4nec2 & Other Antenna Design Modeling Software - Page 84 - Canadian TV, Computing and Home Theatre Forums

 3Likes
Reply
 
LinkBack Thread Tools Search this Thread Display Modes

post #1246 of 1274 (permalink) Old 2017-11-25, 04:04 AM
Veteran
 
Join Date: Nov 2008
Location: San Diego, CA
Posts: 3,395
Ken Nist posted eznec (*.ez) type files at top of HDTVPrimer Antenna Chart Page....I have only converted (and FIXED) a small number of these, mostly so I could add to Comparison Charts or explore modifications:
www.hdtvprimer.com - /SIMS/
http://www.hdtvprimer.com/antennas/comparing.html

4nec2 will Automatically Input and Convert *.ez into *.nec Files so you can view the dimensions. However, most needed FIXES to actually RUN under 4nec2.

I have Analyzed a number of Old/Antique Antennas, based on User Posted measurements...but NOT C-M Crossfire, C-M Advantage nor any of the A-C and W-G Combo's....but would gladly add more if you dig up sufficient info:
Log-Yagi & Dual-Driven Antennas - ImageEvent
VHF LPDA - ImageEvent

Antenna Simulations, Overload Calculations, etc: http://imageevent.com/holl_ands
holl_ands is offline  
Sponsored Links
Advertisement
 
post #1247 of 1274 (permalink) Old 2017-12-30, 09:01 AM
Rookie
 
Join Date: Nov 2017
Posts: 29
Quote:
Originally Posted by rochmndx View Post
Is anyone interested in modeling/building the old Channel Master Quantum VHF antennas, like the 1110?
yes- albeit sub-zero oitside, it is never cold enough for my favorite hobby
Otaobsession is offline  
post #1248 of 1274 (permalink) Old 2017-12-31, 04:54 AM
Veteran
 
Join Date: Nov 2008
Location: San Diego, CA
Posts: 3,395
FYI: CM1111 4nec2 Model (with Dimensions) and Performance Charts posted here....sorry that's the only one for which we have measurements:
https://imageevent.com/holl_ands/zig...uantum23ellpda

Antenna Simulations, Overload Calculations, etc: http://imageevent.com/holl_ands
holl_ands is offline  
 
post #1249 of 1274 (permalink) Old 2019-03-16, 03:43 PM
 
Join Date: Jun 2016
Posts: 497
Hi holl_ands, didn't know where else to ask this.

As you know I have the FF4 antenna and am very happy with it my issue is with the VHF performance.

Previously I was using a Panasonic Plasma tv that only indicated where a signal was available or not, neither 12 or 7 showed as available. What I did was buy a HDHomerun that didn't perform as well as my tv's tuner but did indicate there was a signal coming in most of the time on channel 12, around 6dB ss. Liked the features on the HDHomerun but because the tuner didn't perform as well as my tv's I ebayed it and tried a Homeworx tuner, which was god awful and now resides in the local landfill.

My new Samsung nu8000 has a signal meter and tells me rf 12 and rf 7 are coming in at about 12-14dB snr (still not visible, or sometimes visible only to die a second later).

My question, would a simple compact VHF antenna give me the needed boost to receive these, I was looking at this model, with the swept element.

https://imageevent.com/holl_ands/yagis/k6sti

Thanks in advance.

224 transmitters
Zank Frappa is offline  
post #1250 of 1274 (permalink) Old 2019-03-16, 05:25 PM
Veteran
 
Join Date: Nov 2008
Location: San Diego, CA
Posts: 3,395
I don't know/remember which FF4 version you built, whether 40-in or 48-in High Reflector. Either model indicates that it provides a very healthy 8+ dBi across Hi-VHF Band [rising to as much as 10+ dBi on Ch7 in the Taller Model]....but with miserable SWR:
https://imageevent.com/holl_ands/mul...dblanglreflopt

First of all, I would recommend that you TRY adding various length Coax Extensions....the intention is to MOVE the SWR NULL Pattern along the Coax so that it does NOT occur on the END of the Coax at the desired Frequencies (NULLS occur every Half-Wavelength, which is 33.4-in on Ch7 and 28.5-in on Ch12). So adding a "standard" length coax of say 12-in and/or 18-in should CHANGE the NULL Pattern.

Another "simple" mod that you could try to see if 8 dBi Gain is "enough" would be to move the Reflector further away from the Bowties....such as 6, 10 and perhaps as much as 16-in away to Optimize for Hi-VHF performance at expense of UHF.

=======================================
Instead of a Hi-VHF Band Yagi, you might want to consider one of the alternative "Hourglass-Loops" analyzed here, where you can choose how many Reflector Rods you want to build. Since Wire Size is NOT important, you COULD use AWG12 or larger....incl. the Metal Shield Wires in RG-58 or RG-59 Coax...with NO NEED to remove the Insulation:
https://imageevent.com/holl_ands/loo...fhourglassloop

If you find that you "need" even more Hi-VHF Gain, the fol. Horizontal Stack of UHF Hourglass-Loops (3x Bigger for Hi-VHF) illustrates what can be done that MAY provide 10+ db of Space Diversity Combining Gain [depending on severity of Multipath Fading] in addition to nominal 3 dB Summing Gain:
https://imageevent.com/holl_ands/sta...rglasslooprefl

=======================================
Obviously the K6STI 5-El Yagi will have low SWR in Hi-VHF Band....but it only provides Gain = 7.3 dBi on Ch7 and roughly same Gain as FF4 on Ch12. So it would be useful to know in advance whether 7.2 dBi would be "enough" before you cut and fabricate.....

So if the "SWR FIX" doesn't work, you MIGHT want to consider building a HIGHER Gain Yagi [if a narrower Beamwidth is suitable for your location], such as:
https://imageevent.com/holl_ands/yag...elfdyagi2rropt

And if you eventually find THAT isn't quite enough....build a PAIR, stacked either Vertically or Horizontally, which MAY provide 10+ db of Space Diversity Combining Gain [depending on severity of Multipath Fading] in addition to nominal 3 dB Summing Gain.
If you decide which two Antennas you want to stack, I can calculate a HVH or HHH to interconnect them.

FYI: The only V-Stk Hi-VHF Pair I've Optimized thus far is a Pair of simple CIRCULAR LOOPS with Reflectors...which is also quite good [Screen Reflectors could be Aluminum Foil on Cardboard if in Attic....or Re-Optimized to use Reflector Rods]:
https://imageevent.com/holl_ands/sta...lystackedhivhf

=======================================
BTW: I've been doing some preliminary looks at alternatives for an FF4/FF6 with improved Hi-VHF Performance [at expense of UHF Performance], but thus far the only examples I have ready are as follows:
https://imageevent.com/holl_ands/mul...sreflectorrods [All SAME, NOT FF]
https://imageevent.com/holl_ands/loo...lh410x14norefl [All SAME, NOT FF]

I'm going to go ahead and do some Optimization Runs with a combination of these ideas: FF4 (instead of M4) with two Rectangular Hi-VHF "Resonator" Loops....and then add the two LAYERS of Reflector Rods (same or similar to SBGH Reflector Rod's). PS: This was a Re-Optimized spin-off of alternatives that could be HIDDEN behind a user's 3-ft H x 2-ft W Metal Picture Frame....this fully Optimized alternative assumed two smaller Metal Picture Frames....

Antenna Simulations, Overload Calculations, etc: http://imageevent.com/holl_ands

Last edited by holl_ands; 2019-03-16 at 06:18 PM.
holl_ands is offline  
post #1251 of 1274 (permalink) Old 2019-03-16, 06:47 PM
 
Join Date: Jun 2016
Posts: 497
Wow, thanks for that VERY detailed answer.

I guess I was way off base with regards to SWR, I thought the higher the value the more distorted the gain would be. I measured my coax run and it's 110", my FF4 is the one with the 48" reflector and my reflector is 1"x 1".

I can easily change the spacing of bowties to reflector.

One of the things I have that works against me is I'm in an apartment about 180 feet up and they don't want things extending out far from the railing, I also only have about 54" of height to work with.

Just so you know I am so pleased with the FF4, I get 2 buffalo channels that no one in my area gets, channel 2 (rf33) at -8.2dB NM and channel 7 (rf38), which is 109 miles away 24/7. TV fool doesn't have accurate data for that channel (they show it south of the equator near Nigeria, LOL) but I assume it's got an even larger NM figure based on power and location to channel 2.

The rf7 channel shows as -8.0NM on TV fool, I foolishly thought improving SWR would improve reception.

Given my constraints I'll mess around with the reflector or look at building one of these:

https://imageevent.com/holl_ands/yag...hf6elfdyagiopt

Or just by the stellar labs short boomed VHF antenna and combine it with my FF4.


In that situation is there a minimum distance they have to be from one another using a VHF/UHF combiner?

224 transmitters
Zank Frappa is offline  
post #1252 of 1274 (permalink) Old 2019-03-17, 05:15 AM
Veteran
 
Join Date: Nov 2008
Location: San Diego, CA
Posts: 3,395
For Verticallly Stacked Antennas in Different Bands, Gain isn't degraded until they are about 1-ft apart (Metal-to-Metal)....however, F/B and F/R Ratio start to be degraded when about 4-ft apart....so for "best" performance I recommend 6-8 ft apart [if you had enough room......].
https://imageevent.com/holl_ands/stacked

I haven't done any similar sensitivity studies for Horizontal Stacking....but I would guess we would find similar guidelines.

Antenna Simulations, Overload Calculations, etc: http://imageevent.com/holl_ands
holl_ands is offline  
post #1253 of 1274 (permalink) Old 2019-03-24, 01:08 PM
Veteran
 
Join Date: Jan 2009
Posts: 4,652
nikiml's eval / opt scripts in linux

ever since upgrading my debian based laptop running mint to latest & greatest Mint,
broke the installation of the 'nec2' engine required to get nikiml's scripts going on linux. So was dead in the water for a few months.
Even downloading / installing the last version of 'nec2' available as a .deb package from the debian archives wasn't working.
https://packages.debian.org/jessie/libf2c2
https://packages.debian.org/jessie/nec

Took me half the nite yesterday to get this figured out, but was able to pull the source from the old Debian Jessie repo, and compile my own packages for 'libf2c2', 'libf2c2-dev', and 'nec2', which had to be done the "debian way" which is essentially
'dpkg-buildpackage -rfakeroot'
to get the Debian based Mint machine working again. Which is Cool, but...
That didn't help the Arch Linux machine, since Arch is a rolling distro, and all packages are always 'up to date', they aren't gonna maintain anything like 'nec2'. They only have the more modern 'nec2c', or 'necpp' (which are essentially refactored c translations of nec2 fortran code to c and c++ incompatible with nikml's scripts) packages avail. Turns out there is a way to convert working .deb packages ('debtap') to Arch .pkg.xz that ya can install with pacman. Worked like a champ.

If anyone runs in to a similar issue, I have copied these working packages to my google drive for safe keeping. Just drop me a line here...

Snapper running nikiml's scripts on the Arch machine:

DB8E/VHF Yagi rotor FM Bandstop ap-8700 preamp 8way split LG lcd.
majortom is offline  
post #1254 of 1274 (permalink) Old 2019-03-24, 07:25 PM
Veteran
 
Join Date: Nov 2008
Location: San Diego, CA
Posts: 3,395
Q: Does the the linux package take advantage of Multi-Processor capability???

I would be interested in how it compares to fol. BENCHMARKS [incl Single vs Multi-Processor]:
https://imageevent.com/holl_ands/files/4nec2benchmarks

Antenna Simulations, Overload Calculations, etc: http://imageevent.com/holl_ands
holl_ands is offline  
post #1255 of 1274 (permalink) Old 2019-03-24, 08:03 PM
Veteran
 
Join Date: Jan 2009
Posts: 4,652
I doubt it, as is. Though I believe one can add / link nec2c and / or necpp against the 'ATLAS' library built with 'lapack', which would (from what I read yesterday) Not sure which one... Thinking about it some more, it might be possible if that linking was done when compiling libf2c also.
Pretty sure it is an option when compiling 'gfortran' from what I read yesterday. But I couldn't get an nec2 to function when it was compiled/dynamically linked to gfortran.

https://en.wikipedia.org/wiki/Automa...gebra_Software


But if only nec2c or necpp, that wouldn't do us any good with the scripts if they can't be parsed by nikiml's scripts. If I can figure out how to build nec2c or necpp with atlas I'll let ya know anyhow.

I WAS looking at the differences in the fortran code between a regular build of 'nec2' and what they call 'nec2small' in the debian package though to see what needs to change...

in order to see how people were building different versions to add support for more wires, etc...That part seemed pretty straight forward. Couple of lines need to change before compiling is all. Not sure what the trade offs are though. Maybe that was a big deal back in the day when the program was first written, but maybe not so much anymore???

I don't remember much of anything about fortran. We studied it freshman year of High School (punch cards and all), and Freshman year of college on dumb terminals. about it. So everything I say here could just be BS, as I'm not a sw developer. Just doin a lot of googlin to get sumthin accomplished.

If I download those models and run 'time nec_eval.sh whatever.nec'
for each of those models, that would yield the exact amount of real time used to run the eval.
Would that work for you??

I am currently only using i5-3230M CPU @ 2.60GHz lenovo laptop, 2nd hand. Planning on changing to a better CPU (i7 quad core/8 threads)
some time if I find one cheap enough.
Remember the stellar labs yagi model?
that took 1 second to run.
Code:
time nec_eval.sh stellar_30-2475_holl.nec
with 
Input file : stellar_30-2475_holl.nec                                                                                                                                   
Freq sweeps: [(174, 1.5, 39)]
real 0m1.125s
user 0m3.781s
sys 0m0.076s

edit: apologize, result above was from an old amd quad core h/p laptop. Too many shells open at once
This is the result from the i5-3230M lenovo. A little over 1/2 a second to run the same stellar yagi model.

real 0m0.604s
user 0m1.816s
sys 0m0.040s

DB8E/VHF Yagi rotor FM Bandstop ap-8700 preamp 8way split LG lcd.
majortom is offline  
post #1256 of 1274 (permalink) Old 2019-03-24, 09:40 PM
Veteran
 
Join Date: Jan 2009
Posts: 4,652
Maybe it IS doing some multi-threading holl_ands...
take a look at this, while running the w2imu.nec feedhorn model (freq scaled to CBAND) which is gonna be quite a bit more complicated than a simple wire antenna.

'time nec_eval.sh w2imu.nec'

took 28 seconds to eval that, and all 4 threads (2 cores/4 threads) are crankin while running.
I vaguely recall running that in the past, using 4nec2 in linux under wine, and it was much slower I think.
But that woulda been on an even older computer than this lenovo.
[

real 0m28.434s
user 1m40.236s
sys 0m2.928s

DB8E/VHF Yagi rotor FM Bandstop ap-8700 preamp 8way split LG lcd.
majortom is offline  
post #1257 of 1274 (permalink) Old 2019-03-24, 10:08 PM
Veteran
 
Join Date: Jan 2009
Posts: 4,652
In case anyone is curious of the changes between the regular 'nec2' vs the 'nec2small' fortran source in the debian package source... I assume 'nec2small' can't handle as many wires/nodes as 'nec2'.
I did a 'diff' to compare 'em. then used 'wc' and counted 66 lines are changed.

https://pastebin.com/ZzbaLSjB

DB8E/VHF Yagi rotor FM Bandstop ap-8700 preamp 8way split LG lcd.
majortom is offline  
post #1258 of 1274 (permalink) Old 2019-03-25, 02:09 AM
Veteran
 
Join Date: Jan 2009
Posts: 4,652
holl_ands,
what I saw using ur mclapp_m4_super_dar model.
lenovo T430 notebook from ~ 2012 or 2013??
model name : Intel(R) Core(TM) i5-3230M CPU @ 2.60GHz
My grub bootloader's kernel command line is feeding this to the kernel at boot.
Code:
GRUB_CMDLINE_DEFAULT="intel_pstate=disable"
I added the below 2 lines to ur model so as to use the nec_eval script:
Code:
CMD--EVAL --auto-segmentation=0 --char-impedance=300 --num-cores=2                                                                                                      
CMD--EVAL --uhf-52 --centers  
CE

I ran it twice, once with the cpu governor set for "ondemand", and again set for
"performance"

cpu governor - ondemand (max seen ~ 3 GHz) 2 threads, nowhere near 100%
mclapp_m4_super_benchmark
real 1m53.639s
user 3m30.931s
sys 0m4.593s

cpu governor - performance (stays at 2.6 GHz) 2 threads ~ 100%
mclapp_m4_super_benchmark

real 1m49.651s
user 3m28.730s
sys 0m4.505s

Then change the switch to 4 cores:
Code:
CMD--EVAL --auto-segmentation=0 --char-impedance=300 --num-cores=4                                                                                                      
CMD--EVAL --uhf-52 --centers  
CE
Governor set for "ondemand"
real 1m40.905s
user 6m0.526s
sys 0m7.184s

Governor set for "performance"
real 1m37.779s
user 6m4.926s
sys 0m6.843s

with num-cores=1, it takes 3mins,9secs

DB8E/VHF Yagi rotor FM Bandstop ap-8700 preamp 8way split LG lcd.
majortom is offline  
post #1259 of 1274 (permalink) Old 2019-04-05, 10:26 PM
Veteran
 
Join Date: Jan 2009
Posts: 4,652
put the new to me cpu in the new to me lenovo notebook this evening...
so now with --num-cores=8
the eval time improved to 1 min 8 secs. Not bad I guess, for 6 or 7 year old hardware.

time nec_eval.sh mclapp_super_m4_dar.nec --num-cores=8

real 1m8.545s
user 6m28.936s
sys 0m7.521s

DB8E/VHF Yagi rotor FM Bandstop ap-8700 preamp 8way split LG lcd.
majortom is offline  
post #1260 of 1274 (permalink) Old 2019-04-06, 06:31 PM
Veteran
 
Join Date: Jan 2009
Posts: 4,652
post 1259 is while set for 'performance' at a fixed clock speed of 2.6 GHz.
Can get nearly the same result with --num-cores=4, and set for 'ondemand' allowing
the clock to run up to 3.6 GHz.

time nec_eval.sh mclapp_super_m4_dar.nec --num-cores=4

real 1m6.148s
user 3m48.706s
sys 0m4.795s

DB8E/VHF Yagi rotor FM Bandstop ap-8700 preamp 8way split LG lcd.
majortom is offline  
Reply

Tags
4nec2 modeling tips

Quick Reply
Message:
Options

Register Now



In order to be able to post messages on the Canadian TV, Computing and Home Theatre Forums forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.

User Name:
Password
Please enter a password for your user account. Note that passwords are case-sensitive.

Password:


Confirm Password:
Email Address
Please enter a valid email address for yourself.

Email Address:
OR

Log-in










Thread Tools Search this Thread
Show Printable Version Show Printable Version
Email this Page Email this Page
Search this Thread:

Advanced Search
Display Modes
Linear Mode Linear Mode



Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On

 
For the best viewing experience please update your browser to Google Chrome