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

donate to eham
   Home   Help Search  
Pages: [1]   Go Down
  Print  
Author Topic: calibrating/testing a homebrew antenna analyzer?  (Read 446 times)
AJ6IY
Member

Posts: 3




Ignore
« on: September 24, 2019, 03:05:59 PM »

I'm a terminal cheapskate (not to mention, fiddling with electronics is the fun part of this hobby for me!), so when I got back into radio recently, I threw together a homebrew antenna analyzer, based on the K6BEZ design.  I had most of the parts around from other projects, but I didn't have an MCP6002 or any AA143 diodes.  So, I used an LM358P and a couple of 1N60P diodes instead.

After a couple false starts with a bad Arduino Nano, I got something that is giving me readings that look reasonable for most stuff, but on the high side.  For example, a 50 Ohm QRP dummy load (made from 20 1K 3-watt resistors in parallel) shows an SWR of about 1.6 at 1MHz down to around 1.4 at 30Mhz;  an end-fed half wave I built for 20m doesn't get below a 2.0 SWR anywhere, though the curve itself looks reasonable.

So my question is:  how accurate can I expect this to be?  Did I screw up by using an LM358?  (I have some MCP6002s on order, but they won't arrive for a few weeks).  Is there any way I can (or should, for that matter) compensate for my poor life choices?    Enquiring minds want to know! 

I should point out that right now, this is the only way I have to directly measure SWR.  I do have access to an oscilloscope, multimeter and that sort of stuff, but very little radio-specific measurement tools.

Thanks for any insight,

+j
Aj6IY
Logged
WA2ISE
Member

Posts: 1294




Ignore
« Reply #1 on: September 24, 2019, 03:18:45 PM »

If everything else looks right, other than the SWR numbers, maybe you should divide the SWR variable by 1.5 to get it to match known dummy loads. Oh, check your dummy load with a DVM in ohm mode, to be sure it's really 50 ohms.  See what SWR you get if you use 100 ohms, 300 ohms, 25ohms, and such to see if they yield reasonable SWR values consistent with what you get with 50 ohms.

Once you get it done, be sure to post a construction article so others can make their own too.
« Last Edit: September 24, 2019, 03:21:25 PM by WA2ISE » Logged
AJ6IY
Member

Posts: 3




Ignore
« Reply #2 on: September 24, 2019, 03:34:18 PM »

That was certainly my original thought (dividing things out), but I figured I'm probably missing something important.  The load measures 51.4 ohms with my DMM, so that's pretty good.  I'll give the other resistances a try, too.

As far as a construction article, if/when I ever give the thing its own display (I'm currently graphing with matplotlib on my macbook pro) I'll give it a go...

Logged
VK4FFAB
Member

Posts: 482




Ignore
« Reply #3 on: September 24, 2019, 07:51:40 PM »

That was certainly my original thought (dividing things out), but I figured I'm probably missing something important.  The load measures 51.4 ohms with my DMM, so that's pretty good.  I'll give the other resistances a try, too.

As far as a construction article, if/when I ever give the thing its own display (I'm currently graphing with matplotlib on my macbook pro) I'll give it a go...



Grab a 50 ohm calibration standard off ebay for a couple of bucks. It will have a flat response into the Ghz, and then some 0.1% resistors in 20, 100 and 200 ohms you can then use the results you get there to Math out any errors. It wont be perfect, but good enough to have meaningful data from it.
Logged
WA8NVW
Member

Posts: 91




Ignore
« Reply #4 on: September 25, 2019, 03:51:57 PM »

Also try using the ARRL Lab technique of either 2:1 or 3:1 steps from 50 ohm Zo mid-point calibration of your analyzer: 
Find the closest single resistor to these values in your junkbox.  These are NOT for transmitter testing under power!
So 12.5, 25, 50, 100, and 200 ohms should display 4:1, 2:1, 1:1, 2:1, 4:1 VSWR
and 6, 17, 50, 150, and 450 ohms should show 9:1, 3:1, 1:1, 3:1, 9:1 VSWR.
When you tweak the display math, if you can get VSWR readings at both ends of
either series within 15% and still have 51 in the middle within 5% you are good to go!
Logged
AJ6IY
Member

Posts: 3




Ignore
« Reply #5 on: September 25, 2019, 04:44:56 PM »

Hm, is "single resistor" important?  Seems like this would be a good use for the decade box I built a couple years ago...

Great responses.  Thanks!  I'm looking forward to this weekend and having some time to try them out...
Logged
WA8NVW
Member

Posts: 91




Ignore
« Reply #6 on: September 25, 2019, 04:53:26 PM »

Try to eliminate inductive reactance caused by long component leads, which gets worse at the higher frequencies.  Decade boxes are great for DC, but not so much for RF.
Logged
G8HQP
Member

Posts: 967




Ignore
« Reply #7 on: September 26, 2019, 03:41:47 AM »

Did you build the impedance bridge part of the circuit as small as possible, and as close as possible to the original design? If not you will get poor results as frequency rises. Your mention of using a decade box for testing suggests that you do not understand the vital importance of this.
Logged
KB1GMX
Member

Posts: 1824




Ignore
« Reply #8 on: September 26, 2019, 02:15:16 PM »

best way to go is resistor mounted in a connector.

for suggestions...

use 1/8th watt resistor or other small noninductive resistors.

12.5ohms    4:1 low
16.67 ohms 3:1 low
25 ohms      2:1 low
50 ohms        1:1 to 30mhz if the leads are short.
75 ohms      1.5:1
100 ohms    2:1 high
150 ohms    3:1  high
200 ohms    4:1 high

Those values will be suitable for calibration.

Allison
Logged
WD4HXG
Member

Posts: 379




Ignore
« Reply #9 on: September 26, 2019, 05:07:01 PM »

Measure the DC Resistance of the Dummy Load you built. Your resistor might be off a tad and it does not take much offset from 50 ohms to yield a 1.6:1 VSWR.
Logged
W9IQ
Member

Posts: 3521




Ignore
« Reply #10 on: September 26, 2019, 05:14:47 PM »

Measure the DC Resistance of the Dummy Load you built. Your resistor might be off a tad and it does not take much offset from 50 ohms to yield a 1.6:1 VSWR.

Generally a DC resistance check is not too meaningful as the problem is often inductance. Even some laser trimmed film resistors will exhibit inductance.

As far as a "tad" off, if the resistor presents no reactance, it would have to be either 31 ohms or 90 ohms to create a 1.6:1 SWR50.

- Glenn W9IQ
Logged

- Glenn W9IQ

I never make a mistake. I thought I did once but I was wrong.
WD4HXG
Member

Posts: 379




Ignore
« Reply #11 on: September 27, 2019, 02:32:40 AM »

You are absolutely correct Glen.

Regards

Chuck
Logged
VK2TIL
Member

Posts: 648




Ignore
« Reply #12 on: September 27, 2019, 05:52:55 PM »

I had a look at the manual for this instrument; its heart is a basic return-loss or RF bridge;



At balance, ie when the load is 50 ohms, the voltage between X & X should be zero.

This voltage, DC by courtesy of the diode, can be measured with a DMM; the point marked A is appropriate.

A basic check would be to insert a 50 ohm load and check that this voltage is zero; there may be a "dead spot" caused by diode action but you should find a zero.

If that is OK, the problem is further down the line.

The LM358s should be OK as substitutes; they are a good component.
Logged
K9FV
Member

Posts: 496




Ignore
« Reply #13 on: October 01, 2019, 04:04:52 PM »

......... For example, a 50 Ohm QRP dummy load (made from 20 1K 3-watt resistors in parallel) shows an SWR of about 1.6 at 1MHz down to around 1.4 at 30Mhz; 
That sounds a lot like inductance/reactance issue since it's getting closer to correct at lower freqs.  Try 2.0 mhz and see what you get.
Logged
Pages: [1]   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!