Multimeters are indispensable tools when working on electronics. It’s almost impossible to build any but the most basic of circuits without one to test and troubleshoot potential issues, and they make possible a large array of measurement capabilities that are not easily performed otherwise. But when things start getting a little more complex it’s important to know their limitations, specifically around what they will tell you about circuits designed for high frequency. [watersstanton] explains in this video while troubleshooting an antenna circuit for ham radio.
The issue that often confuses people new to radio or other high-frequency projects revolves around the continuity testing function found on most multimeters. While useful for testing wiring and making sure connections are solid, they typically only test using DC. When applying AC to the same circuits, inductors start to offer higher impedance and capacitors lower impedance, up to the point that they become open and short circuits respectively. The same happens to transformers, but can also most antennas which often look like short circuits to ground at DC but can offer just enough impedance at their designed frequency to efficiently resonate and send out radio waves.
This can give some confusing readings, such as when testing to make sure that a RF connector isn’t shorted out after soldering it to a coaxial cable for example. If an antenna is connected to the other side, it’s possible a meter will show a short at DC which might indicate a flaw in the soldering of the connector if the user isn’t mindful of this high-frequency impedance. We actually featured a unique antenna design recently that’s built entirely on a PCB that would show this DC short but behaves surprisingly well when sending out WiFi signals.
Click bait :(
It’s not lying, ‘you’ just don’t know what you’re doing.
Maybe a click bait, but it worked. Not only did you clicked, but also commented. Me? I came here to drop this link to article I wrote almost two years ago:
https://www-elektroda-pl.translate.goog/rtvforum/topic3805119.html?_x_tr_sl=pl&_x_tr_tl=en&_x_tr_hl=pl&_x_tr_pto=wapp
Google Translate messed up LaTeX formula syntax…
With MathML being what it is, that’s why some make math formula into image snippets.
There’s the phrase “know their limitations” which really should apply to any tool. e.g. shovel as a cheese spreader.
“A man’s got to know his limitations.”
-Dirty Harry
The tool you are looking for is an LCR meter ($$$, only works at a few fixed frequencies). A spectrum analyzer and sweeping sine wave/function generator would be preferable ($$$$). Or if you are really serious a vector network analyzer ($$$$$). It’s also possible (but tedious and time consuming) to manually test the frequency response of a circuit using an oscilloscope and function generator that you “sweep” yourself while and take amplitude and phase measurements.
Another tool for measuring an antenna’s impedance or resonance was the noise bridge, I vaguely remember.
For measuring resonance in circuits, a grid dip meter was also helpful. However, most modern users/operators aren’t accustomed to using that anymore.
NanoVNA, not even $$$, more like the price of a couple large pizzas. The accuracy isn’t terrible, and it can discern these situations with ease. Idk if I’d believe it on the high end (they claim 1.5GHz) but it’s decent enough up to probably 300MHz. Vector network analysers are no longer the exclusive domain of research labs!
Yep, cheaper than the black-box antenna testers and cheaper than a multimeter good enough to have a good version of any of the same features. Just as long as you don’t need the high end frequencies, it works fine.
Yeah if you are working with RF systems especially more complex systems with combiners, tower top amplifiers, duplexers, and other antenna arrays you have to have a good service monitor capable of line sweeping, spectrum analyzer and more. The smallest thing can cause issues even with that equipment. Even using the wrong barrel adapter just for testing can throw off readings or disconnecting/reconnecting lines can leave metal shavings inside the connector. I’ve had older cat 3 cable show it was good but once connected couldn’t handle the signal
Or how about a multimeter where the voltage displayed varied on the state of the internal battery with no indication of battery state. That was fun trying to diagnose a 3phase furnace fault…
That £100 meter went into the bin.
If it’s using a 9v battery, then there’s a workaround, maybe.
Those “10 year” batteries based on Lithium Ion technology do contain an electronic circuit that tries to keep the voltage constant.
It also prevents explosion caused by shorts or a too high power draw in general.
The circuit is vital for the battery to be used as a drop-in replacement for 9v alkaline/carbon-zink batteries.
https://lygte-info.dk/info/BatteryDisassembly9VLiIon%20UK.html
Yes, my multimeter literally lies to me when the battery is low, it starts to read double the actual voltage.
It has been relegated to the carage where seeing 24V on a motorcycle circuit is just as good as seeing 12V, and indicates that it is time for a new battery.
Inre: your motorcycle
I have often wondered if Pugh owns a Puch?
B^)
I used to, when I was 16 I had a Puch moped.
Thanks, one question scratched off my bucket list!
NanoVNA H4 —– My MFJ-269 had not been out of the box in over a year now.
Also–
TinySA Ultra and a simple “sampler” and a decent scope and you have
(almost) all you need for RF stuff.
And yes—Click bait
Be careful with the NanoVNA. Some versions make use of harmonics to increase operation frequency, which makes them less reliable.
For trustworthy results, something like a RigExpert AA170 is a better solution.
I’m speaking from experience, by the way. A friend has an AA170 that is much more reliable than my humble NanoVNA. It got same results for multiple measurings of same test object, while my NanoVNA was not reliable. Hand capacity is also an issue. The AA170 has a proper chassis that provides insulation.
Below 300MHz the v1 nanoVNA (original design) uses the fundamental, not harmonics.
So it goes higher than that AA170, at a fraction of the price.
Additionally, in my opinnion they are fine for testing amateur radio antennas up to 1300MHz, even in harmonic mode.
SAA-2/NanoVNA V2 uses fundamental across the whole 50kHz – 3GHz span. And some later versions have an improved bridge design, so the specification is extended up to 4GHz or 4.4GHz.
With both the buyer should beware when using amazon/aliexpress/ebay, as there are plenty of clones out there.
The LiteVNA makes it all the way up to 6!
And “using harmonics makes it less reliable” is just complete and utter hogwash.
“trustworthy results”
I have a AA230 and the H4 from R&L.
In over a year of testing I have found the H4
to be at least as good up to 2M. And it was a heck
of alot $$.
I would sell the 230 if I would not lose so much $.
Like the IC-7300 changed radio, the NanoVNA/TinySA
changed “testing/measurement”–Both for the better.
Knowing what I know now, I would not have spent the
$$ on the 230 and just waited on the H4.
(I also have the MFJ-269—–Sitting in a box :( )
Phew, user error then.
I was expecting yet another subtle plug for Fluke.
Maybe not so subtle…
Just ditch the (fluke) TL175 test leads for starters. While they may last for a while on the bench they don’t hold up in the field in my experience causing all kinds of errors at the worst times IMO
How about a follow-up article: “What to do when you don’t know if your meter is lying to you”?
Bottom line, person doesnt know what they are doing. Not a meter issue, and not worthy of HAD.
This is what happens when the editorial team are more writers than tech people. I miss the old HAD so much.
The first rule is to have *two* good meters. If they read the same, you can be reasonably confident in the reading. But if they differ… “A man with a clock knows what time it is. A man with two clocks isn’t sure.”
There are many common cases where the readings will be in error. The AC range on most meters measures the average value, but the scale displays the equivalent RMS value. This only works if the waveform is a sinewave. You need to use a “true RMS” meter for anything but a sinewave.
AC circuits also have phase shifts and power factor. AC volts x AC amps is *not* power; unless you also know the waveform and power factor!
A meter’s AC ranges also have a frequency response. They generally work OK at audio frequencies, but are increasingly inaccurate above 100 KHz. Don’t expect it to tell you if your micro’s clock is oscillating!
Measuring voltages around RF sources can also produce screwy readings. For an example, hold your meter near your microwave oven. Cheap meters will often have large errors.
When measuring current, most meters use their lowest voltage scale to measure the drop across an internal resistor. If the meter is 400 mV full-scale, then it has a 400mV drop at full scale on its current ranges. That’s enough drop to cause significant changes in a circuit’s operation.
this sentence about inductors turning open and capacitors turning short at high frequencies reminds me. i once decided to ‘solve the antenna riddle for once and for all’ with 5 minutes of googling. and what i came at was a formula which a better writer would reproduce here (but i won’t), which basically said that whether your circuit is a short or open basically changes by the distance measured in wavelengths. like half (or a quarter? oh i don’t remember) a wavelength down the wire from a short, it is open. and then another quarter wavelength away from that, it is short again. i thought it was pretty neat that a little piece of math could say that so simply, and then that the human mind could accept it. what a great way to think of a circuit, where the two most opposite circuits can be created or found by merely travelling along the wire.
This is how I memorized some of the Amateur Extra test questions. They ask what the transmitter sees for an open or shorted transmission line at a particular fraction of the wavelength. You imagine the actual standing wave on the line, and look at where the nodes and peaks are. If the end of the line is at a node, there’s no potential at that point, because the top and bottom waves have no distance between them, so if the line is shorted at the end, you have a high impedance. If the end is at a peak, there’s a high potential, because the top and bottom waves have maximal distance between them. If the line is shorted in this configuration, you’ve got an actual short, which is a low impedance. If the line is not shorted, which means high impedance. If the end of the line isn’t on a node or peak (1/4 and 1/8th wavelengths), move up one odd harmonic and try again. Thus 1/4th wavelength is the same as 1 full wavelength, and 1/8th is the same as 1/2. Also, if the line is shorted, the reactance is inductive, and if the line is open, the reactance is capacitive.
If you understand the relationship of wavelength, transmission line length, and the visual representation of the standing wave, you don’t have to memorize the answers, because you can derive them yourself on-the-spot.
You guys really need to make this fail gracefully. I’ve now had my comments magically disappeared twice, with the response “Nonce Failed”, because no one thought that maybe when a comment fails to post, the text should be preserved for the writer to at least copy/paste for another try. I’m not going to repeat another half hour of free labor purely for the benefit of others, if I’m going to be treated so poorly for doing so.
If you want to learn what I attempted to post, go watch Veritasium’s videos on how power doesn’t flow through the wires, and then watch this ~2 hour training presentation on grounding for RF circuitry: https://www.youtube.com/watch?v=ySuUZEjARPY
Sorry, I tried to summarize it, but crappy UX resulted in my work being utterly wasted.
There are browser add-ons that serve that function.
“You can jump through extra hoops to get what should be part of the site in the first place.”
Thanks for the oh so useful advice.
I always hit copy before submitting a long post to a forum because posts do frequently get eaten (not only Hackaday, though their “failed nonce” does sound a bit pervy)