Actions
Forum Thread Details
  • Replies 21 replies
  • Subscribers 106 subscribers
  • Views 4474 views
  • Users 0 members are here
Related

Dealing with Unexpected Emissions: The Fellowship of the Ferrite Ring

shabaz


I was experimenting with bits of circuit, hopefully, to become a power supply eventually. It is a work in progress.

At this stage, I’m using bits of scrap copper-clad board, each with a sub-module. It’s too early for a PCB.

image

However, during the prototyping, I noticed (amongst other issues that I need to solve!) a particular artifact on the ‘scope trace when looking at the output.

This situation ended up being a good example of where it pays to zoom out and look at a lot more of the spectrum view than expected!

Admittedly this was a lot easier to identify using the MXO 4 ‘scope, but it would be possible to spot with other test equipment, such as a spectrum analyzer and a near-field probe. I just used the normal 'scope probe, I didn't do anything special.

With the MXO 4, the anomaly was zoomed in, and then using a logarithmic-frequency spectrum view, it was possible to see the extent of the problem, with the left side of the spectrum view showing the wanted oscillation (fundamental frequency) and its harmonics, and the right side showing the spurious output.

image

Now I could try to do something about it!

I used tiny ferrite rings (legend has it they were forged by Fair-Rite, part code 2643001501 ) on the base input of the transistors. Essentially, they act as a bit of additional resistance for the higher frequencies. The rings cost $0.02 currently (although more expensive in the UK).

image


Back to the ‘scope:

image

Much better!

As mentioned, I still have other problems to deal with, but I thought it was a nice example of why it makes sense to do testing to frequencies far higher than expected, just in case there are spurious emissions, and why it can be a good idea to possess the rings because one never knows when they are needed.

Thanks for reading!

Parents

  • I've frequently just used tape to hold wires together, but it's very fiddly with small ferrite cores. I've tried adhesives, but superglue just gets on hands more than on the core! Impact glue takes too long to take hold. Mixing epoxies is too wasteful and time-consuming for these tiny quantities.

    I decided to try some sort of optical glue. It's not perfect, but I'm liking it so far.

    The glue is used for nails (I believe). Anyway, it is low-cost from Amazon or AliExpress. Both pots in the photo below contain the same stuff, it is called "Solid Sticker Adhesive Gel". The consistency is like plasticine (kid's modeling clay) or slightly softened toffee. I would prefer it to be a bit thinner.

    image

    For the UV light, I'm currently using a handheld light, and closing eyes and looking away for about 20 seconds until the adhesive sets. I don't like this method, it doesn't feel safe, so it's probably better to find some sort of covered UV light.

    image

    The UV lights come in different wavelengths. I'm using 395 nm, I have no idea if that is optimal or not, since there's no specification for the adhesive.

    The adhesive needs to be pushed well into the wires on the toroid (at just the ends, not the entire winding of course) using the spatula since the adhesive is quite thick. 

    The end result is great though. After 20 seconds of UV illumination the adhesive sets well enough and needs to be cut away with wirecutters if the ferrite core ever needs to be reused.

    image

    Any suggestions for a better method would be great to learn. Also, if anyone knows of a low-cost optical adhesive that is slightly softer than the one described above, it would be interesting to hear about it.

Reply Children