Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: TW-2010 vs MFJ-1786  (Read 6463 times)
KK6GMN
Member

Posts: 132




Ignore
« on: September 12, 2013, 02:11:53 PM »

I don't want to get into a discussion about how bad these antennas are and why am I not putting up something else.  Thanks for understanding.

I have a special situation and am looking at these two antennas to solve the issue.  They both have merits for my situation.  I have limited mobility and cannot easily mount an antenna on my roof.  I have a VERY small backyard that is mostly pool, however I could place the TW-2010 in a corner of my yard and the MFJ-1786 could easily be placed on top of a 15' pole on my fence. Both are fairly close in price.

If you had to choose between the two, which would you choose and why. 

Thanks!
Logged

-SeanM
KK6GMN

"No man is a failure...
...who has friends." --Clarence

Weather at my shack
http://www.pegnsean.net/~sean/weather/wx.htm
JAHAM2BE
Member

Posts: 276


WWW

Ignore
« Reply #1 on: September 12, 2013, 06:37:11 PM »

In your situation, I would choose the MFJ-1786 as the better all-around antenna, for the following reasons:

- Considering your mounting possibilities - the TW-2010 near ground, or the MFJ-1786 up higher on a 15' pole - the higher antenna will probably work better (assuming it is in the clear).
- The MFJ-1786 covers 30 meters, which the TW-2010 does not.
- The MFJ-1786 is continuously tunable from 10-30 MHz, so you can use it for shortwave listening as well.
- The radiation pattern of the MFJ-1786 includes both high-angle and low-angle radiation, making it suitable for both local and DX work.

On the other hand, the advantages of the TW-2010 include:
- Bandwidth: The TW-2010 does not need retuning within a band, while the MFJ-1786 will be very narrow-banded and will frequent retuning as you change frequencies within a band.
- Portable operation: The TW-2010 can be folded up and transported compactly, unlike the MFJ-1786.
- DX signals: The radiation pattern of the TW-2010 (a vertical dipole) will tend to weaken close-in signals and emphasize DX signals coming in at low radiation angles. This can allow you to better hear the DX signals. Also, if operated near the ocean and at the water's edge, a vertical dipole can have good gain at low angles.

Efficiency-wise, I would guess the two antennas are pretty close. The TW-2010, being a physically larger antenna with straight-line current paths, has the theoretical potential for higher efficiency, but based on the size and placement of the coils in the TW-2010, I don't think it will be more efficient than the MFJ-1786.

One other thing to be aware of is that the loop, if mounted vertically, will have two deep nulls in its radiation pattern, where RF energy will be poorly radiated and poorly received. The vertical dipole will be omnidirectional.
Logged

KK6GMN
Member

Posts: 132




Ignore
« Reply #2 on: September 12, 2013, 07:57:40 PM »

Excellent response.  It points out the things I did not know about both antennas which is what I was looking for.

Thanks!
Logged

-SeanM
KK6GMN

"No man is a failure...
...who has friends." --Clarence

Weather at my shack
http://www.pegnsean.net/~sean/weather/wx.htm
HFHAM2
Member

Posts: 31




Ignore
« Reply #3 on: September 13, 2013, 11:20:27 PM »

The MFJ 1788, although more expensive, will give you 40 meters thru 15 meters coverage and works reasonably well on 40 meters (contrary to some reports). 40 meters CW is my favorite band so that's why I went with the 1788 this time, but I've had great results with 1786s in the past (30 thru 10).

For best DX (although you'll "overshoot" closer-in contacts), MFJ recommends that you mount your HI-Q loop antenna horizontally and at a minimum of 23 feet above ground.

Alternatively, if you mount it vertically, height above ground doesn't appear to make any appreciable difference as it then radiates both low and high take-off angle waves; it's also somewhat directional when mounted vertically. I have my MFJ 1788 mounted on a single six foot mast section supported by a tripod and can routinely work out to 3000 miles plus on 20 and 15 meters (50-100 watts CW); I don't use SSB much so can't really comment).
Logged
WX7G
Member

Posts: 5917




Ignore
« Reply #4 on: September 14, 2013, 02:44:06 AM »

The TW-2010 is the clear winner

I have done extensive antenna modeling and have constructed several short dipoles and monopoles. Today I modeled the TW-2010 and the MFJ-1786 with the bottom of each 3' above average ground. Adding loading coil loss until the specified 14 MHz, 200 kHz, 1.5:1 VSWR bandwidth was achieved the TW-2010 has a radiation efficiency of 50%. This is comparable to the MFJ-1786.

Where the real difference shows up is in the radiation patterns over average ground. The TW-2010 has a classical vertical dipole pattern and is omni-directional with good gain near the horizon. The MFJ-1786 mounted vertically has maximum radiation straight up with the gain decreasing at lower angles. The azimuth pattern shows a dipole pattern with maximum gain end-fire. Broadside, at a take-off-angle of 10 degrees (DX), the gain of these two antennas differs by 15 dB with the TW-2010 being better. End-fire the difference in gain is 5 dB.

I would go with the TW-2010 for these reasons:
1) 5-15 dB greater gain at a take-off-angle of 10 degrees (more at lower angles)
2) Near-instantaneous band changes
3) Wide instantaneous bandwidth
4) Handles 1200 watts PEP vs 150 watts



« Last Edit: September 14, 2013, 02:52:53 AM by WX7G » Logged
WX7G
Member

Posts: 5917




Ignore
« Reply #5 on: September 14, 2013, 10:41:17 AM »



Here is a comparison of the TW2010 and the MFJ-1786 end-fire (maximum gain). The BLUE trace is the TW2010. Both antennas are modeled over "average ground" with the bottom of the TW201 3' high and the MFJ-1786 15' high. Resistive loss was added to the TW2010 to obtain the specified 200 kHz, 1.5:1 VSWR bandwidth. No loss was added to the MFJ-1786.
« Last Edit: September 14, 2013, 10:49:40 AM by WX7G » Logged
KK6GMN
Member

Posts: 132




Ignore
« Reply #6 on: September 14, 2013, 12:45:58 PM »

Very interesting stuff there!  Thanks for the information.
Logged

-SeanM
KK6GMN

"No man is a failure...
...who has friends." --Clarence

Weather at my shack
http://www.pegnsean.net/~sean/weather/wx.htm
JAHAM2BE
Member

Posts: 276


WWW

Ignore
« Reply #7 on: September 14, 2013, 06:57:36 PM »

Here is a comparison of the TW2010 and the MFJ-1786 end-fire (maximum gain). The BLUE trace is the TW2010. Both antennas are modeled over "average ground" with the bottom of the TW201 3' high and the MFJ-1786 15' high. Resistive loss was added to the TW2010 to obtain the specified 200 kHz, 1.5:1 VSWR bandwidth. No loss was added to the MFJ-1786.

I ran a model of the TW2010 and was not quite able to reproduce the high gain you achieved. Here are my model and results, which are slightly worse than yours:

Model (created in 4nec2):
Code:
CM
CE
SY h=36
SY L=5.476e-6
GW 1 13 0 0 h 0 0 h+94 0.625
GW 2 13 -34 0 h 0 0 h 0.625
GW 3 13 0 0 h 34 0 h 0.625
GW 4 13 -34 0 h+94 0 0 h+94 0.625
GW 5 13 0 0 h+94 34 0 h+94 0.625
GS 0 0 0.0254
GE 1
LD 5 1 1 13 24900000
LD 5 2 1 13 24900000
LD 5 3 1 13 24900000
LD 5 4 1 13 24900000
LD 5 5 1 13 24900000
LD 0 1 7 7 0 L
GN 2 0 0 0 13 0.005
EK
EX 0 1 7 0 1.0 0 0
FR 0 0 0 0 14.05 0
EN

Model parameters are as follows.

Dimensions: Using page 8 of http://www.twantennas.com/resources/TW2010_Trans_World_manual_v1.3.pdf as a guide, I generously estimated each capacity hat spoke's length at 34 inches. I estimated the total length of the vertical portion to be the stated length of the middle section (30 inches) plus the length vertical bars in the top and bottom sections, which I generously estimated as having a combined total length of 64 inches. Conductor diameter is assumed to be 1.25 inches (0.625-inch radius) based on the minimum dimension of the permanent mounting assembly.

Loss resistance: Conductor material is modeled as aluminum material "Alu-T6". Inductor loss is very generously assumed to be zero.

Resonating inductance: I used a single resonating inductance at the dipole feedpoint. Its value, calculated by the optimizer for zero reactance at 14.05 MHz, is 5.48 uH.

Height and ground: The bottom of the antenna is placed 36 inches above average ground.

Results:


The maximum gain is -1.2 dBi, and this is assuming zero coil loss. If we add in realistic coil loss assuming Q=500, maximum gain drops slightly to -1.5 dBi. If we assume coil Q=200, maximum gain drops to -1.85 dBi.

In any event, I am not able to get the 0 dBi gain you achieved. The difference with your results is small, but puzzling. Do you have any idea what could explain the discrepancy? Perhaps it's within the margin of error for these sorts of simulations?

Also, it seems the TW2010 manual does not mention the use of a balun, and recommends a minimum coax length. Could the feedline be radiating?

Regarding the small loop, I'm also rather surprised at the dismal simulation results you achieved. For comparison, here are my results simulating a square vertical loop, 0.9m per side, constructed from a 1cm-radius aluminum conductor, placed 5m above average ground:



The model for this antenna is as follows:
Code:
CM
CE
SY C=3.98e-11
GW 1 9 0 -0.45 6 0 0.45 6 0.01
GW 2 9 0 -0.45 6 0 -0.45 5 0.01
GW 3 9 0 -0.45 5 0 0.45 5 0.01
GW 4 9 0 0.45 5 0 0.45 6 0.01
GE -1
LD 0 3 5 5 0 0 C
LD 5 1 1 9 24900000
LD 5 2 1 9 24900000
LD 5 3 1 9 24900000
LD 5 4 1 9 24900000
GN 2 0 0 0 13 0.005
EK
EX 0 1 5 0 1 0 0
FR 0 0 0 0 14.05 0
EN

In my simulation, the maximum gain of the loop (assuming no losses other than conductor losses) is in fact greater than that of the vertical dipole.

The MFJ-1786 is round, not square, but I would expect its radiation pattern and gain to be in the ballpark of the square loop.

Comments?
« Last Edit: September 14, 2013, 07:50:26 PM by JAHAM2BE » Logged

KD2CJJ
Member

Posts: 368




Ignore
« Reply #8 on: September 15, 2013, 05:14:57 AM »

Let me comment on the 1786...I can't comment on the tw antenna.

The 1786 antenna is truly a compromise antenna compared to anything else in my opinion.  This includes a ground mounted verticle dipole.  My father ran one for a few months due to living Ina hoa and sold it after much frustration.

1.  You will need to retune every 10 to 20khz.  If you like to jump around then this is not for you.  Tuning at first is very tedious and has a learning curve.

2.  It is very very sensitive to its surroundings and feedline quality.  Any metal structures, fence, shed, siding, etc. will detune the antenna and you will become frustrated.  You need touse high quality feedline or again tuning will become a nightmare.  Other antennas are more resilient to lower quality feedlines.

3.  SSB its performance is no where near as good as cw.  You will not achieve the same distances.

4.  Mfj quality is horrible.  It took him 3 antennas before he got one that worked.  He still sent it back due to the above points.

For these reasons he put up an elevated Tarheel.  I do not recommend this unless you can get it up 20 feet or so and have elevated radials.  If the tw antenna performs as well or better than the tarheel then I would go with that antenna for the reasons above.  I'm not sure about reliability though.
Logged

73

Mike
KD2CJJ
WX7G
Member

Posts: 5917




Ignore
« Reply #9 on: September 15, 2013, 05:08:16 PM »

JAHAM2BE, thank you for pointing out the error in my simulation. Originally I did wonder where all the gain went and spent 10 minutes checking things and trying a few things but didn't catch the problem. After you posted your simulation I looked some more and found the problem. The problem turns out to be that I resonated the loop with a series capacitor at the top. For some reason this corrupts the simulation and without the capacitor the gain appears more realistic. I'll post a correct simulation that we can compare to yours.
Logged
JAHAM2BE
Member

Posts: 276


WWW

Ignore
« Reply #10 on: September 15, 2013, 05:48:08 PM »

The problem turns out to be that I resonated the loop with a series capacitor at the top. For some reason this corrupts the simulation

That's a rather interesting failure mode (that I've not experienced, though I use 4nec2 not EZNEC). Does an Average Gain Test catch the problem with the model?
Logged

WX7G
Member

Posts: 5917




Ignore
« Reply #11 on: September 15, 2013, 07:22:07 PM »

I'll check. Can you see if 4NEC2 is corrupted by adding a resonating capacitor?
Logged
JAHAM2BE
Member

Posts: 276


WWW

Ignore
« Reply #12 on: September 15, 2013, 08:15:27 PM »

I'll check. Can you see if 4NEC2 is corrupted by adding a resonating capacitor?

The line "LD   0   3   5   5   0   0   C" in my small loop model is the resonating capacitance, located at the bottom of the square loop (opposite of the loop feedpoint/excitation), with C being calculated by the optimizer to be 39.8 pF.

Running an AGT on that model in 4nec2 gives 1.010 (0.04 dB) as the result, which seems acceptable.

Interestingly, swapping the location of the feedpoint and the capacitor, so that the capacitor is on top and the feedpoint on bottom, significantly changes the pattern and improves low-angle radiation:

                           

AGT results with capacitor at the top: 1.012 (0.05 dB). Efficiency with the capacitor at the top is 52.2% while with the capacitor at the bottom it is 61.49%, which is opposite of what I would expect.

This all is a bit surprising to me. The pattern difference was also observed when running the AGT (where wire loss and ground loss are set to zero), and also when using all ground types (fast, perfect, real, MiniNec). No pattern difference was observed when simulating in free space. Therefore, it may be the interaction of the electric field around the capacitor with the ground that is causing the difference in pattern. The size of the loop is 0.9m per side, so a height difference of 0.9m (0.04 wavelengths) of the capacitor placement above ground is enough to cause the difference in pattern. (NOTE: merely raising the entire structure height of the capacitor-at-bottom loop by 0.9 meters, so that the capacitor is still at the bottom of the loop but is as the same physical height as the capacitor-at-top loop, still results in essentiallly the same difference in pattern between the two configurations, so it's not just the physical height, but the relative position of the capacitor within the loop that appears significant.)

It would be interesting to run calculations of the near field to understand more clearly what is going on with the capacitor placement (and to run similar tests on loops of different shapes and sizes), but I'll save that for another day.

-----

EDIT: The phi for the 2D slices is not consistent among all my graphs, so here are 3D plots of the far field, which are perhaps easier grasp in their entirety.

Capacitor at bottom:


Capacitor at top:
« Last Edit: September 15, 2013, 09:15:11 PM by JAHAM2BE » Logged

VA2PBJ
Member

Posts: 160




Ignore
« Reply #13 on: September 15, 2013, 09:35:13 PM »

How much does the TW2010 rely on loading the coax? I looked at several of their manuals and the all suggest at least of 65 ft of coax. Would this thing cease to work with a 1:1 balun on the line?
Logged

7 3 Peter VA2PBJ
JAHAM2BE
Member

Posts: 276


WWW

Ignore
« Reply #14 on: September 15, 2013, 09:51:53 PM »

How much does the TW2010 rely on loading the coax? I looked at several of their manuals and the all suggest at least of 65 ft of coax. Would this thing cease to work with a 1:1 balun on the line?

I'm guessing (based both on the simulation results from this thread and on my experience constructing a similarly short vertical dipole) that the TW2010 should work fine with a balun on the line. Maybe they decided to eliminate a balun as a cost-saving measure.

It's possible to model the outer surface of the coaxial cable shield in NEC-based programs to evaluate current on the feedline. I have limited experience doing this, but in my limited experience, I did notice that with the coax running away at a 45-degree angle from a vertical dipole (as the TW2010 manual suggests) there was noticeable current on the coax shield. Ideally, in addition to the use of a balun at the feedpoint, the coax should be routed away from the vertical dipole at a 90-degree angle for as far as possible, to prevent the antenna from coupling into the coax. But in practice this is difficult mechanically, so again, I'm assuming that the TW2010 manufacturer chose to recommend the compromise solution of a 45-degree angle plus a minimum coax length (whose purpose, to be honest, is not clear to me).
Logged

Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!