DIY Biconical antenna 30-300mhz?

[RogerThat]

Hi,

Been reading up on EMC testing the last weeks.
To be able to do basic pre-compliance tests I need to add a 5uH LISN (I'm just doing DC 12-24v products)
and a biconical antenna to my arsenal of test stuff. The antenna should be the basic 30-300mhz.

After some intense days of searching the web I found a decent design for the LISN (here on the forum) but with the antenna
I'm drawing a blank. Can someone point me to a proven design for this antenna?

[tautech]

For such low frequency you shouldn't need much more than a good scope, maybe with FFT and a near field probe, just something like Dave cobbled together recently:
https://youtu.be/2xy3Hm1_ZqI

[RogerThat]

If I use a near field probe , how do I translate the measurement to that of the EMC standard (biconical antenna at 3m distance? I think).

[David Hess]

If I use a near field probe , how do I translate the measurement to that of the EMC standard (biconical antenna at 3m distance? I think).

The measurements can be translated by also using an EMC standard biconical antenna at 3 meters distance and comparing them.   

This is not quite as useless as it sounds.  Compare them once at the beginning when you pay for EMC testing of your design and it fails and then use the near field measurements to determine when you have gained enough reduction to pass.  There is just too much going on between the near and far field in a given design to compare near and far field measurements except through direct measurement and sometimes even that is not enough.

[EEVblog]

For such low frequency you shouldn't need much more than a good scope, maybe with FFT and a near field probe, just something like Dave cobbled together recently:

A near field H-probe is quite different to the biconical (EM) far field antenna the OP is talking about.
I've been looking myself and it seems to be rare as hens teeth, even 2nd hand.
The DIY log-periodic PCB based antennas are usually only useful for above several hundred megs.

[tautech]

For such low frequency you shouldn't need much more than a good scope, maybe with FFT and a near field probe, just something like Dave cobbled together recently:

A near field H-probe is quite different to the biconical (EM) far field antenna the OP is talking about.

Quite so but for DIY EMC, near field is where it's at.
Leave the far field to the experts in the labs.

Or could this be a good subject for a future vid ?

Further, far field testing is at risk of erroneous results without using a anechoic chamber.

[EEVblog]

Quite so but for DIY EMC, near field is where it's at.
Leave the far field to the experts in the labs.
Or could this be a good subject for a future vid ?

I think it's possible to get decent far field measurements in the lab (or at a DIY OATS) just need a decent low cost DIY biconical for the 30M-300M range.
I want to do a video on it but getting a suitable antenna is not easy. I could rent one or ask a manufacturer to send me one, but I'd prefer a DIY solution.
I'm surprised some chinese company hasn't just taken an existing biconical and simply copied the design down to the mm, would be fairly trivial. BUt then you have the issue of the balun matching.

[EEVblog]

You can get these cheap to do >200M, it's the lower frequency stuff that is tricky.

https://www.aliexpress.com/item/1PC-Broadband-wideband-antenna-EMC-EMI-antenna-Directional-antenna-290M-1-1G-EMI-rectification/32966314464.html

[tautech]

A customer and member here sourced a range of PCB antennas for his EMC work using a SSA3021X.
AFAIK they served his needs well.
Might have been these guys:
https://www.wa5vjb.com/products1.html

But here, the OP hasn't given us much to work with, not even what equipment he has or what he's trying to get compliance for.

[hagster]

Just make a 1 meter tall bow-tie antenna(copper tape on a bit of wood will work). It will be close enough. As shown on Antenna theory, spot solder the coax to one arm of the antenna, thia will help isolate the cable.

http://www.antenna-theory.com/antennas/wideband/bowtie.php

You may need to do a characterisation of signals that are already present so you can discount them.

[tautech]

As mentioned in previous posts in an exchange with Dave here's what my customer did:
https://www.eevblog.com/forum/testgear/howto-calibrate-diy-emc-pre-compliance-antenna/

Some good advice and links to suppliers and papers in that thread.

[Jay_Diddy_B]

Hi,

To the original poster.

You need to find a copy of MIL-STD-461A. This is the first version. Fortunately, I have a copy and I have attached here.
There is a dimensioned drawing for the biconical antenna in this spec. If you build it to the drawing it will match the curves in the spec.

The dimensions are in inches.

I have built a copy of the antenna from these drawings.

Regards,
Jay_Diddy_B

[HalFET]

A fellow student built one for his master thesis back in the day because the university was too cheap to buy one. I don't know how he designed it specifically, but I do remember how he physically constructed it. He started from two circles cut out of insulating foam panels. To keep them at a distance from each other he used pieces of PVC pipe around the periphery of the foam panels. The "cones" were constructed from a radial mesh of thick copper wires tensioned between the periphery of the disks and the feed gap. To support them in the centre he had a thick PVC tube that they ran through. Think he either crimped something onto them or knotted them to keep them in place. It wasn't pretty, but it performed quite well in measurements. I suspect if you want something more solid you could probably build it out of thin copper plumbing pipes. Keep in mind, at 300 MHz the wavelength is still quite long, so you don't need an ultra dense wire/rod mesh to represent a solid structure. Alternatively, you could grab some thick copper foil and put it over something made out of rotan or wood I suppose?

In terms of design, I think the evaluation version of Antenna Magus was capable of calculating biconical antennas.

[David Hess]

Self-calibrated far field measurements are possible but time consuming.  I did it by building reference antennas to generate calibrated far field intensities to then calibrate wideband directional antennas which made the actual measurements at a relatively RF quiet and flat location.

The better option if you can afford it is to treat the near field measurements as relative so when the first EMC compliance test fails, you know how much they need to be reduced.  And before the first expensive EMC compliance test, use near field measurements to evaluate basic good EMC practices.

[T3sl4co1l]

I built a bowtie antenna that I'm happy enough with.  I don't have any way to test its response, mind -- it seems to have good reception from ~100MHz to 1GHz+, but relying on ambient sources, this isn't very meaningful...

Quick terrible pic,



Anyway, it's flat, about a meter long, constructed from 18ga. bare wire, using five radials fanned out about +/- 30 degrees, then a few odd-angle crossings to stiffen things up.  Wires are twisted and soldered together.  The on-axis wires are supported by a 1"-wide strip of G10.  At the vertex of the bowtie, a 1:2 balun is made from two twisted pairs and a couple ferrite beads.  The pairs are wired in parallel going to the BNC connector (each pair is about 100 ohms characteristic impedance, so two in parallel gives 50 ohms), and wired in series (centertap grounded) at the antenna side (for 200 ohms differential, close enough to what I would expect from a bowtie).

Much better closeup of the balun,



Note that the left pair doesn't need to go through ferrite beads (it's wired ground to ground, no common mode voltage ideally), it just looks nicer this way.

As for bicons -- you'll want one that's constructed right to begin with, and preferably that has some kind of calibration.  I've heard reasonable deals (like under $1000) can be found on eBay if you keep looking around a while?

Tim

[tautech]

Self-calibrated far field measurements are possible but time consuming.  I did it by building reference antennas to generate calibrated far field intensities to then calibrate wideband directional antennas which made the actual measurements at a relatively RF quiet and flat location.

The better option if you can afford it is to treat the near field measurements as relative so when the first EMC compliance test fails, you know how much they need to be reduced.  And before the first expensive EMC compliance test, use near field measurements to evaluate basic good EMC practices.

This ! ^^
It really doesn't need to be anymore complex, the KISS principle need always be tried first. 

[T3sl4co1l]

Self-calibrated far field measurements are possible but time consuming.  I did it by building reference antennas to generate calibrated far field intensities to then calibrate wideband directional antennas which made the actual measurements at a relatively RF quiet and flat location.

The better option if you can afford it is to treat the near field measurements as relative so when the first EMC compliance test fails, you know how much they need to be reduced.  And before the first expensive EMC compliance test, use near field measurements to evaluate basic good EMC practices.

This ! ^^
It really doesn't need to be anymore complex, the KISS principle need always be tried first. 

And to add to that -- there are design practices you can follow to facilitate testing and early passing.  The most obvious case, throw everything inside a solid metal box, and filter all connections passing through it.  You can connect these to a CDN or LISN, to measure the RF coming out directly, or going in (if you have a generator and amplifier) to do immunity testing as well.  In short: you can approximate these connections as zero-dimensional connections -- ports -- and do conducted testing to much higher frequencies than is justified at the end of a long power cable.

This extends to the board level, where for example, a ground plane design allows you to know, at a glance, where the image currents are flowing, and what trace impedances are (and therefore equivalent capacitance and inductance).

Such a design may not be the cheapest possible approach (it may use more space between components and routing, it may use more layers, or it may use more metal, including the PCB, shields, enclosure, etc.), but it reduces extremely costly lab time, and time-to-market.

Tim

[tautech]

From another thread:

I too am looking for some EMI probes as well.

I've seen some low-cost PCB material probes on ali express.
     - https://www.aliexpress.com/item/1PC-EMC-EMI-Near-Field-Probe-Conducted-Radiation-Correction-Simple-Magnetic-Field-Probe-9KHz-6GHz-Type/32966266773.html

$40 for a set of 5 
How could you go wrong ?

[cpuerror]

I'd really like to see someone just come up with a design for this sort of antenna, and have it 3d printable. The elements can be painted with conductive spray paint to make it effective.

[coppercone2]

thanks j, another project

looks  very home build able even on a simple drill press so long you have reamers


lots of filing and grinding for the base pieces though if you don't have a lathe. 

[Jay_Diddy_B]

thanks j, another project

looks  very home build able even on a simple drill press so long you have reamers


lots of filing and grinding for the base pieces though if you don't have a lathe. 

Hi Coppercone2 and the group,

I am going to share some pictures that show the construction of my biconical antenna. I am not going to provide dimensions for the parts, they can be found in the document attached to this message:

https://www.eevblog.com/forum/testgear/diy-biconical-antenna-30-300mhz/msg2257035/#msg2257035


Overview of the completed antenna



This picture shows the complete antenna assembled and ready for use.

Center Structure



This is the center rod assembly. Two of these are required.



Cone 1



This is the inner cone. It has a section of threaded rod, to screw into the center assembly. The cone is held on the rod with a 10-32 set screw.


Cone 2



This is the outer cone. It is also held in place with a set screw.



Antenna Elements



The construction of these differs from MIL-STD-461A, in that they were bent from one piece of rod. The MIL-STD uses to straight rods and coupler. I cut these to length after bending.



Center Connector



The housing for the center connector is waterproof electrical conduit. I made aluminum inserts that the two antennas screw into.


Center Insulator open



This picture shows how the thin coaxial cable is attached.

N connector



The N connector is also mounted in a plastic conduit fitting.


Balun - Not Shown

There is a balun consisting of ferrite beads threaded on the coax inside the support arm.


The aluminum is all 6061 purchased from the Metal Supermarket.
The plastic conduit fittings were purchased from the Home Depot (or similar store).

Part were machine on a 8x14 lathe and a mini-mill.

Regards,

Jay_Diddy_B

[coppercone2]

I am thinking about casting those parts

[coppercone2]

I noticed it looks like you decided to go for a bend rather then make the elbow pieces . I was thinking about this myself

why do you think they went with them? higher precision then bent aluminum?

[Jay_Diddy_B]

Hi,

I used a pipe bender like this one:



Very cheap from Harbor Freight:

https://www.harborfreight.com/tubing-bender-3755.html


Jay_Diddy_B

[Jay_Diddy_B]

Hi,

I should add, although I have built this antenna, I don't find it very useful.

Jay_Diddy_B