New measurement equipment – LISN – insight into power supply noise!

Your author does like measuring equipment. It says “your author” deliberately, because not all authors are into scopes, resistors and analyzers. Anyway: to each his or her “thing. The danger of this “thing”: it never ends. And yes: something has been added again: a LISN. What? Well… listen..

LISN stands for Line Impedance Stabalyzer Network. It is also called an Artificial Mains. We apply it in two ways; it is a very effective disconnection from the mains – which is often very convenient! – and it makes it possible to measure supply noise. Very accurately even, since this LISN can measure from 9 kHz to 30 MHz.

We did not purchase the Tekbox software with it (about 500 Euro extra, so something to think about). This software works seamlessly with the Rigol Spectrum Analyzer and allows very accurate measurements by applying stitching. (Six separate, highly accurate blocks combined into one large measurement). So it is interesting and should it prove valuable at a later stage, we can think about purchasing it.

Transformer

An LISN must be connected to an isolation transformer. We found that out ‘the hard way’…. the fuse blows right out if you just plug it in and then turn it on. Very nice for the PFsense router, Synology NAS and Unraid server that are – apparently – connected on a defected UPS. The second UPS switched over nicely. So the switches and other servers and NAS systems are simmering along nicely. Oh well: now we know we need to replace the other unit. Once hooked up to the transformer, everything is running smoothly. The first test measurements on switches show nice results that also show similarities with the other measurements we performed with white noise. So, promising!

Solid ground

What is important in a LISN measurement is that the equipment is grounded to a metal plate. So the LISN is attached to the metal plate with a grounding strip. The CDN T8 unit (Coupling / Decoupling Network for 4 twisted pairs, so also Ethernet) also will be fixed to that plate. The ‘device under test’ (DUT – Device Under Test) itself is separate from that and should be about 10cm above the plate on a non-conductive elevation. We just grabbed a piece of wood. We will use some extra foam to raise it a bit more. Just as the cables need to be a little raised from the metal plate for good measurements. All this is necessary to avoid interference in the measurements.

What do we get out of these measurements?

Interference going through the power supply to the mains. And how immune a device is to external interference. Both are crucial in a hi-fi system. Power is everything. And a good power supply will send little noise towards the mains and will care little about external sources of interference.

We now have two major tests we are working on. A large switch test with over ten switches and a mass test of integrated amplifiers of 2500 Euro. In both cases we can add this new measurement to the measurements we are already doing. We then immediately have a good idea of what is ‘normal’. After all: we can directly compare.

Why even more?

We understand that the number of measurements is growing rapidly. However, we try to check all sides, so that you get a complete picture of a device. And also gain insight into what is and what isn’t important. Because we’ll also figure that out; how audible are some distortions or how damaging is interference? Does it matter?The direct link between measuring and hearing is still far from being made. We will keep trying!

Leave a Reply

}