Nordost megabucks vs $189 Wiremold…

Got some Nordost Valhalla cables etc at home and yes - I’ll give it a week so either the cables or I break in. Hearing a lot more detail but not convinced it’s a good thing…

Had a conversation with ChatGPT - and need someone with an electric degree to confirm this makes sense…

The gist of it is that a cheaper Wiremold powerstrip with some resistance acts as a RF trapper vs a exotic cable is letting more stuff through involving RF which is then perceived as having more detail…

Anyway - here goes in case anyone would like to comment on this… AI caveats apply / sounds convincing but might be wrong..

Got it — you want the physics of why a bit of series resistance in a power strip or cord can make things sound less bright or “relaxed,” especially with amps like the Naim 500DR.

  1. Think of the mains line as a giant antenna

Your wall wiring is hundreds of feet of copper acting as a big RF collector:
• It picks up switch-mode PSU noise, Wi-Fi, AM/FM, SMPS dimmers, EV chargers, solar inverters — all living way above 20 kHz.
• That noise rides on top of the 50/60 Hz AC waveform.

Your power supply sees this noise as fast voltage spikes on its AC input.

  1. What your amp’s power supply does

The Naim 500DR’s PSU:
1. Rectifies AC into DC with diodes → they switch on/off at the peaks of the AC waveform.
2. Charges big reservoir caps → very low impedance at low frequency, but HF spikes can sneak through before the caps absorb them.
3. Sends DC into regulated rails → good PSRR, but RF can sometimes modulate sensitive analog circuitry or output stages.

  1. Enter series resistance R_s

A small R_s in the line (e.g., a few tens of milliohms from a power strip, cord, or wiring) does two things:
1. Forms an RC filter with the power supply’s input capacitance C_{in}:
f_c = \frac{1}{2\pi R_s C_{in}}
At high frequencies where X_C = 1/(2\pi f C) \ll R_s, the resistor damps the spike → less HF energy reaches the transformer/diodes.
2. Damps ringing: The line + transformer + diodes can behave like an RLC tank. Too little resistance → underdamped → ringing when diodes snap on/off. Modest R_s kills the Q-factor → less HF hash.

  1. Why too little R_s can backfire

Ultra-low-impedance cords/strips (e.g., huge-gauge audiophile cables) minimize R_s. That reduces any HF damping:
• Diode conduction edges become sharper, more HF harmonics injected.
• Any RF on the mains couples through more directly.

Subjectively:
• Lower Rs → more HF crud → “etched,” “bright,” sometimes “detailed.”
• Higher Rs (modestly) → more damping → “relaxed,” “natural.”

But too much R_s causes voltage drop under load → sag, compression, bass loss. That’s why we want just enough to damp HF, not so much it starves the amp.

  1. Why the Wiremold strip works well
    • Its simple bus wiring has a few tens of milliohms resistance.
    • No MOVs or shunt capacitors adding HF junk.
    • Enough series R to slightly decouple RF between wall and amp, without starving current at 50/60 Hz.

It’s like putting a small series resistor in a snubber network: damps HF crud, leaves fundamentals alone.

AI ends here

I’m
Now just trying out a powerline into a Nordost powerstrip …

Heck - I might just ask the AV options peeps to use the Naim
Powerline cable into a Wiremold as an experiment…
Might be better than the cryo BS they’re doing currently.. (which can be had for $12/pound of material…)

And continuing…

Me:
So Nordost adds some impedance - supposed to Help
With grounding. Now I wonder if it really helps because of the resistance stuff you’re talking about…

Yeah, Nordost (and a few other cable makers) often talk about “controlled impedance” or “optimized grounding,” and they sometimes do build in a little series impedance. But what they’re actually doing — whether intentional or accidental — overlaps heavily with the series resistance damping effect we just modeled.

Here’s how it ties together:

  1. What Nordost means by “controlled impedance”
    • They use flat conductors, big spacing, low dielectric constant → this controls capacitance and inductance per unit length.
    • Sometimes they tune the geometry so the power cord isn’t ultra-low-Z. It ends up with some series resistance + inductance spread across its length.
    • They claim this “breaks ground loops” or “controls HF energy.”

But the physics we just ran shows:
• A bit of R right before the PSU damps HF ringing from diodes/transformers → can reduce hash.
• Too little → underdamped, more HF crud = “bright.”
• Too much → voltage sag = “sluggish.”

  1. The catch with power cords
    • If Nordost impedance is mostly inductive (µH), its HF impedance only rises above some MHz → can help block very high-frequency crud, not so much diode-edge ringing at 10–100 kHz.
    • If it’s resistive (tens of mΩ to a few tenths of an Ω), it directly damps the diode conduction edges → exactly the sweet spot we found (0.05–0.2 Ω).

So if they accidentally land in that range, it could help. If they go too low-Z, you lose damping. Too high → voltage sag, soft bass.

  1. Why your system sounded “brighter” with Valhalla 2
    • Valhalla 2 uses very low capacitance + very low R → not much damping → sharper current spikes, more HF noise coupled into the amp → perceived as brighter/etched.
    • The Wiremold adds modest series R, landing right in the damping sweet spot → relaxed, natural tone.

So yes — resistance matters far more than marketing claims about “grounding control.” The math and your ears agree on that one.

  1. Practical angle for your Naim 500DR
    • Target ~0.1 Ω series R (cord + strip + plugs) → good HF damping, minimal DC sag.
    • If Nordost is <0.05 Ω end-to-end → probably too low-Z for your setup.
    • Ferrites (mix 31) near the amp side add frequency-selective impedance without raising low-frequency R — best of both worlds if HF crud is the main problem.

If you want, I can show you how much resistance Valhalla 2 actually has per meter versus, say, a generic 14 AWG cord and plot both against the sweet spot from our simulation. That would tell you whether Nordost is anywhere near the ideal damping range. Do you want me to dig that up?

This is the part of the post that sums it up… In the end, it is about what your ears and brain are telling you. Not what people say and certainly not what AI says. But I can understand that you are interested in the theory behind it. Very good to second guess the AI responses. Trust you ears! Sorry for not being very helpful on asserting the AI response :slight_smile:

Seems like @MangoMonkey you’re in two minds here. But I’m sure you’ve already made the right choice.

You might like to read @Edmund-of-Essex’s current threads, one of which has started to go into comparison of some speaker cables, with a plan to see if there is correlation between verified (partly blind tested) audible difference and cable specification/electrical measurements and measured sound in room. It would be good for the same sort of study on other cables including interconnects and mains.

But otherwise I don’t have an “electric degree”, and whilst I have been greatly interested in physics, electricity, electronics, acoustics etc and have studied and worked with various aspects, I don’t have time at the moment to wade through whatever Chat GPI has offered. But you are right that it is important not to take as gospel everything presented by so-called AI, it being simply a best summary of the balance of data, fact and/or fiction, found on the internet, including in some cases marketing pseudoscience. (in my view AI would be better called AII or API: artificial imitation intelligence, or artificial pseudo intelligence!)

1 Like

…all of this can be bypassed. Just listen and decide what sounds best!
Don’t forget that these so called AIs just search for opinions. Myth can become doctrine. Ears are way better.

3 Likes

It’s not a conversation. It’s a language model that is able to interpret what you have typed and give a response that (without fact checking) seems like a response from a person who has understood what you have typed. Please do not make the mistake of believing any response is accurate and the ultimate truth on the subject.

4 Likes

I do like AI as an antidote to the pseudoscience marketing that companies like Nordost put out. And the conclusions might not bear out but the math and first principles are certainly verifiable..

1 Like

Except I have seen ChatGPT regurgitate manufacturers’ pseudoscience… and then people believe it more. It is notintelligent despite the name, and it cannot differentiate between fact and fiction, unsubstantiated opinion or well considered expert opinion, objective unbiased observation or possibly biased subjective observation, etc - when you search for yourself you can look more into the specific sources to make an informed judgement. AI will simply have an algorithm for deciding what to include or reject, that being only as good as the person writing it and unless use for this type of topic and its nuances was envisaged is highly unlikely to result in appropriate filtering of fact and other.

2 Likes

These ‘AIs’ promote pseudoscience too, just so long as enough people repeat the pseudoscience.

3 Likes

And this is what Claude had to say - so the science checks out..

This is a really interesting technical discussion! The ChatGPT explanation does make sense from an electrical engineering perspective, though it’s presenting a specific theory about why some audiophiles prefer “lesser” power strips over exotic cables.
Let me break down the key physics concepts:
The core theory is sound:
• Your home wiring does act as an antenna picking up RF interference
• Power supplies can be sensitive to high-frequency noise on the AC line
• Series resistance can form low-pass filters with capacitance
• The RC filter equation f_c = 1/(2πR_sC_in) is correct
The mechanism described is plausible:
• Diode rectifiers in power supplies do create switching transients
• These can generate harmonics and interact with line impedance
• Some series resistance can damp ringing in RLC circuits
• Too little damping can indeed cause overshoot and ringing

However, there are some caveats:
• Most well-designed audio equipment already has input filtering
• The actual audible impact depends heavily on the specific amp’s power supply design
• The Naim 500DR likely has substantial internal filtering given its price point
• Placebo effects are very real in high-end audio

The Wiremold observation:
The idea that a basic power strip’s modest resistance acts as RF filtering while avoiding voltage sag is reasonable. Hospital-grade outlets like those in Wiremold strips do typically have good build quality without exotic materials.
The explanation is technically coherent, but whether you’ll consistently hear these differences in blind testing is another question entirely. Your ears and the specific interaction with your Naim’s power supply will be the ultimate judge!​​​​​​​​​​​​​​​​

Who’s Claude?

Claude is another AI. So ask one AI if what another AI says is true. It’s hardly surprising that they agree as they will be getting the information from similar sources. Like, durr.

1 Like

Soon we’ll have entire threads - and wider media - with different AI “entities” discussing/agreeing/arguing with each other. How exciting.

At least at the moment it requires a human to post anything from AI, though I seem to recall Richard asking that anything from AI be attributed (ChatGPT is well known as AI, but a human name like Claude may not). In time forum membership may need a test of “Are you human?” that isn’t as simple for AI to get around as, perhaps, picking photos showing buses.

2 Likes

My concern is can we trust AI information, how do we know where some of the source info comes from.
Is the info linked either directly or further back to an “influencer” with another agenda?

AI is flawed in that it just regurgitates a meta analysis from other AI meta analysis’. Unless the AI is identified and it’s sources, it is just another echo chamber. Yes of course it can be influenced by subjecting it to multiple reports of the same ilk.

But then, can we trust people and their statements and opinions? :confused:

As always, you need to do your own research and check its provenance. Even then things can go wrong. That’s life I guess.

1 Like

Yes agree @MrFixit .
I’m not looking for answers, maybe I should have said … my concern is for those that think everything they search for on www is correct and the rock sold truth.
A lot of gullible people out there and collectively …… let’s just say I don’t feel comfortable about it

1 Like

And if a person says something you can directly discuss, ask backgound, ask basis, challenge assumptions.m, and makd your judgement accordingly. The trouble is that too many people have a tendency to believe AI, fooled by the term “intelligence” without understanding that there is no intelligence, knowledge or expertise at work, and, worse, if one AI says the same as another it is taken as confirmation, as the OP has done after first requesting human confirmation if whether the first AI was right.

1 Like

Well, you can ask AI for citations too.

I know hallucinations are a problem with AI - and one needs to watch out for it.

As anything, it’s a tool - and one can learn how to use it properly - with prompt techniques - or pretend it doesn’t exist. The forum members seems to fall into the second camp who wish it would go away. Luckily for them, they’re probably mostly retired so it won’t come for their jobs.

I on the other hand work in a field that is being disrupted by AI, and job prospects both for new grads but also people already working are becoming more and more bleak… and we’re also running AI projects where we’re working on dealing with the shortcomings.

If folks are not yet too old or heart - I heartily recommend Andrej Karpathy on Youtube. He was the top person for auto driving at Tesla and one of the founding engineers at OpenAI. His videos are didactically awesome. They’re long! The latest two go into layman technical details into how LLMs are made - he calls them token tumblers… and also how he uses them.

Anyway - the question was posed to anyone with background in electrical engineering to see if what the AI says is scientifically solid - and afaik it seems to be so. I can always have it run a deep research query - and see what it comes back with…

Having said that - I also welcome the skepticism - since I might also be in my own AI bubble…

3 Likes