Search This Blog

Saturday, 1 July 2017

EMC 'n' all that Jazz

Well,

I've been having some issues when I TX on 6M CW. Very strange in that my Radio PC (the one sending the CW) shuts down - it doesn't crash - it performs an orderly shutdown.

This is definitely an RF issue as it only happens when the TX power is above a certain value.

So, by using a process of elimination, i.e. removing cables from the back of the PC one at a time and seeing if the problem goes away, I concluded that it's probably the HDMI cable to the monitor (well, one of the monitors) that's causing the problem.

This has lead me to question the effectiveness of ferrite suppression and other such gubbins.

Now, all of us hams will have purchased a bunch of clip on ferrites at a rally; these are supposed to be made of type 31 material which is rated up to 500MHz.

I mean something like this:


Now, these are designed to clip on a cable, effectively providing one turn through the ferrite. How well does that work then? So here's the spectrum analyser showing 0 to 100 MHz and a simple loop back from the generator to the input - no ferrite here:


So, lets now add a single turn of the clip on ferrite and see what difference it makes:



So the answer is, quite expectedly, not a great deal. So, let's increase that to 5 turns:



So, thats much more like it.

We have to conclude that clipping these ferrites onto cables around the shack is next to useless at HF - we need at least 6 turns through the material before we see any significant attenuation.

There are a number of larger ferrites available, using the same material, but bigger:


These will allow you to get multiple turns of coax or mains cable, or in my case a HDMI lead through the core, and most importantly they are also clip on.

I've tried a few combinations of different cores on my HDMI cable to see what works best; there are all sorts of other issues creeping in now though, like the resonant frequency of the cable itself:




The bottom picture above effectively gives me 6 turns by using 6 cores; the image above uses much more expensive cores and passes the wire through multiple times. They both have much the same impact.

So, I plan to add the cable above as an extension to my existing HDMI cable and see if the problem is solved.

Here's our very beautiful Elmo enjoying the fact that summer has finally arrived:


Local conditions.

3 comments:

  1. I wonder if these bargain unmarked ferrites found at rally's and on eBay are using genuine type type 31 material? It would be useful to compare them against the real things from for example FairRite Grade 31 which are considerable more expensive.

    ReplyDelete
    Replies
    1. Agreed; my large ferrites are fairRite type 31, but the smaller cheep rally sourced ones are completely unknown.

      Delete
  2. I have the opposite problem with my external monitor connected to my laptop. On 4m and 6m the HDMI cable radiates noise that is picked up by the antenna on the tower. Puttng clip on ferrites at either end of the cable has improved it to the point that 6m is now OK ( the band noise is the limiting factor). But on 4m I still have some birdies from it.

    ReplyDelete