Re: MFJ-868 SWR/Wattmeter

On 8/28/2011 11:33 AM, Jeff Liebermann wrote:
On Sun, 28 Aug 2011 10:14:06 -0400, Renee wrote:

On Sat, 27 Aug 2011 21:22:32 -0700, Jeff Liebermann<jeffl@xxxxxxxxxx> wrote:

On Sat, 27 Aug 2011 21:40:29 -0500, tom<news4792@xxxxxxxxxx> wrote:

The majority of what they sell does exactly what it's advertised to do.
You have to read the specs.

Generally true. However, I spent some time modeling the MFJ-1800
2.4GHz antenna and came the conclusion that it's not even close to 50
ohms and more like 300 ohms. I guess they forgot to include a balun:
<http://www.mfjenterprises.com/Product.php?productid=MFJ-1800>
<http://802.11junk.com/jeffl/antennas/mfj1800/index.html>
While one screwup is not sufficient to claim that the entire product
line is lacking, it does make me very suspicious and wary.

I'm curious, did you do any measurement and testing with an analyzer, etc., or
are all the figures simply calculated estimates?

There are no affordable "analyzers" that work at 2.4Ghz. I also have
not owned, or plan to own an MFJ-1800 antenna. The physical
measurements were done by a friend that owns this antenna and was
wondering why it didn't provide anywhere near the gain promised (as
compared to a commercial panel antenna).


Hmmm.. while you might not be able to buy an analyzer, that doesn't mean that you couldn't have done the measurement by other means. A power meter (perhaps an eval board for one of the analog devices parts)? Maybe even using a WiFi card as a field strength meter at a little distance away, and then using an adjustable stub tuner (e.g. microstripline on a PC board. (or one of those "check your microwave oven for leakage" meters from radio shack)

Sure, it's not as nice as that VNA from Anritsu or Agilent, but people have been making microwave impedance measurements for decades with pretty primitive gear.

Building a directional coupler in microstripline is a cookbook endeavor. An hour with some copper clad board and a razor blade would do. If you happen to have some copper foil tape around, it's even easier.

An impedance measuring system along the lines of the "3 meter" technique (where you have a known L or C in series) is a possibility.

Not to mention that there are a lot of folks on this list who DO have access to suitable analyzers.

So Renee's question was pretty reasonable.



Incidentally, I recently purchased a pair of totally misdesigned
2.4Ghz yagi's on eBay for about $8/ea.
<http://802.11junk.com/jeffl/crud/wi-fi-yagi-that-sucks.jpg>
Note the crude driven element, exposed coax conductors, and lack of a
balun. What happened was that this was a cheap clone of another
similar antenna, which did provide all the required matching
circuitry. This vendor decided it was too much trouble to copy the
matching network and balun, so he just left it out. It has more gain
to the sides of the yagi, than in front.


Fascinating... especially since the matching network wouldn't normally have any effect on the pattern. Balun/Choke having an effect, I could see (coupling to the feedline screws up the pattern).





Has anyone confirmed what the piece is that connects the driven element with the
coax connector that looks to be a short piece of coax covered in heat shrink?
Has anything been found or measured hiding under the heat shrink?

There's a ferrite bead under the shrink tube. If you can find a
ferrite material that actually works at 2.4GHz, it might pretend to do
something useful. However, no way is it going to provide a 4:1
impedance transformation.

2.4 GHz is 12 cm wavelength

4-6cm of transmission line is a good fraction of a wavelength. How do you know it's not 3/4 lambda of 75 or 92 ohm line, or something like that. (1/4 wavelength of 75 ohms would transform 200 ohms to about 30 ohms, for instance)

I've worked with enough UHF and microwave antennas with weird funky feedpoint things to not trust in first impressions. There could be all sorts of parasitics that help the situation out, and they're tough to model accurately (you're certainly not going to do it with NEC, for instance.. I'd trust NEC for the feedpoint impedance without matching/balun/stubs..)


Yeah, it's probably not a nice 1:1 match over the entire band, but then, most antennas aren't. It just has to be "good enough" and in that business, the loss in the feedline probably dominates anyway. Figure 3 dB of loss in the feedline, and you get a 6B improvement in return loss.


The model you have shows a beamwidth around 30 degrees (H) and 28 degrees (V).. that's a directivity around 16 dBi

4NEC2 is giving you 14dB gain.. that's moderately consistent, although 2 dB is something I'd want to figure why the difference. I'd want to figure out why 2dB different, especially since you didn't put resistive loading into your model, which, at 2.4GHz *will* make the match better

MFJ claims 15dBi, and that is also consistent with your model and the back of the envelope from the pattern.



I realize we can all assume the antenna impedance is 50 ohms, but is it actually
stated in any of the specifications from MFJ? I didn't see any mention of
impedance specification at all for this antenna, but I could only find
information on the web page at the link you provided as well as in the pdf file
of the instructions for the MFJ-1800. The instructions are rather sparse.

Good point. One can assume that if it uses a 50 ohm coax pigtail, a
50 ohm N connector, and is designed to connect to a wireless device
that presents 50 ohms, it just might need to be a 50 ohm antenna.


Nope.. not if the feedline is a couple dB loss (which wouldn't be unusual for a laptop pigtail using RG-174 or smaller).. All they really care about in this kind of application is the directivity.

After all, how many ham antennas for HF have a PL-259/SO-239 "UHF" connector on them. I wouldn't take the fact that it has a 50 ohm connector as indicating anything other than it's a common connector that people are likely to have mating connectors for.

75 ohm BNC, Ns, etc exist, but how many people use them on dipoles (which have 72 ohm impedance)... nahh. they just use a 50 ohm cable, UHF connectors, and are done with it.

If they used 75 ohm coax on this antenna, and it's close to a multiple of 1/4 wavelength, then the match at the N connector is probably not a lot worse than using 50 ohm coax on a dipole.
.

Relevant Pages

  • Re: VHF antenna placement question
    ... antenna as far away from anything else up there that's metal. ... Coax cable is made for smooth radius bends, ... Put another PL-259 connector on the bottom or inside the removable mast ... broken cables. ...
    (rec.boats.electronics)
  • Re: Low cost coax connectors
    ... You might have to modify the plug and socket a bit, ... power jack as an antenna connector does present another problem. ... requires a "unique" coax connector to prevent users from substituting ...
    (sci.electronics.design)
  • Re: Low cost coax connectors
    ... into the plug and vaporizes your radio. ... power jack as an antenna connector does present another problem. ... power jack as a coax connector would certainly not qualify as unique ...
    (sci.electronics.design)
  • Re: Low cost coax connectors
    ... You might have to modify the plug and socket a bit, ... power jack as an antenna connector does present another problem. ... requires a "unique" coax connector to prevent users from substituting ...
    (sci.electronics.design)
  • Re: Low cost coax connectors
    ... You might have to modify the plug and socket a bit, ... power jack as an antenna connector does present another problem. ... requires a "unique" coax connector to prevent users from substituting ...
    (sci.electronics.design)