We never had transformer hum in our previous houses, nor in the current one in the early years.
However, we started with 1 house ( ours) and a security cabin (the builder’s) on our sub-station, so our mains feed was awesome.
Then the builder built more houses, a lot more houses, 4,500 more houses. Along the way, all our full width Naim power supplies started to hum, with the new 555PS being the most sensitive.
Our mains noise didn’t come from fridges, hair dryers or electric blankets in our house. It came from the dirty loads that other houses were adding to the mains e.g. the oft used Powerline ethernet over mains.
In our situation, the dedicated radial may have lowered supply impedance but it did nothing to isolate us from the incoming noise. A separate consumer unit would be equally ineffective.
It took a BPS to stop the humming.
Our situation cannot be unique. As more noisy loads are added into most homes via SMPS, cheap solar PV inverters, ethernet over mains etc., so I would expect the issue of transformer hum to become more prevalent, not an isolated (no pun intended!) series of one-offs.
Hi Bluesfan,
On another thread, now removed, you gave some specifics around the actual product that I bookmarked but have lost.
Could you possibly provide again?
Many thanks,
I spoke with the very helpful Engineering Director at Airlink Transformers in the UK.
He recommended adding together the VA capacity of all the Hifi transformers, then doubling the total to reach a starting point for the size of the BPS.
This led us towards a 3kVA balanced power supply but for the extra £50 or so, we went for the larger, lower impedance, 5kVA one. By selecting the MP variant, we have the option to change the output voltage in approximately 10-12V steps.
We ended up with a BPS5120MP, selecting an output tap to drop the output voltage from a measured 242V at the living room sockets to 230V.
All the above was checked with Naim before proceeding.
If you do follow this path, then it is essential that the BPS is fitted with double pole isolators to ensure that both the live and “neutral” are isolated in the event of a short circuit or earth leak. The Airlink BPS comes with this as standard.
A poster wished to directly involve a Naim member of staff, I asked poster to refrain but they wished to continue.
I have contacted Naim independently on this matter through normal channels so did not want this to compromise my process in any way.
I asked for thread to be closed as a result.
The poster, on another thread, is free to do what they wish I have no concern I did not wish to be dragged in.
The big question is does anyone have any hard measurements on how it affects the D.C. rails that supply the head units. I think the main advantage with the 552 is a differential that effectively removes noise with a common mode rejection
Toroidal transformers are favoured over other linear designs like EI core because the flux they generate is confined to the core. In simpler terms, they cause very minimal EMI. So with an EI core, the EM radiation is quite high and this has to be mitigated by means of some shielding and distance from sensitive components. With a toroidal, the EM flux is so tightly confined that you do not require shielding and the layout of the product is simplified because it can be closer to sensitive components.
While the above is fine, it is not the whole truth. Toroidal cores behave like that in optimal conditions. In other words, in optimal conditions, a toroidal core far outperforms an EI core. But (and there is always a “but”), in sub optimal conditions the reverse is true. EI core behaviour varies very little when presented with DC offset or over voltage. It’s not great in either situation. But it doesn’t really degrade much and is less likely to hum. With a toroidal, over voltage or offset cause the core to saturate. And when a core saturates, that nice donut shaped sharp flux path that the toroidal was actually selected for no longer exists. It’s all over the place. Except now with a transformer that isn’t shielded in a product that probably doesn’t have a board layout to minimise the effects. In other words, once a toroidal starts humming, it is behaving badly negating all its benefits and potentially inducing EMI in sensitive analogue components or passing induced noise over DC rails that pass through the now unconstrained flux path.
Basically, it is causing more problems than just humming. The hum is generally a sign it is not performing well in the current conditions.
This is actually why a few high end audio makers opt for EI core. Consistency of performance. I guess since the forum loves car analogies, it’s like a Porche 911 that always performs versus a McLaren that’s always in the shop getting something fixed.
Just a reminder to all, this thread is not a continuation of the other closed thread. This is not a, “let’s whine about a problem” thread. This is a data collection thread to find (or not) anything of potential use to users and/or Naim (if they care).
We just need votes and any first hand observations about the nature of this if anything interesting was experienced. No axe grinding please.
Interesting and if that’s the case then Naim have really made a mistake in pursuing the use of such devices then. I would hope that they have fully analysed and tested all this, which I assume they have. It would be interesting to see something from @110dB in response to your post.
I remember , and I read that relatively often, that some guys have bought a Naim component that was humming too much , returned it, and got a replacement for the same component that was not humming.
It’s a bit a lottery sometimes.
