Did I build a low parasitic Resistor or a dud.

Hi guys, I need a good test of my new Antenna Analyzer before I decide whether to return it. In preliminary tests I found a 14% error at 500 ohms, so I have attempted to build a low parasitic resistor for the second test. I built the resistor using Four-2000 ohm 0805 resistors It measures 501 ohms on my Fluke 79 and 499.8 ohms on my DE-5000 LCR meter. On my new Antenna Analyzer, I'm measuring the 500 ohm resistor, as shown in the Dropbox photos below.

At 1.1MHz it measures 427.0 ohms and -62.3 ohms Xc. 2400pf

At 3.5MHz it measures 420.3 ohms and -72.2 ohms Xc. 629pf.

At 10MHz it measures 406.0 ohms and -104.8 ohms Xc. 151.9pf.

At 20MHz it measures 373.3 ohms and -146.8 ohms Xc. 54.2pf.

At 30MHz it measures 334.7 ohms and -174.4 ohms Xc. 30.4pf.

formatting link

All I need is confirmation that my resistor should not have parasitics as high as measured and there must be a problem with my new AA-35 Antenna Analyzer.

Thank, Mikek

Reply to
amdx
Loading thread data ...

Your numbers are off and rather strange. I would guess(tm) that the measurements should be purely resistive up to about 10 MHz or so.

There's no way you have 0.0024uF across 500 ohms at 1.1Mhz in your dummy load. Measure the capacitance with your LRC meter. I think it should work across 500 ohms.

Next, setup your MFJ259 and try again. The MFJ-259 isn't particularly accurate, but it is far better than the numbers your AA-35 is belching. I don't know what's causing the weird numbers, but I would guess(tm) that the AA-35 has gone insane.

"Calibrating Your MFJ-259" There are more elaborate procedures, but this one will show any gross errors. Unfortunately, the symptom of blown input diodes is inaccurate numbers: so I would make sure that the MFJ-259 is actually working correctly before declaring something else to be broken.

Incidentally, at 1MHz, you could build your Beverage load out of a bundle of #2 pencils and connect them to the antenna with clip leads. Any stray capacitance or lead inductance is not going to be noticeable at 1MHz and the value of the load resistor is not that critical unless you want the best possible directionality (F/B).

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

A cheap 0805 resistor, mounted properly, will be nicely resistive to several GHz. Those capacitance numbers are not reasonable.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

[...]

0805 resistors have a parallel parasitic capacitance of about 50fF and a series inductance of about 900pH.

Jeroen Belleman

Reply to
Jeroen Belleman

Your resistor has to be fine with that construction.

My guess is that the analyzer is not good so far from its nominal design impedance (50 ohms). If I were you, I'd build another resistor for 50 ohms and try again.

--

-TV
Reply to
Tauno Voipio

Even more accurately/specifically/pedantically, it's approximately three transmission lines in cascade: a slightly lower impedance for each of the pads/end caps, and slightly higher for the resistive element itself (plus a lot of loss, of course).

The above figures are, of course, the LF approximations of this structure.

In short, any wavelength many times the body length of the resistor (and connecting leads/traces, if applicable!) will see a damned good resistance and nothing more.

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
 Click to see the full signature
Reply to
Tim Williams

That depends on the value of the resistor, of course. Whenever

1/(2 pi f 50fF) is lower than R, the reactance of the parasitic capacitance dominates the impedance. So for a 100k resistor, that would be above 32MHz, even if the wavelength is still hugely greater than the resistor size.

None of this matters to the OP, is true.

Jeroen Belleman

Reply to
Jeroen Belleman

I was wondering if there was a short length of coax in the analyser to the internals, but even that can't explain the numbers.

--
Mike Perkins 
Video Solutions Ltd 
 Click to see the full signature
Reply to
Mike Perkins

It's much better at 50 ohms with, About 1/2 ohm low and a couple tenths of an ohm reactance. But you did bring up the question I have and that is, How accurate is it at resistances other than 50 ohms. I have had an email exchange with RigexpertsUSA and gave them all my measurements up to 800 ohms and received the following.

"Although most of our Analyzers are within about 2% and at RigExpert USA we try to keep ours to as close to 1% as possible, it could be possible that depending on the type of resistor, the tolerance of resistor, and its ability to handle RF your readings could be affected, or your Analyzers tolerance could be off." So, I don't know if he didn't actually read my numbers and just assumed I was near 50ohms or if it is really accurate at large R values. I will say I got an email response from RigexpertUSA within 3 hours.

PS, here's where the idea for the GHz termination came from.

Mikek

Reply to
amdx

Thanks, John

I can't get Jeff's post to display, so copied from Google groups. >All I need is confirmation that my resistor should not have parasitics >as high as measured and there must be a problem with my new AA-35 >Antenna Analyzer.

Your numbers are off and rather strange. I would guess(tm) that the measurements should be purely resistive up to about 10 MHz or so.

There's no way you have 0.0024uF across 500 ohms at 1.1Mhz in your dummy load. Measure the capacitance with your LRC meter. I think it should work across 500 ohms.

I can't see both Ohms and Capacitance, But I do see Zero phases angle at 100kHz, with my 500 ohm resistor.

Next, setup your MFJ259 and try again. The MFJ-259 isn't particularly accurate, but it is far better than the numbers your AA-35 is belching. I don't know what's causing the weird numbers, but I would guess(tm) that the AA-35 has gone insane.

"Calibrating Your MFJ-259" There are more elaborate procedures, but this one will show any gross errors. Unfortunately, the symptom of blown input diodes is inaccurate numbers: so I would make sure that the MFJ-259 is actually working correctly before declaring something else to be broken.

I have calibrated it many times, I need to do it again, just now it says 500 ohm with a 50 ohm resistor. It has a problem with an intermittent part. Can't do it this morning, maybe in the afternoon.

Incidentally, at 1MHz, you could build your Beverage load out of a bundle of #2 pencils and connect them to the antenna with clip leads. Any stray capacitance or lead inductance is not going to be noticeable at 1MHz and the value of the load resistor is not that critical unless you want the best possible directionality (F/B).

Maximum nulling of anything but the signal I want is the optimum, so variable is the way to go. btw. The graph I showed was after adjusting the termination in the shack at the end of my speaker wire, then going out and testing the actual BOG wire. If the termination wasn't right, I did get a changing reactance on the graph.

I ran across these methods of setting the termination yesterday. From W8JI's page. There are three fast, simple ways to test for proper termination: With an Antenna SWR Analyzer

" Connect the antenna analyzer at the Beverage feedpoint through a good matching transformer Sweep the analyzer frequency from 1.8 to 7 MHz (or over a ~4:1 frequency range near the frequency intended for antenna) while watching SWR Adjust termination for minimum SWR variation (not minimum SWR, minimum SWR variation!)

When installation (including grounds) and termination is proper, SWR VALUE will remain nearly the same regardless of frequency With an Antenna Impedance Meter

Measure the feedpoint impedance (right at the feedpoint) of a roughly terminated antenna at the frequencies of highest and lowest resistive impedance. You can do this through a known good transformer by correcting impedance for use of the transformer Multiply the lowest measured impedance by the highest, and then find the square root of that number. This will be the correct termination impedance of the antenna"

Thanks, Mikek

Later for work, hopefully my wife will fire me!

--
Jeff Liebermann     je...@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
amdx

Sigh. Fix your Thunderbirdie mail program index per instructions I previously posted. Or, better yet, get a decent news reader (Forte Agent).

Bad news. That's what a blown diode acts like. Calibration isn't going to fix that. Two ways that I've found that blows them up. One is to leave it connected to a Hi-Z antenna for too long where it builds up a static electricity charge and blows the input diode(s). That's why I like antennas that look like a short circuit at DC. You can put a resistor across the antenna to bleed off most of the charge. A better way to blow out the front end is walking across a plastic carpet in rubber sole shoes, building up a static charge, and then dumping the static charge into the MFJ259 when you connect the PL259. The PL259 is perfectly designed for doing this. You have one hand on the MJF259 case and the other on the grounded sleeve of the PL259. The projecting center conductor of the PL259 makes contact first, which completes the circuit and unloads the static charge into the SO239 receptacle. I've been replacing the SO239 connectors with Type-N panel connectors whenever possible.

Anyway, replace all 4 diodes. $0.67/ea. Buy some extras and match them. A curve tracer is useful, but not necessary. Hmmm... my collection is down to 10 that are probably mismatched. Time to restock. The hard part is getting the RF board out of the box. Be careful.

Ok. A variable termination might be justified. Personally, I think it's a waste of time.

When you blow up your last remaining optically isolated CdS resistor, think about using an ordinary potentiometer and a radio control servo actuator. I have a design for such a contraption to turn the butterfly capacitor on a magnetic loop antenna. It would also work for a motorized antenna tuner.

I'm not familiar with the AA-35 or the BOG antenna, so I can't tell you how either is expected to act. Considering the frequency and the non-resonant low-Q(?) design, I would not expect to see much of a peak (or null) on the graphs.

W8JI also wrote the detailed calibration instructions for the MFJ259.

Are you using a transformer?

Ok, the antenna is low-Q and you need a wide sweep to see anything. I sweep every antenna I design or build to see what they do. My weapon of choice is one of several sweep generators depending on frequency range: The impedance is displayed using a (Texscan) return loss bridge: or a directional coupler with a diode RF detector. The display is an oscilloscope. I use an MFJ-269 when I need to measure something on the roof, and I don't want to drag all the equipment and cables outside.

You can only do that when you see the frequency sweep. I think your AA-35 will do that with the computah interface. Try it.

Sure, as long as your antenna isn't resonant. Since the wire portion of your Beverage antenna is much less than a wavelength at 1MHz, it won't be resonant and therefore will be fairly flat. Same with the feed line.

Yep. However, if you want to compensate for changes in ground resistance, wire droop, foliage, and such, and you want the best possible front to back ratio, you might need to tweak the value.

May I suggest that before proceeding with this antenna project, you verify the operation of ALL your test equipment and calibrate where necessary? As near as I can determine from here, literally every instrument that you own, including the new AA-35, has something wrong with it. With electronics and RF, we're essentially blind. We can't see electrons (or holes) moving. We can't see RF. We're totally dependent on various instruments to see what is happening. If those instruments are in any way lacking, we get a distorted picture of reality.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

The data sheet at: says: "Measurement for 25, 50, 75 and 100-Ohm systems" Again, I suggest you build a 4:1 impedance broadband transformer and downconvert your relatively high impedance measurements to something that the AA-35 can handle.

The MFJ-259 spec sheet doesn't offer a number for upper impedance measurement range. The MFJ-269 says: Measure SWR and loss of coax with any characteristic impedance, 10 to over 600 Ohms, including 50, 51, 52, 53, 73, 75, 93,95,300,450 Ohms -- MFJ-269C exclusive! Fix your MFJ-259 or borrow one that works and compare numbers. Better yet, use a 4:1 transformer to get it down to 50 ohms.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

For a 50 ohm resistor, with those numbers, a current step rings at about 20 GHz, Q around 2. 250 ohms is about critically damped.

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

That's how I got the 259 it had bad diodes. I replaced them, probably

20 years ago. I don't suspect the diodes, but you could be right.

It's already done.

Yep.

Yes, I've tried three different ones. A Laird toroid, don't have the number, (wrong computer) and a 4C6 potcore and a 73 binocular core. I wanted to use the AA-35 to compare the xformers, but the numbers were nothing like I expected, so I put that on hold for now. The potcore is in there now.

Or, that's where he expects to use his beverages.

I bought the AA-35 to replace the MFJ. That doesn't seem to have worked out so well.

With electronics and RF, we're essentially blind. We can't

Reply to
amdx

The question still stands, is my AAA-35 out of specs? I'm waiting for the company to respond to my second email. I did build a transformer and yes, the impedance is transformed. I'm starting to think the info I got from RigexpertUSA of 1% to 2% accuracy lacked some additional info, like "from 25 ohm to 100 ohms". But I gave him measurements up to 800 ohms, so it seems like he would have clarified that.

I just got an email back from RigexpertUSA, it gives me a little pause, in that, I think they are about to learn something I would have expected them to know already. Their response, "I'm having our Bench engineer look into this more thoroughly. This will require him having to make a bench test jig as well. All of our calibrations are done using the center band using a 50 Ohm load, and not

500 Ohm resistors. Therefore we do not have this available for testing purpose."

I found this look at four antenna analyzers and it confirms what I'm seeing. Accuracy falls of away from 50 ohms. A QST article.

I guess I need 2 or 3 transformers to analyze my wacky antennas. Now, what am I to make of the transformed reactance, Damn, it's always something! Couldn't they do some correction in software?

Graph of BOG at the end of my ~100 ohm speaker wire in the shack.

Mikek

Reply to
amdx

Am 22.06.2018 um 15:09 schrieb amdx:

I think in the noninv. Smith diagram you see series caps and parallel inductors easily.

FWIW I have build a similar thing this evening: SMA connectors,

50 Ohm microstrip, 49R9 and 4* 1960 Ohm termination and put them on both a R&S ZVB8 network analyzer with a Z51 active cal unit and a 54754A 18 GHz TDR.

I think that a VNA takes quite a hit already at 500 Ohms, even if calibrated.

The pictures are on <

formatting link
>

and those when you follow the right arrow. We can assume that the termination presented by amdx is blameless.

Bed time NOW.

cheers, Gerhard

(there is also a nice dragonfly to the left!)

Reply to
Gerhard Hoffmann

Just to clarify, I have a DE-5000 LCR meter, with my 500 ohm resistor, it reads 499.8 ohm with zero phase angle. It does not display C or L, when measuring a resistor.

formatting link

I never felt I was blamed, just looking for the adults in the room to look over my shoulder and check things. When I saw the article showing all four antenna analyzers lose accuracy when you get away from 50 ohms, I assumed my expectations were to high. Thanks for the testing. It will be interesting to get the results from the distributors bench tech, I'm surprised they don't already know this.

Reply to
amdx

That was meant in the sense of Douglas Self's blameless AF amplifier. ;-)

<
formatting link
>
Reply to
Gerhard Hoffmann

I just found a menu item that switched from Series Model to Parallel Model. That does give a somewhat more reasonable number for the capacitance if parallel, 45.2pf. Still much higher measurement than the actual, but it is now a bad tolerance, rather than just crazy. Mikek

Reply to
amdx

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.