Continuity Measurement with a Multimeter

[cobalt232]

hello,

i was wondering why when people demonstrate continuity sensitivity they insist on using the ends of the probes like they're striking a flint rock.
when i try this with my multimeters with a rapid motion the results are always spotty at best.

when i use an intermediary, say a knife blade or any conducting metal, the results are fantastic. the difference is quite noticeable.

so why do many reviewers use the first method instead of the second?
am i missing something?

thanks

[tautech]

hello,

i was wondering why when people demonstrate continuity sensitivity they insist on using the ends of the probes like they're striking a flint rock.
when i try this with my multimeters with a rapid motion the results are always spotty at best.

when i use an intermediary, say a knife blade or any conducting metal, the results are fantastic. the difference is quite noticeable.

so why do many reviewers use the first method instead of the second?
am i missing something?

thanks

Welcome to the forum.

Most everybody wants a fast continuity test with low burden voltages and current to reduce the chance of damaging components on a PCB. Manufacturers implement this as a feature or just another thing their DMM can do. 

In this thread Joe uses a sig gen to find the minimum on time a meter can detect and the max frequency it can also detect.
https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/
Many meters that Joe tests are documented in his spreadsheet and you can find the fastest listed.
Some are so fast the continuity test can be used to play a tune.

[RayRay]

i was wondering why when people demonstrate continuity sensitivity they insist on using the ends of the probes like they're striking a flint rock.
when i try this with my multimeters with a rapid motion the results are always spotty at best.

If you ask me, that is most likely due to using cheap probes.
Now, some people would recommend buying a quality set of probes to resolve this problem, but I'm not gonna do that (as there's another solution to this!) you could simply use sandpaper on the metal ends (sideways) to sharpen em a bit, and then the continuity speed would work just as good as using more expensive probes (unless your multimeter of choice has a really slow continuity check that is).
I've used this technique multiple times with great success.

[xrunner]

i was wondering why when people demonstrate continuity sensitivity they insist on using the ends of the probes like they're striking a flint rock.
when i try this with my multimeters with a rapid motion the results are always spotty at best.

Another interesting thing you might run into is on higher end DMMs, you can set the resistance that the DMM considers to be "continuity". For example you might set continuity to be <= 15 ohms. So if you test for continuity across a 20 ohm resistance, it won't beep, but if it's 10 ohms, it will beep.

[cobalt232]

thanks all for the advice.

i'll definitely check out that link.

i didn't consider that the quality of the probes could be a determining factor.
your thinking is probably right in that using the sides of the probe ends isn't as reliable or practical as using the actual point on the end.
good catch!

and just for something to do, i will see if adjusting the continuity resistance also changes things.

[SeanB]

Before sanding the probes first clean them with some IPA and a lint free cloth, and if that does not improve things then use some metal polish to get the oxide layer off. Sandpaper destroys the probe coating, and this then causes them to degrade further rapidly.

[RayRay]

Sandpaper destroys the probe coating, and this then causes them to degrade further rapidly.

Never had this problem.
As long as you don't use an overly abrasive sandpaper, and only sand em lightly, they'd be fine.

[2N3055]

I personally never liked latching continuity. I used non-latching continuity testers many times to locate intermittent connection problems.. Those manifest as noisy, scratchy tone.
Solid tone, solid connection. It is additional info.
 

[rsjsouza]

The goal is to test the meter's latching logic, and see how fast it responses to short circuit input without the buzzer being scratchy.

One practical aspect is to get a faster way to find the pinout of a complex connector, for example: put one of the probes at a given pin and the other can be quickly zipped through a row of pins on a PCB - if it does not beep you know it is either NC or the cable is broken. If it beeps, you know the surroundings of the pin location.
 

I don't think there is a particular need for a latching continuity mode, it's most like a psychological fantasy -- if I spend $500 on a meter, it better not sound like a scratchy $5 DT830B.
And no, not all manufacturers give this a $hit. Some high end Keysight meters don't implement latching (selectable, the latching mode works really badly, so practically non-latching mode is used all the time).

There are several discussions about this around here, with the occasional mud flinging.

[joseph nicholas]

I like the non-latching probes.  I also like that some cheap analog meters will give a audio indication of how much resistance there is between test points.  When doing a quick check for continuity you get some feedback from other non continuity test point such as diodes while in continuity mode, then you can go back an test the suspects using the proper diode mode.

[kripton2035]

here you have many continuity testers to build yourself :
http://kripton2035.free.fr/continuity-repos.html

[retiredcaps]

i didn't consider that the quality of the probes could be a determining factor.

TechnologyCatalyst, a member here, did a video on the above subject.

[RayRay]

Those common UNI-T probes (as shown on the above video) are actually not bad at all (once fixed with sanding^)
I have a pair of em myself, and they work perfectly for continuity.
Buying a new item to resolve an issue with the current is not always the best option.
In case of crap probes & continuity, some light sanding and it'd be just as good as those $20-30 probes.

[xrunner]

Those common UNI-T probes (as shown on the above video) are actually not bad at all (once fixed with sanding^)
I have a pair of em myself, and they work perfectly for continuity.
Buying a new item to resolve an issue with the current is not always the best option.
In case of crap probes & continuity, some light sanding and it'd be just as good as those $20-30 probes.

OK I guess I'll bite ...

What are we sanding off the "crappy" probes? They look bright and clean. Has anyone looked at them under a microscope?

What is the bad layer that is being removed by sanding? 

[joeqsmith]

Those common UNI-T probes (as shown on the above video) are actually not bad at all (once fixed with sanding^)
I have a pair of em myself, and they work perfectly for continuity.
Buying a new item to resolve an issue with the current is not always the best option.
In case of crap probes & continuity, some light sanding and it'd be just as good as those $20-30 probes.

OK I guess I'll bite ...

What are we sanding off the "crappy" probes? They look bright and clean. Has anyone looked at them under a microscope?

What is the bad layer that is being removed by sanding? 

I don't get it myself.  Why would I want to ruin the plating?  For the most part I use a set of Probemaster's and a set of Fluke needles.

I am not a big fan of the latching type myself. May have been a deal breaker on the BM869s.  But to each their own. There is fast, then there are some that basically have no filter at all.

[tooki]

Those common UNI-T probes (as shown on the above video) are actually not bad at all (once fixed with sanding^)
I have a pair of em myself, and they work perfectly for continuity.
Buying a new item to resolve an issue with the current is not always the best option.
In case of crap probes & continuity, some light sanding and it'd be just as good as those $20-30 probes.

OK I guess I'll bite ...

What are we sanding off the "crappy" probes? They look bright and clean. Has anyone looked at them under a microscope?

What is the bad layer that is being removed by sanding? 

I'm thinking I should do a video on this myself. What I've observed is this: with any probes other than my ProbeMaster ones, the measured resistance of the probe tips themselves varies significantly depending on the pressure applied, and whether they're touching in the little grippy ridge or not. (As measured by touching the probe tips to each other.) The ProbeMasters, on the other hand, show almost no variability in resistance with pressure applied. I'm not a chemist, but I assume there's an oxide layer on the chrome plating whose resistance is higher than that of the gold plating of the ProbeMasters. This might explain why sanding and cleaning helps for some people.

When used with a non-latching continuity tester (in my case, a cheapie 830L and an okay Mastech), this resistance variability manifests as scratchiness. On good latching meters (in my case, a Fluke 87V and a Keithley 2015), it manifests as taking longer to register and beep. Don't get me wrong, the 87V and 2015 are extremely fast even with bad probes. But with the ProbeMasters, they're even better.  The ProbeMasters produce a rock-steady tone on the non-latching meters.

And mind you, the other probes I'm comparing to aren't crappy ones, they're Fluke (i.e. Pomona) and Oldaker. (The Mastech probes actually don't fare any worse than these, to be honest.) But the ProbeMasters beat them all handily. (I've got the Probemaster 8017S standard probes and the 8150 fine probes that come with both fixed and spring-loaded tips. The latter have become my daily-use probes.)

[RayRay]

@xrunner Never looked at em under a microscope, but I'd assume there was some kind of oxide layer on em.
Anyhow, what happened was that I got a new multimeter a while back, and noticed the continuity wasn't working proper, I had some sandpaper around, and decided to give that a go, and it worked wonders!

@joeqsmith Lightly sanding the sides of the needle ends with sandpaper that is not overly abrasive will not ruin the plating in any way.
It's nothing too aggressive, and mine still have their original color and work fine for all uses.

[xrunner]

@xrunner Never looked at em under a microscope, but I'd assume there was some kind of oxide layer on em.
Anyhow, what happened was that I got a new multimeter a while back, and noticed the continuity wasn't working proper, I had some sandpaper around, and decided to give that a go, and it worked wonders!

I believe the claims - I simply want to know why?

What is being sanded off so as to make the continuity work better? I don't have any Probemasters (but I guess I'll order a set) to compare against, but I do have a lot of marginal quality probes. I guess I'll break out my USB microscope and find a set that seems to cause the continuity fail a little and see what happens. 

[SingedFingers]

I'd argue this is all semi-religious rubbish.

Even the shittiest probes that barely worked that came with my UT61E work really well on a Fluke 8021B or an 8024B with the continuity beeper on. The meter is where the problems are if there are any. Sanding your probes I suspect is just fixing the meter's insensitivity.

I've got a pair of Pomona probes for the 8024B above and it's simply the best continuity implementation of any meter I've used. It has a properly triggered monostable, not something derived from the ADC.

[xrunner]

Dug out my USB microscope. I used a UNI-T UT136B for the tests. I got a pair of the crappiest probes I had, really worthless, but came free with something I got once. I tried them on the cont. test first and got very intermittent.

But - it's not quite so simple. I also tried another DMM - an ANENG AN8002. The crappy probes worked worse on that meter. So it's not just the DMM that's the issue.

I then tried the probes that came with my Rigol DM3058E. They worked much better. I took pics of them and made a side-by-side pic of the crappy probe tip vs the Rigol probe tip, please find attached.

Next I'm going to do another experiment ...

[tooki]

I'd argue this is all semi-religious rubbish.

Even the shittiest probes that barely worked that came with my UT61E work really well on a Fluke 8021B or an 8024B with the continuity beeper on. The meter is where the problems are if there are any. Sanding your probes I suspect is just fixing the meter's insensitivity.

I've got a pair of Pomona probes for the 8024B above and it's simply the best continuity implementation of any meter I've used. It has a properly triggered monostable, not something derived from the ADC.

Trust me, it makes a difference. As I already stated, even on meters with top-notch continuity (87V and K2015), there's a difference. A much smaller difference than on the cheap meters, but a difference nonetheless.

Yes, a really good meter can mask poor probes, but that doesn't mean good probes aren't better.

[rsjsouza]

One practical aspect is to get a faster way to find the pinout of a complex connector, for example: put one of the probes at a given pin and the other can be quickly zipped through a row of pins on a PCB - if it does not beep you know it is either NC or the cable is broken. If it beeps, you know the surroundings of the pin location.

The best latching one can't beat the speed of a non-latching one.

And nor I am claiming so (if you think I implied that, otherwise disregard).

Dug out my USB microscope. I used a UNI-T UT136B for the tests. I got a pair of the crappiest probes I had, really worthless, but came free with something I got once. I tried them on the cont. test first and got very intermittent.

But - it's not quite so simple. I also tried another DMM - an ANENG AN8002. The crappy probes worked worse on that meter. So it's not just the DMM that's the issue.

I then tried the probes that came with my Rigol DM3058E. They worked much better. I took pics of them and made a side-by-side pic of the crappy probe tip vs the Rigol probe tip, please find attached.

I have observed a variability of scenarios that confused me about making any implications regarding the ability to beep. One thing I think it is happening is that output voltage/current may make a difference in breaking the thin oxide barrier created on the probes. Depending on the material of the probes the oxidation may happen almost instantaneously, thus they would be back to their previous "oxidized" state in the small time interval between switching DMMs. 

[xrunner]

I polished the crappy probes with Simichrome polish. Now they work a whole lot better. So part of the problem is the finish on the probes, but as I said, it's not simply that issue, even with the original crappiness, the continuity test works better on some multimeters than others.

[PA4TIM]

I use a whetstone to sharpen my probes.  The only downside up to now is  I have learn not to catch them between my knees when they drop from the table. 

[xani]

Another interesting thing you might run into is on higher end DMMs, you can set the resistance that the DMM considers to be "continuity". For example you might set continuity to be <= 15 ohms. So if you test for continuity across a 20 ohm resistance, it won't beep, but if it's 10 ohms, it will beep.

I'd love if some multimeter vendor did advanced version of that, with 2-3 tones (say <7, <15, and <30 ohm) to say how good connection is

[kripton2035]

this one make a continous sound with variable frequency depending on the ohms it measures :
https://www.elv.de/elv-kurzschluss-detektor-kd100-bausatz.html

[SingedFingers]

This thread actually made me try something.

I've just tested my Fluke 8024B continuity on the end of two 500m reels of 7/0.2 wire. That's 1km path between the terminals.

It's still fast and sensitive!

[Zbig]

I'd love if some multimeter vendor did advanced version of that, with 2-3 tones (say <7, <15, and <30 ohm) to say how good connection is

Keysight U1270 U1273A and U1273AX have a mode like that.

[Wytnucls]

Not quite. The beep frequency can be changed by the user. There is no frequency change according to the resistance encountered.

[xrunner]

In any case, I did order a set of Probemasters for myself yesterday. Should have done that years ago I guess. 

[Zbig]

Not quite. The beep frequency can be changed by the user. There is no frequency change according to the resistance encountered.

On 1273A, you can choose SINGLE or TONE. The chosen frequency affects the constant tone in SINGLE mode only. In TONE, the meter produces a cacophony of different pitch noises and my understanding is, the frequency is directly dependent on the measured resistance in this mode.

[Wytnucls]

Tone is only available on the 1273A and 1273AX, not on the whole series.
My understanding is that tone is just a long beep, instead of a single short beep for continuity or lack of it.
There is no evidence in the user manual that the frequency changes with resistance.
Of course, if you have personal experience with the meter, I stand to be corrected.

[Zbig]

Tone is only available on the 1273A and 1273AX, not on the whole series.

You're right. I thought the only difference between U127[1/2]A and U1273A(X) was the display but confirmed what you say with the manual and corrected my post.

My understanding is that tone is just a long beep, instead of a single short beep for continuity or lack of it.
There is no evidence in the user manual that the frequency changes with resistance.
Of course, if you have personal experience with the meter, I stand to be corrected.

No, it's just labeled in a bit misleading way. "BEEP" is actually classic latched beep. It sounds for as long as you keep the probes shorted with the added "inverse" (shifted) mode when it only ceases to beep when shorted (normally open/normally closed). The "TONE" mode produces a kind of "singing" tone varying between several discrete frequencies. I agree the manual doesn't elaborate on that at all but I own two of these meters.

[Wytnucls]

A resistance-dependent tone frequency would be easy to test with a bunch of decreasing low value resistors. Have you tried the experiment yet?

[Zbig]

A resistance-dependent tone frequency would be easy to test with a bunch of decreasing low value resistors. Have you tried the experiment yet?

Not yet, but I might. I don't expect it to be dependent on anything other than resistance, though. The only other option would be a random generator and that wouldn't make whole lot of sense

[Wytnucls]

I just don't see how they can implement that with a latched system. Perhaps the meter goes into a non-latched mode with tone.

[Zbig]

I just don't see how they can implement that with a latched system. Perhaps the meter goes into a non-latched mode with tone.

Maybe I was wrong saying it's latching. I mean I think it is but it has some scratchiness to it so perhaps it's not. Anyway, I'll try to make a short video later as I wanted to test out my camera's video functionality anyway

[rsjsouza]

Not quite. The beep frequency can be changed by the user. There is no frequency change according to the resistance encountered.

On 1273A, you can choose SINGLE or TONE. The chosen frequency affects the constant tone in SINGLE mode only. In TONE, the meter produces a cacophony of different pitch noises and my understanding is, the frequency is directly dependent on the measured resistance in this mode.

Zbig, you are correct. I just tested this. From 0 to 300\$\Omega\$ the U1273A produces a tone that decreases the frequency as the resistance increases.

To access this, go to Setup menu and select "Tone" on the Continuity selection.

Quite interesting this meter...

[Wytnucls]

Can you confirm that it is a non-latching mode, like possibly on the new U1282A.

[rsjsouza]

Wytnucls, it is non-latching. I am rendering a short clip for demonstration.

(edit) the video:
https://youtu.be/fH6rL2U2oAI

[tpowell1830]

Hi Cobalt232,

Welcome to the forum.

The reason that I use the method that you describe is to test the speed at which the buzzer sounds. When testing multiple points to a common point for continuity, I rely on hearing the buzzer to confirm continuity, without looking at the meter, but focussing my eyes on the point I am testing. Using this method, you don't want to have to wait for the buzzer to sound. If you have to wait too long for the buzzer, quite simply, it takes longer to ring out the points in question, not to mention disrupting your rhythm.

Other folks may have different ideas, but that is why I do it this way.

[joseph nicholas]

I don't get it.  Why are buzzers latched at all?  Is it for people who are preoccupied with other things when testing continuity and need to shift their attention to the sound of the buzzer?  Is this for stubborn people who refuse to use their hearing to confirm until they look at the continuity led to confirm what their ears have already told them?

[dimkasta]

For me, it's just part of a good user interface.
Both work, it is just a continuity test after all... But a solid tone is a far more non-intrusive experience. Even slightly satisfying, if you are geeky enough.

I got a couple of sets from probemaster last week, and it feels so nice testing stuff

[Wytnucls]

Thanks fort the informative video. I'm surprised that the high-end U1273A doesn't have a latched continuity mode.
Perhaps Keysight can't do a latched continuity properly, because after watching Dave's review of the little brother U1272A, I learned that the continuity test is latched, but is painfully slow.

The fast frequency change in tune with resistance value is interesting, but is there a real world use for that kind of continuity test? Just wondering...

PS: Waiting for the video of the 5th.

[Rick Law]

I personally never liked latching continuity. I used non-latching continuity testers many times to locate intermittent connection problems.. Those manifest as noisy, scratchy tone.
Solid tone, solid connection. It is additional info.

For me, it's just part of a good user interface.
Both work, it is just a continuity test after all... But a solid tone is a far more non-intrusive experience. Even slightly satisfying, if you are geeky enough.

I got a couple of sets from probemaster last week, and it feels so nice testing stuff

I am of the school that solid tone should indicate solid continuity.  Speed is a good thing, but failing to show intermittent connection by continuing with a solid tone is in a way the meter lying to me.

Now different frequency for different resistance would be a great thing.

By the time a manufacturer come up functions something that every member here is happy, the UI will need a 16 page setup menu.  So simplicity is probably best.

[helius]

I don't get it.  Why are buzzers latched at all?

As you can see in rsjsouza's U1273A video, some aren't. On a Fluke meter, the latched continuity provides a consistent and dependable signal that you have found a low-impedance connection. You can sweep over a whole connector and even if you only make contact for a few milliseconds, the beep will be consistently audible. This isn't the case for non-latched meters, which might not let through enough cycles of the tone for you to hear.
Other companies' implementations of latched continuity are usually much worse than Fluke's and correspondingly less useful (the typical complaint is that they are too slow to be used this way).
One device that does do it well is the Wavetek SF10 (but as hinted in that thread, it may actually be manufactured by Fluke anyway!)

[dimkasta]

I am of the school that solid tone should indicate solid continuity.  Speed is a good thing, but failing to show intermittent connection by continuing with a solid tone is in a way the meter lying to me.

Now different frequency for different resistance would be a great thing.

By the time a manufacturer come up functions something that every member here is happy, the UI will need a 16 page setup menu.  So simplicity is probably best.

Why would you consider this lying?
It is our job to know what tool to use.

The frequency change is a nice idea. It would be nice to see new features for a change. I bought my first DMM ~25 years ago, and it had the same functionality with the ones I use today... With today's tech we should be able to get rock solid DMMs for peanuts. Not paying hundreds of dollars/euros for features that existed in DMMs 20 years ago...

[mmagin]

I've thought about designing my own continuity tester +  milliohm meter with Toneohm-like functionality.  The ability to configure the latching is an interesting idea.  More of what I wanted though was:

• < .1 volt to avoid turning on semiconductor junctions
• use of an AC signal (ala HP 4328a) to allow insensitivity to thermal EMFs
• possibly a variable resistance-to-tone-frequency response, at least because I have no idea what I'd want in practice

[dimkasta]

How about pulsing?

[Rick Law]

I am of the school that solid tone should indicate solid continuity.  Speed is a good thing, but failing to show intermittent connection by continuing with a solid tone is in a way the meter lying to me.
...
...

Why would you consider this lying?
It is our job to know what tool to use.
...
...

Hmmm...   It is just a personal preference (bias) of connotation. so I find it hard to respond to your "why I consider this lying?"  It would be like trying to explain why I like green better than yellow.

Let me try this:  A perfect solid tone for a connection implies/connotes a perfectly good connection.  If the connection is intermittent, the tone should reflect that imperfection.  When an intermittent connection generates the same tone as a perfect connection, it hid the problem - so as I said, "in a way the meter [is] lying to me."

[dimkasta]

Yeah I got what you meant.  I can see how a non-latching one can be useful. I have used my dmms like that a couple of times to test scratchy interconnects.

My "objection" was about the word "lying". I find it a bit unfair. DMMs and especially latching ones never claimed that they reveal the presence or absence of intermittent connections with a scratchy tone. It is a symptom, not a feature.
They can only be trusted to identify existing connections or non-connections on that specific time.
Anything else is subject to personal judgement and proper use or misuse of specific characteristics of the DMM

If someone decides to use a latching DMM to identify an intermittent connection, it is not the DMM that is lying. It is him the one using an unsuitable instrument.
The same way it would be his fault if he used a non-latching one and got unlucky enough to get a solid tone

[Rick Law]

Yeah I got what you meant.  I can...
...
My "objection" was about the word "lying". I find it a bit unfair. DMMs and especially latching ones never claimed
...

We are on the same page.  I agree it really isn't "lying".  I just can't think of a better way to express it so I soften it by say "in a way it was lying".

Note of course it was not my short coming.  It was the fault of my thesaurus - it failed to give me better suggestions when I needed them.

[Wytnucls]

If someone decides to use a latching DMM to identify an intermittent connection, it is not the DMM that is lying. It is him the one using an unsuitable instrument.
The same way it would be his fault if he used a non-latching one and got unlucky enough to get a solid tone

A latching continuity detects an intermittent connection just like the non-latching meter, but instead of a barely audible signal, you get a loud 200ms beep to confirm a short or open connection. The main difference is that a latching continuity will pick up a one-event very short transient, which could easily be missed by the user's ear on a non-latching meter.

Master Handbook of Acoustics:

[Rick Law]

If someone decides to use a latching DMM to identify an intermittent connection, it is not the DMM that is lying. It is him the one using an unsuitable instrument.
The same way it would be his fault if he used a non-latching one and got unlucky enough to get a solid tone

A latching continuity test detects an intermittent connection just like the non-latching meter, but instead of a barely audible signal, you get a loud 200mS beep to confirm a short or open connection. The main difference is that a latching continuity will pick up a one-event very short transient, which could easily be missed by the user on a non-latching meter.

Wytnucls, thanks for the explanation!  I didn't know that a latched meter would warn of the open connection.  A good day is one that I learned something new.  So today started well!

Thanks

[Wytnucls]

The meter with the best fast latched continuity mode I know of, is the Fluke 867B which can detect transients as short as 1us. One doesn't even have to be close to the meter, as they also register on the trend plot.

[Simon]

@xrunner Never looked at em under a microscope, but I'd assume there was some kind of oxide layer on em.
Anyhow, what happened was that I got a new multimeter a while back, and noticed the continuity wasn't working proper, I had some sandpaper around, and decided to give that a go, and it worked wonders!

I believe the claims - I simply want to know why?

What is being sanded off so as to make the continuity work better? I don't have any Probemasters (but I guess I'll order a set) to compare against, but I do have a lot of marginal quality probes. I guess I'll break out my USB microscope and find a set that seems to cause the continuity fail a little and see what happens. 

I sell some of the probemaster leads if you in Europe 😁

Sent from my Moto G (4) using Tapatalk

[xrunner]

I sell some of the probemaster leads if you in Europe 😁

Sent from my Moto G (4) using Tapatalk

I don't live there but I got a set last week - very nice. Should have done it long ago.