Ferrites/cable advice

I wouldn't bother with sticking ferrite cores on the mains cables. Just make sure the mains cables hang freely and are kept away from any signal cables. With the system switched off, just plug and unplug the mains plugs and IEC connectors in and out a few times to clean them and you'll be sorted. 

Thanks James & Mike for the quick replies. So I'll just treat it as a tidying job.

I thought of making power cables to the right length and this seems the best way to go. I don't want to cut the original Naim cables so is there a good basic alternative cable type I should look for off a reel, or is it just a matter of getting the right gauge and then fitting a good quality plug? It's English 3pin outlets here in Singapore.

I recently had some very bad experience with ferrites. Don't take my experience as gospel but as a warning that there is more to it. I took the advice of some very respected members but it just didn't work out for me.

I had recently posted that I was having some terrible RF issues. The use of ferrites as a possible way to mitigate this came up but I held off until I located the source of the RFI and removed it. Then I thought about adding ferrites for good measure and purchased a bunch of different sizes from MacLab and set about attaching them to:

• The Ethernet cable nearest the NDX
• Both ends of the twin SNAIC leads between 282 and SCDR
• The XLR cable nearest the 250
• The speaker cables near the speakers (not over the whole cable but smaller ones on each inner cable)
• The non grounded DC power cable from the NAPSC nearest the 282

The result was immediate - and quite bad. Somehow the punch was all gone and things sounded slightly compressed and held back. A noticeable veil had come down. Within 3 mins I know this was a mistake. I left them on for a week anyway to make sure I was not knee jerking but they had to go.

First up, I have to say I do not come from an electrical engineering background. I know enough to be dangerous as they say. So when things were not sounding right, I went off and did some reading about the correct use of ferrites to see what I had done wrong. Some interesting points came up

• No use sticking a ferrite core on a cable without knowing what frequency range of RFI you are attempting to absorb. They are made with different materials to absorb specific ranges. So when using them to blocking incoming RFI, you need to know what frequency range you are attempting to block and choose accordingly.
• As per above, the ferrites pre-attached to some cables are not to block incoming RFI but to prevent them transmitting RFI known by the purpose (and therefore the frequencies carried) of the cable so that they meet legal standards for RFI radiation.
• Indeed you only need to stick them on the receiving end. Putting on both ends has no meaning but chaining them in the receiving end can add absorb more but it is not linear. You don't get twice the benefit from adding extra ferrites.
• On a differential cable like speaker cables, indeed you just put them on the individual conductor, not around the pair. So this guess was right.
• Adding a ferrite core increases the resistance of the cable so it in effect adding in a resistor to it to act as a high pass filter. But changing the resistance of the cable changes it's intended design parameters.
• They are not recommended for earthed mains cables and can in fact make things worse because you want the least resistance possible on the earth drain. 

I also have to admit, I ran the idea by a few others outside of the forum who have degrees in electrical engineering and did not get consensus. One advocated them and the other (my dealer) said it was a last resort when you know it is the only way left to address a problem - he strongly advised against it. I ignored my dealer, bought 3 bags of 10 at different sizes and now have to figure out which recycle bin they go in.

Hi DC1

My limited trials of ferrites on individual naim audio power supplies has provided musically 'strange' results; eg on the DAC or XPSDR - eg increased apparent definition which could be initially beguiling, but also with a 'flangy' effect and undermined solidity and whole ness.  My advice is to trial carefully.

However I recently taped two large 42mm ferrites across my flat 'cooker cable' mains spur just prior to the eqpt socket which virtually eliminated a hardness traceable to the wifi.  The Naim Unitiserve power filter only treats phase and neutral with the ferrite, and bypasses ground straight through which is probably the optimum approach 

Generally I prefer to 'quieten' other noise producers at the source, eg on different spurs.  Look at your fridge/freezer, microwave, alarm clocks, PCs, NAS, TV - turn them successively off and see where the significant improvements come from.  In my experience multiple noisy sources inter modulate, and often an element is more noticeable when it's elevated upon an underlying noise base originating from something else. If you can cure the root cause the others may lose their significance.

I once tried a filter with the 'more turns is better' approach and found it was actually having an adverse effect.  More turns == higher Q == susceptibility to resonance.  My advice is for incremental test trials before committing to anything permanent.

Also different ferrite materials are effective at different frequencies.  For unavoidable switching power supplies, you may find HY2 powdered iron rings far more effective at their lower troublesome frequencies of 10s to 100s of KHz into MHz.  I have a set of digital network power supplies off a good isolation transformer and keep my filtering on the secondary side of this.

You may also be able to persuade household members into turning off some noisy elements when not actually in use!  Explaining that the alternative is a $5,000 power supply upgrade can be motivating ;-)

Test methodically and diagnose before attempting a cure!

It does depend on your environment.

If you have a real RFI problem (i.e. significant electrical or radiated EM noise pollution) in the frequency band 0.5-500MHz they will almost certainly be beneficial.  If not they won't give you any advantage.  In most cases they won't cause a problem where there's no RFI present.  However there is always the case where a small amount of RFI alters the 'balance' of the system in a favourable direction; reducing this will be perceived as detrimental.  Their use on unbalanced digital cables is also a little more complicated.

I also agree on using them on the mains leads of many household appliances.  Because I now use a non-standard type of mains filter especially designed for my Naim system, I no longer need this, but previously I found this beneficial.  I also agree that standard mains filters are a bad idea unless the mains conducted RFI is terrible (mine is, hence designing a special unit to remove it).

Feeling_Zen,

Some of your information is incorrect, ferrites don't affect resistance at all, they only affect inductance, and then only in their stop bands.

Also, did you try using just some of the ferrites rather than putting them all on at the same time?  You may have found them beneficial in some positions, less beneficial in others, and the detrimental effect may have been caused by only a few of them.

You are particularly right about the case where there is an issue with a specific cause of the RFI.  Here indeed, injudicious use of ferrites can make matters worse not better.  The main place for ferrites is where the the RFI is non-specific broadband interference as in my case.  If you have a specific cause of RFI that needs to be attended to specifically.

It's not the cable that's differential, it's the mode of the RFI.  Cable usage is either balanced or unbalanced, the RFI can be common mode or differential mode.  Ferrites are used in different ways for CM or DM RFI.

However you are fully correct in pointing out that the only way to know is to try it.

P.S. pity you're not in the UK I'd have taken those spare ferrites off your hands!

@Huge, I have a feeling, giving that they cost next to nothing, the postage would be worth more than the ferrites are worth even in the UK. Otherwise you would be welcome to them.

Though after reading your post, I may give them a try on the non earthed mains cables of other appliances.

I did indeed put them all on together and this was already after the specific RFI problem I was resolved. Notice the one place I did not put them was on the HiLine since I did not have one that made a nice fit that would not pinch the cable and did not want to put additional weight on an already fragile plug.

It's interesting. I was always under the impression that it was purely an increase in inductance that was offered but got conflicting reports on this when reading up which seems strange since this, I assumed was governed by well understood non ambiguous laws of physics. I'll take your word for it over an untrusted source. You'd think (hope) a component dictionary published by a retailer specialising in radio parts would be more accurate. I'd be interested in your input on a few other things that did not match up on the sources I compared:

• One mentioned that they should not be used on any earthed cable especially a mains cable.
• The other mentioned that their use was detrimental on cables carrying AC current. I have to admit working in data centres, I see hundreds of factory fit ferrites on many DC power cables but never on a mains AC cable.
• One source advocated putting one ferrite only on speaker cables and only on one wire (+ or - they did not specify).
• The other mentioned putting ferrites on each wire of the speaker cable (so 4 in total) but at the amp end since this is were the RFI would generally arrive back to the unit rather than the speaker end.

I suppose I could have another whack if there is another way that is more beneficial. But the results the first time were awful.

I wouldn't discourage anyone from trying since you can't damage anything with normal use and they are incredibly cheap.

Hi Huge,

yes ferrites do not increase resistance, but they increase impedance at higher frequencies.  Many people seem to think that these two are the same thing, which is not true. Maybe one of Zen's sources got this wrong as well.

Hi Feeling_Zen,

I think trying them on two core mains cables and on the DC cord of SMPSs is a good idea; and yes I really agree on the high line, for both reasons.

You're also absolutely right about the laws of physics, there's no ambiguity.  Unfortunately some people don't rigorously distinguish impedance from resistance (presumably because they're both specified in Ohms).  On the specific points

Use on an earthed mains cable will be unpredictable, there is a one type of RFI where it may be the right thing to do (common mode RFI affecting different linked components to a different degree); unfortunately, this is also one of those situations where in very similar circumstances with the same type of RFI it can make things worse!

Use on AC cables depends on the frequency, the current carried and whether they are used in Common Mode or Differential Mode.  Common mode is normally OK, but can get troublesome if the AC signal approaches the ferrite's absorption band.  However high current with ferrites in differential mode is a no-no as the magnetic field saturates the ferrite.

I do use ferrites on my speaker cables (and haven't heard any difference, but they are used to increase the HF inductance of the speaker cable to make life easier for the amp as I don't use NAC A5).  I pass the + and - speaker leads through in opposite directions to increase the DM inductance.  There's no point in ferrites at the speaker end as the speaker isn't sensitive to RFI and has a low impedance anyway.

But yes trying it is a good idea, it may help and won't do damage (except hi-lines!).

@Huge,

Always a pleasure to read your input. I may go back and give myself some homework this weekend to try differing configurations. Am starting to thing ferrites in the SNAICs might have also been wrong since I believe they carry an earth pin but need to check that again.

feeling_zen posted:

@Huge,

Always a pleasure to read your input. I may go back and give myself some homework this weekend to try differing configurations. Am starting to thing ferrites in the SNAICs might have also been wrong since I believe they carry an earth pin but need to check that again.

You know what, that's a very good point, I think you may well be right on that.

SNAICs carry DC power, the signal and the zero volts line, but they are quite sensitive to disturbance affecting the performance and stability of the amplifiers - hence the talk of 3rd party SNAICs being banned on these forums.

I think ferrites may very well be a bad idea on SNAICs.

I'm also going try removing some of the ferrites I have (the ones I put on earlier and haven't subsequently tested individually), I'll let you know if I get an improvement.  Maybe we can both learn from this, it could be a very enlightening exercise.

Interestingly I do have ferrites on the lavender interconnect, and I did specifically test them for my set-up: They are definitely beneficial in reducing background clutter and revealing details in the music without any detrimental effects.  This is despite the fact that the inteconnect caries the zero volts line just as the SNAIC does.  On the other hand the interconnect isn't designed as an integral part of the amp / PSU system, so the system is designed to be far more forgiving of the characteristics of the interconnect.  Oh dear, making guesses on what to do does get so complicated!

Thanks guys for a lot of in-depth advice and discussion. I guess since I'm not feeling/hearing any specific issues with my setup on its dedicated spur, I'll only do some experimenting with ferrites if I come across some without too much effort, and then probably only trying them for unearthed power cables on the non hi-fi equipment.

Next step is to approximate a homebrew version of the Naim power cables in a shorter length so I can tidy up the power cabling without coiling or trailing around the floor behind the shelf unit. I'll need half-metre lengths for Qute and power amp. 

I'd welcome your recommendations for cable type/brand preferably at DIY rather than audiophile prices!

DC71 that discussion may well be outside of the Forum AUP.  Discussion of DIY and electricity - not always happy bedfellows and something that Naim don't wish to encourage here.  Thanks.

Noted Richard, sorry. Do you know if Naim dealers are able to supply 0.75m or custom length approved power cords?

I believe that the length of the Powerline was optimised through R&D at Naim, i.e. shorter doesn't sound quite as good.  On that basis I doubt they would do a shorter than standard length.

However, maybe a custom Hydra such as those supplied by Grahams Hifi would suit your needs?

@Richard, since you have popped in on this thread, to the best of your knowledge, have Naim ever expressed an opinion on ferrite cores?

I don't know, FZ.  You could always put the question to Naim - send an e-mail to Steve Hopkins.

Done. Will share response.

feeling_zen posted:

Done. Will share response.

FZ, would you be kind enough to share what Naim has responded you in reply pertaining to the ferrite core?

I took the advice from Simon to clamp one on the DC1 cable but most are on motor driven home appliances such as refrigerator, air-con compressor, washing machine, computers so on. The idea is to mitigate the RFI being crept into the mains.

Let me clarify that the point of asking Naim is not to get a definitive answer or contradict any opinion. The physics around this topic might be unambiguous but the opinions on their impact on SQ may vary. If there was only a single definitive answer for these things, there would be no diversity in the hi-fi market at all. There are some very intelligent members on this forum and Simon and Huge are a couple examples. Even after taking advice and getting bad results, I am convinced enough to have another go in a different way. Any response from Naim is just another view as valid as those expressed.

With that in mind, here is the diplomatic response I got from Steve verbatim with his recommended reading links.

Hi

Thank you for your enquire, Naim has nothing for or against ferrite cores, in fact HDX has one on its mains input at the product end but we do not have a guideline on the use. I would have thought the suppliers of the cores would give guidelines on their best use, see

http://www.cnet.com/forums/dis...r-not-to-use-578132/

http://www.murata.com/en-eu/ap...m_xxx_xxx_cmn_nv_xxx

Regards

Steven Hopkins

I'd previously tested the following ferrites individually and found them to be beneficial and removing them removed the benefit.
The Ethernet cable, switch to ND5 XS
The interconnect cable
Several SMPS for other devices

I'd also specifically tested the ferrites on the speaker cables and left them because they make no difference sonically, but will give the amp a slightly easier time by increasing the cable inductance at HF.

I hadn't tested the ferites on the mains cables (there were two on each).
Removing both the ferrites improved the sound.  Replacing just the one that works at higher frequency (5-500MHz) cause a loss of resolution, but replacing the lower frequency ferrite not only brought back the lost resolution in the system but slightly improved it over the situation with no ferrites at all.  Go figure that one!

Regarding the OP, I think Huge has offered the best advice.. The effect of ferrite cores will depend on the electrical environment... If there is HF noise on the conductor, then a ferrite ring will add reactance to the conductor and impede the HF energy. On mains I would use large quality ferrite and check they don't get warm (from losses in the ferrite), also a larger ferrite is less liable to saturate and cause a worse outcome.. but I can only see this being a factor if you have seriously bad electrical HF noise issues.(RFI)

With interconnect, Ethernet and other small signal leads, ferrites are easier to administer without side effect. However do bear in mind you might have got used to the effects of RFI on your system.. and it gives a character you like and become used to.. so removing it could produce a worse outcome for you.. Life is never straightforward  

Yes looping a conductor through a ferrite ring increases the reactance and therefore the impedance and effectiveness of impeding HF noise.

On mains leads another effective RFI inhibitor is using conductors that are braided together..I am not sure whether in fact this is what is done with the Naim Powerline.

Simon

Thanks Simon, Huge, FZ and all the others who provided a lot of food for thought on this and helped to answer my general questions about mains, ferrites etc.

Since I'll be in the UK over Xmas I have ordered a couple of shorter mains leads (thanks Mike for the MCRU pointer) which will be my first step towards tidying up and separating all the cables. I went for the basic IEC plugs as I have a feeling that some of the bigger ones may reduce the benefit of Naim's floating sockets (due to the extra weight/leverage from the bigger plug). I'll feed back in Jan on how it pans out and whether I try some ferrites but I'll also be running in a new amp at that time too so I'll be doing cable swaps bit by bit after a run in period.

Cheers,

Dylan