/N is going to inject noise on ground wires too. Of course, depending on the equipment, the spike can be many 10's of amps and 100's of usec in duration, so don't be surprised that the instantaneous drop on the AC line depending on location is rather large.
 
I managed a alternative energy and battery shop for a short while and the life expectancy of a SLA (sealed lead acid) battery was not that great. They have lotsa issues. If memory serves me correct it was very thin sulphate deposits on the plates and other recharging issues stuff. I always bought fresh for my own needs like my mountain bike headlamp system that used a halogen bulb system. :D

Depending on the construction, and quality of the maintenance charge, a good SLA can last 10+ years on standby. The batteries in cheap UPS are not this quality and the UPS does not treat them well. SLA can last 7+ years in cars, though less in warm climates.
 
It looks like I've been wasting my time: https://www.audiosciencereview.com/...ne-noise-emi-filtering.976/page-2#post-554128

Looking at the Class D amp modules in my speakers and subwoofer, it looks to me like there's a mains filter right on the board:

Ah well, it was fun building the DIY filters and the sale of the ISOL-8 filter, and the collection of Russ Andrews power cables I had, paid for the parts with a significant amount of change left over. :)

You missed a few parts :) Technically what you circled is the common mode filter, though the leakage inductance of the common mode inductors coupled with the X-capacitors will help out on the differential mode noise as well. I added a box around the specific differential mode filter. If I am not mistaken, and I certainly could be, but the capacitor, diode, resistor looks like a non-sacrificial surge circuit.

1622656744935.png
 
Whilst a lot of hifi equipment contains AC filtering, most equipment with a toroidal transformer doesn't cope well with a DC offset (which is often caused by common white goods appliances). This often results in the transformer buzzing and the only way to get rid of the offending appliance/s or purchase a DC filter to remove or at least reduce the DC offset, which in turn removes or reduces the buzzing.

I've had this problem (over 80dbA @1m) and have a mains DC filter connected to my voltage input of my Arcam AV amplifier, which significantly reduces the toroidal transformer buzz.
 
Last edited:
In a residential environment, there are few appliances capable of causing a DC offset that operate 24 by 7. So most DC offset problems should come and go, yet we don't read about audiophiles noticing this on/off situation.
 
I personally have never experienced this issue in my house, but know it exists. I purposely made a circuit once to create it once, just some diodes and a big resistive load close to the transformer load.
 
In a residential environment, there are few appliances capable of causing a DC offset that operate 24 by 7. So most DC offset problems should come and go, yet we don't read about audiophiles noticing this on/off situation.
Actually I have - one of my O110s (just one) seems to be quite sensitive to this and will occasionally emit quite audible transformer hum. Then, at some point, it'll suddenly stop again. I have yet to identify the cause but it only seems to happen on workdays, so I presume it's something of an industrial nature (which is odd, given that I live in a residential area). Usually this problem is most prominently seen when 500+ VA toroids are involved.

Hello Amir, I just remembered that in 1984 the German Test Institute tested amplifiers with a refrigerator. It was switched on and off to simulate spikes in the power grid. If I remember correctly, only two out of a dozen amplifiers did not give a cracking sound.
Hmm, my speaker setup will occasionally emit one of these, and it had never occurred to me to try and correlate it with my fridge. Silly, given that the fridge giving trouble tends to identicate ground loop issues quite effectively.
 
What is the oldest appliance in your house. Look for that as the culprit. Can be something unexpected like an old diode based step light dimmer.
 
Actually I have - one of my O110s (just one) seems to be quite sensitive to this and will occasionally emit quite audible transformer hum. Then, at some point, it'll suddenly stop again. I have yet to identify the cause but it only seems to happen on workdays, so I presume it's something of an industrial nature (which is odd, given that I live in a residential area). Usually this problem is most prominently seen when 500+ VA toroids are involved.....................................................
That's exactly how I would expect a DC offset problem to occur. It takes some using a significant amount of power to cause a DC offset. Not many items in a residential area do that for long periods of time. For the most part, these items have to be powered from the same utility power transformer.
 
I found out a good reliable way to introduce noise is using power line adapters (Devolo) and download a big file. A filtered extension was unable to remove (all) the noise I heard via my headphone DAC/AMP. The noise was still there but softer, even when the adapter was in a socket 5m across the room.
Maybe unfair to blame the dac/amp, since these adapter purposely and forcefully create this modulated and pulsecoded (network packet bursts) 'noise' to communicate.

They didn't work reliably and this issue was the last straw causing me to return them.
I would not recommend them, other than for the purpose of testing noise filters :p
 
I found the buzzing from the toroidal transformer on my Arcam AVR-850 so annoying and even spoke to Arcam about it.

I measured the DC offset to be just over 1.5v, so purchased an ATL audio DC filter (diodes up to 2v) and this has worked exceptionally well and has removed the loud buzzing.

This is really worth knowing if anybody has a similar problem, as I could easily hear the buzzing from over 17ft away and through a BDI based cabinet.
 
Back
Top Bottom