Multimeters With Low Ohms Function

[EEVblog]

I have seen multimeters malfunction doing Ohms measurement on a DC rail with capacitance and semiconductors present.

Starting at the lowest Ohms range, beep and charging happens until the multimeter upranges with the corresponding lower compliance current then causing the capacitor's voltage to start dropping. This can also be aggravated due to semi's becoming biased on and current drain increases as the voltage goes up. It's a non-linear load. Auto-ranging will oscillate between ranges hunting all the time.

I believe the test current on the Brymen is quite low now at 0.3mA compared to 1-4mA on other multimeters. I just manual range to look for shorts.

Some old school meters have a "Low ohms" function that has a compliance voltage below 0.6V so it doesn't switch on semiconductor junctions in-circuit. Seems to have fallen out of favour since the 80's and 90's. My first digital meter, a Soar had it.

[EEVblog]

Some old school meters have a "Low ohms" function that has a compliance voltage below 0.6V so it doesn't switch on semiconductor junctions in-circuit. Seems to have fallen out of favour since the 80's and 90's. My first digital meter, a Soar had it.

e.g



And Metex:

[rsjsouza]

The Keysight U1272A/U1273A have a very low ohms voltage - I did a comparison a while ago with a few other multimeters (U1282A, 189, UT61E) and was quite pleased to find out this.
AFAIK it is not configurable like your examples.

[AVGresponding]

I have seen multimeters malfunction doing Ohms measurement on a DC rail with capacitance and semiconductors present.

Starting at the lowest Ohms range, beep and charging happens until the multimeter upranges with the corresponding lower compliance current then causing the capacitor's voltage to start dropping. This can also be aggravated due to semi's becoming biased on and current drain increases as the voltage goes up. It's a non-linear load. Auto-ranging will oscillate between ranges hunting all the time.

I believe the test current on the Brymen is quite low now at 0.3mA compared to 1-4mA on other multimeters. I just manual range to look for shorts.

Some old school meters have a "Low ohms" function that has a compliance voltage below 0.6V so it doesn't switch on semiconductor junctions in-circuit. Seems to have fallen out of favour since the 80's and 90's. My first digital meter, a Soar had it.

This sent me down the rabbit hole investigating the continuity compliance voltage on all my meters (no small task!     ).

I found some (to me) interesting results.

Most were in the range from 3-4.5V, with some outliers: F87V ~7V, Keithley 2015 6.6V, Gossen M242A 8.4V (running off 2 AA Ni-MH!), Fluke T5-1000 ~1.1V, UT139C 1.04V, and finally a winner-winner-chicken-dinner, the Mastech 2108A with 0.44V!
I had a strange result by accident when I tried my Fluke 99B S2, got around 500mV, but then realised it was in auto-ranging ohms mode, and not continuity. Put it in continuity, and it went right up to 3.2V!
Decided to try just ohms mode on a few other meters, and found pretty much the same thing; a compliance voltage of around 400-500mV, the one outlier I had there was the Tektronix DMM912 with ~850mV.
Obviously, using ohms instead of continuity has some drawbacks, like response speed, no audible signal (unless your meter has an audible output on %dev function, you could set that up), but it's worth bearing in mind, if you find something with continuity, check it with ohms too.

[EEVblog]

I'm thinking of a vidoe titled: "What happened to the Low Ohms Function?"
I could dig my original Soar out of the archives.
It would be interetsing to know what modern meters with have such a function, or have a complaince volatge under 0.6V?

[EEVblog]

This sent me down the rabbit hole investigating the continuity compliance voltage on all my meters (no small task!     ).

I found some (to me) interesting results.

Most were in the range from 3-4.5V, with some outliers: F87V ~7V, Keithley 2015 6.6V, Gossen M242A 8.4V (running off 2 AA Ni-MH!), Fluke T5-1000 ~1.1V, UT139C 1.04V, and finally a winner-winner-chicken-dinner, the Mastech 2108A with 0.44V!
I had a strange result by accident when I tried my Fluke 99B S2, got around 500mV, but then realised it was in auto-ranging ohms mode, and not continuity. Put it in continuity, and it went right up to 3.2V!

Yes, you might have to auto-range to get the desired result. the high ohms ranges have to have a higher compliance voltage otherwise the noise becomes an issue.

[bdunham7]

I'm thinking of a vidoe titled: "What happened to the Low Ohms Function?"
I could dig my original Soar out of the archives.
It would be interetsing to know what modern meters with have such a function, or have a complaince volatge under 0.6V?

Keithley 2010--low power ohms and dry circuit test <20mV OCV

https://download.tek.com/datasheet/2010.pdf

I have a dual-banana plug with a 10R resistor that I use for short circuit testing.  It keeps OCV down to 10mV with a 1mA test current and doesn't affect readings very much below 1R--this is an easy ad-hoc dry circuit testing method as well.

[EEVblog]

I'm thinking of a vidoe titled: "What happened to the Low Ohms Function?"
I could dig my original Soar out of the archives.
It would be interetsing to know what modern meters with have such a function, or have a complaince volatge under 0.6V?

Keithley 2010--low power ohms and dry circuit test <20mV OCV

https://download.tek.com/datasheet/2010.pdf

I have a dual-banana plug with a 10R resistor that I use for short circuit testing.  It keeps OCV down to 10mV with a 1mA test current and doesn't affect readings very much below 1R--this is an easy ad-hoc dry circuit testing method as well.

Interesting, front page brochure feature:

Unique Resistance Measurement Functions
Characterizing the resistance, linearity, or isolation of contacts, connectors, switches, or relays
completely and efficiently demands an uncommon combination of ohms measurement capabilities.
The 2010 offers:
• Low-power ohms measurement mode. Low-level resistance measurements can be made with
source current as low as 100µA, an order of magnitude lower than is possible with other DMMs,
so device self-heating is minimized. Among other benefits, this low-power measurement capability
makes the 2010 suitable for end-of-life contact testing per ASTM B539-90.
• Dry circuit test function. When measuring contact and connector resistances, it is important to
control the test voltage carefully in order to avoid puncturing any oxides or films that may have
formed. A built-in clamp limits the open circuit test voltage to 20mV to ensure dry circuit conditions.
• Offset compensated ohms function. This function eliminates thermal effects that can create
errors in low-level resistance measurements in system environments.
• Extended ohms measurement capability. The 2010 provides a 10W range for more precise
measurements of low resistances

[Neutrion]

This reminds me of the discussion somwhere about why the low vs high continuity voltages are good or bad.
Although I don't remember any positiv point for the high ones apart from getting through dirt easyer.

AVGresponding: Did you also measure the currents?


On the other hand the voltage vs. semiconductors is just one thing, to not to have false alarm on capacitances would reqiere an other solution.

[floobydust]

Low-power Ohms feature was well known and common in the early 1970's to 1980's. It was great for repairs involving BJT's, you could fast test for shorted components. I remember the boss in the repair shop harping at us about the feature, it was kind of a standard feature that was expected but not really known about. Some multimeter models advertised it, and others allowed you to select the feature. A few I can find:

Beckman Tech 310/310/320, HD110 max in-range voltage 250mV, OCV 500mV
Beckman HD110B, 320B max in-range voltage 200mV, test 300mV except 3V on 200R range
Sencore I forgot the model numbers
Simpson 260-6XLM and 260-XLMP  "low power resistance" Rx1 100mV 4.9mA, Rx10 100mV 490uA
Fluke 8010A, 8012A, 8050A, 8060A
Sperry AWS DM-3010, Elenco M1200B, Triplett 601/603, Viz WD758 thru 763, Omega HHM57
Tektronix DMM155, 157 450mV, normal is 900mV.
Agilent 34420A low power mode is 10% limited power 1mA vs 10mA on 100R, and voltage limiting 20,100,500mV selectable clamp
Keysight 34460A, 34461A, 34465A have "low power" ohms mode but poorly documented.
TeleDyne TSC805, TelCom TC818 DMM IC switchable 0.35V on all ranges above 200R.

So it was great for servicing TV's, stereos (not germanium) and any IC's without low V substrate diodes or Schottky diodes. You could test for shorts in-circuit, no need to pull parts.
I dislike the creep to lower and lower ohms test currents in multimeters- away from the pristine lab/bench conditions, these choke and die out in the field.

[kripton2035]

self-promoting, but this is really kick-ass to locate short circuits...
http://kripton2035.free.fr/Projects/shorty-display.html

[Squarewave]

That Soar looks rather like the AVO DA212

Some old school meters have a "Low ohms" function that has a compliance voltage below 0.6V so it doesn't switch on semiconductor junctions in-circuit. Seems to have fallen out of favour since the 80's and 90's. My first digital meter, a Soar had it.

e.g

[Kleinstein]

One can still test for shorts in circuit, if the open circuit voltage goes higher. A diode junction would still not measure very low ohms, but what you get with some 0.6 V at a resistor and this is usually quite high.

There are 3 points with a limited voltage/power: one is protecting sensitive parts that can not withstand a higher voltage and could be damaged (e.g. BE junction in reverse from more than some 6 V, some LEDs - at least according to the abs. max specs and some were actually sensitive to > 5 V).

The other point is measuring sensors like PTCs / NTCs, where the test current may already cause significant self heating. A small PT1000 tested with 1 mA test current is not ideal. A 10 K NTC with 1 mA test current can be way off. 

The 3 rd point is dry contact testing, so testing the low votlage performance of contacts, where a higher voltage can destroy an oxide layer at the contact and this way hide a contact problem. This dry ohms testing wants a rather low maximum voltage, often realized with a prallel resistor.

A high voltage is nice for diode testing to also measure a blue / white LEDs and maybe a 5 V zener.
On the other side high resistor ranges may need a relatively high voltage to gets low senstivity to hum.

[AVGresponding]

This reminds me of the discussion somwhere about why the low vs high continuity voltages are good or bad.
Although I don't remember any positiv point for the high ones apart from getting through dirt easyer.

AVGresponding: Did you also measure the currents?


On the other hand the voltage vs. semiconductors is just one thing, to not to have false alarm on capacitances would reqiere an other solution.

I did not, but it's easily arranged. Give me a couple of days to perform the tests and tabulate the results though.

Regarding the issue of turning semiconductor junctions on, it's also worth mentioning that quite a few here like to tinker with older equipment, that may well be full of germanium transistors. None of the compliance voltages I've seen so far would be low enough to not turn a germanium junction on.

[nali]

My Meterman 38XR circa 2005 vintage is about 480mV on all ranges

[alm]

Some Fluke manuals have a note to this effect in their manual:

Most in-circuit resistance measurements can be made
without removing diodes and transistors from the circuit.
The full-scale measurement voltage produced on ranges
below 40 MΩ does not forward-bias silicon diodes or
transistor junctions enough to cause them to conduct. Use
the highest range you can (except 40 MΩ) to minimize the
possibility of turning on diodes or transistor junctions. Full-
scale measurement voltage in the 40-MΩ range does
forward-bias a diode or transistor enough to cause it to
conduct.

In all fixed resistance ranges (200Ω to 200 kΩ), the test voltage is less than
that required to turn on most semiconductor junctions. This feature,
sometimes referred to as “low power” ohms, aids in troubleshooting by
allowing you to measure resistors independent of the effects of in-circuit
transistors and diodes. For the fixed ranges the maximum full scale voltage
across the circuit being measured is less than 250 mV. The autoranging MΩ
ranges have enough voltage to turn on semiconductor junctions (maximum
2.5V full scale), but the current is very low (2.2 μA maximum).

The resistance function can produce enough voltage to
forward-bias silicon diode or transistor junctions, causing
them to conduct. To avoid this, do not use the 30 MΩ or
500 MΩ ranges for in-circuit resistance measurements.

It's not about the open circuit voltage, but about the maximum voltage for an in-range resistance reading. So as long as the meter is in a range below 40 MOhm (for this particular meter), and it gives an in-range reading, you won't be forward-biasing any normal semiconductor junctions. I guess that's why these meters don't have a separate low Ohms function, because any range below the tens of Megaohm ranges is effectively a low Ohms range?

[Neutrion]

• Low-power ohms measurement mode. Low-level resistance measurements can be made with
source current as low as 100µA, an order of magnitude lower than is possible with other DMMs,
so device self-heating is minimized. Among other benefits, this low-power measurement capability
makes the 2010 suitable for end-of-life contact testing per ASTM B539-90.
• Dry circuit test function. When measuring contact and connector resistances, it is important to
control the test voltage carefully in order to avoid puncturing any oxides or films that may have
formed. A built-in clamp limits the open circuit test voltage to 20mV to ensure dry circuit conditions.
• Offset compensated ohms function. This function eliminates thermal effects that can create
errors in low-level resistance measurements in system environments.
• Extended ohms measurement capability. The 2010 provides a 10W range for more precise
measurements of low resistances

[/quote]

In which arrangement could be more than 100µA be a problem because of device self heating?(Apart from the mentioned PTC from Kleinstein.)
Also I am not qiet sure if I understand the last two points exactly.

Dave, if you do a video about it, might would be nice to see on the scope, how the different meters clamp back the voltage and how fast. Some people defend the 7V of the 87V with the claim that it would be clamped back fast. Now It would be interesting to see, if it could cause a problem in any arrangement.

self-promoting, but this is really kick-ass to locate short circuits...
http://kripton2035.free.fr/Projects/shorty-display.html

Does your circuit avoid the the false positives of fast continuity beepers caused by capacitances?
Because the actual original "problem" in the bm786 topic was that, I don't even know why we drifted away completely into the low voltage/current domain.

This reminds me of the discussion somwhere about why the low vs high continuity voltages are good or bad.
Although I don't remember any positiv point for the high ones apart from getting through dirt easyer.

AVGresponding: Did you also measure the currents?


On the other hand the voltage vs. semiconductors is just one thing, to not to have false alarm on capacitances would reqiere an other solution.

I did not, but it's easily arranged. Give me a couple of days to perform the tests and tabulate the results though.

Thanks, it would be nice to see.

From the bm789 manual:
"Beeplit diode tester test current typical 0,35 mA" But it can go up to 0,5mA at least if I attach it to a 10ohm multimeter input. No maximum is specified. But in continuity mode it stays around 0,35 mA.

"Short beep-alert treshold: Drop across 0,85V."   Is this the short false positive beep? Drop across the leads from the open circuit voltage?

"Beeplit  continuous ON treshold: <0,1V " Drop across the leads?  I measure 3,6mV with the arrangement above measured paralell with another meter.

Will do some more measurement soon.

[kripton2035]

Does your circuit avoid the the false positives of fast continuity beepers caused by capacitances?
Because the actual original "problem" in the bm786 topic was that, I don't even know why we drifted away completely into the low voltage/current domain.

don't know. on what kind of device(s) can I test that ?

[Neutrion]

If the beeping function is as fast as on the Brymen, you can just start to probe around any circuit for shorts, to see wether you get any short false beeps where there are no shorts. But again, this only happens if the beeper function is fast enough. But if you did not design your circuit to avoid it, it would be a wonder if it would work.

[gamalot]

I just did a simple test on my Kyoritsu KEW1021R. Its open-loop voltage is about 3V (the voltage of two AA batteries), the voltage drop across a 10M resistor is 180mV, and the voltage drop across a 1K resistor is 95mV. The smaller the resistor, the lower the voltage drop. 

[Per Hansson]

Should not the name of the thread (and any future video) include the word power or compliance voltage?
As I read it now it sounds like a meter that can show low resistances, like a 4-wire measurement or the dedicated low ohms mode of the Fluke 289?

[Fungus]

Dave, if you do a video about it, might would be nice to see on the scope, how the different meters clamp back the voltage and how fast. Some people defend the 7V of the 87V with the claim that it would be clamped back fast. Now It would be interesting to see, if it could cause a problem in any arrangement.

Yep. These are current sources so measuring the voltage into a high impedance doesn't tell us much.

I assume the clamping will happen at the speed of analog electronics (ie. sub-microsecond) but it would be good to test this.

[kripton2035]

If the beeping function is as fast as on the Brymen, you can just start to probe around any circuit for shorts, to see wether you get any short false beeps where there are no shorts. But again, this only happens if the beeper function is fast enough. But if you did not design your circuit to avoid it, it would be a wonder if it would work.

the bip on the shorty has an audio tone proportionnal to the value of the short. so if there is a charging capacitor, the tone is changing, and you can hear it.

[AVGresponding]

I just did a simple test on my Kyoritsu KEW1021R. Its open-loop voltage is about 3V (the voltage of two AA batteries), the voltage drop across a 10M resistor is 180mV, and the voltage drop across a 1K resistor is 95mV. The smaller the resistor, the lower the voltage drop. 

This is because meters use a constant current source, and measure the voltage across the resistor under test to derive its resistance.

[AVGresponding]

Ok the tests took less time than I thought, so here are the results.

[David Hess]

Some old school meters have a "Low ohms" function that has a compliance voltage below 0.6V so it doesn't switch on semiconductor junctions in-circuit. Seems to have fallen out of favour since the 80's and 90's. My first digital meter, a Soar had it.

I bought my Tektronix DMM916 in the late 90s and it has 4 "resistance" modes:

Ohms   860 millivolts
Low Ohms   280 millivolts
Continuity Test   3.2 volts   Reads out ohms
Diode Test   3.2 volts   Reads out volts

My Beckman RMS225 from about the same age has:

Ohms   650 millivolts
Continuity and Diode Test   3.2 volts   Reads out volts

[EEVblog]

Ok the tests took less time than I thought, so here are the results.

Thanks for those.
The V in High voltgae will be reduced by divider of the 10M input impedance of the meter unless you used a High-Z meter input.

[ermionesrl]

Hello.
May I add some measurements of my own multimeters?

(EDITED: sorry, I mistakenly wrote some units, this is corrected table)

Meter; V in Cont; A in Cont; V in High R; A in High R; V in Low R; A in Low R
Fluke 177; 2.6941; 980.4u; 1.2439; 120n; 1.2440; 311.5u
Fluke 17B+; 0.5316; 159.2u; 0.5301; 45n; 0.5304; 158.8u
Fluke 15B; 0.4410; 140.7u; 0.4408; 45n; 0.4410; 140.7u

I used the Agilent 34401A in >10Gohm input mode, both for voltage and for current. For current, a 100 ohm 0.01% metal foil resistor shunt has been added. In some time I could eventually repeat the measurements with an electrometer, for better current values. I have a second 34401A and eventually I can measure it with its twin.

AVGresponding: you have used the Fluke 289, but its 10Mohm input voltmeter impedance made your measurement of high ohm V almost surely incorrect, the current source got loaded. You see that even the >10Gohm of 34401A (very) slightly load it.
I think that in most (if not all) cases, the burden voltage is the same between low and high ohm. As you own a 34401A, maybe you can repeat the measurements with it at >10G.

[AVGresponding]

The V in High voltgae will be reduced by divider of the 10M input impedance of the meter unless you used a High-Z meter input.

AVGresponding: you have used the Fluke 289, but its 10Mohm input voltmeter impedance made your measurement of high ohm V almost surely incorrect, the current source got loaded. You see that even the >10Gohm of 34401A (very) slightly load it.
I think that in most (if not all) cases, the burden voltage is the same between low and high ohm. As you own a 34401A, maybe you can repeat the measurements with it at >10G.

Fair point, I should have thought of that. Just got home from work, and I'm warming up my Fluke 8840A (it's the most recently calibrated high-impedance meter I have).

[Neomys Sapiens]

One can still test for shorts in circuit, if the open circuit voltage goes higher. A diode junction would still not measure very low ohms, but what you get with some 0.6 V at a resistor and this is usually quite high.

There are 3 points with a limited voltage/power: one is protecting sensitive parts that can not withstand a higher voltage and could be damaged (e.g. BE junction in reverse from more than some 6 V, some LEDs - at least according to the abs. max specs and some were actually sensitive to > 5 V).

The other point is measuring sensors like PTCs / NTCs, where the test current may already cause significant self heating. A small PT1000 tested with 1 mA test current is not ideal. A 10 K NTC with 1 mA test current can be way off. 

The 3 rd point is dry contact testing, so testing the low votlage performance of contacts, where a higher voltage can destroy an oxide layer at the contact and this way hide a contact problem. This dry ohms testing wants a rather low maximum voltage, often realized with a prallel resistor.

A high voltage is nice for diode testing to also measure a blue / white LEDs and maybe a 5 V zener.
On the other side high resistor ranges may need a relatively high voltage to gets low senstivity to hum.

Another very good reason for a low test voltage and limited energy would be the presence of squibs/detonators.

That said, both the Tektronix DM501A and DM502A have a HI/LO section switch for their kOhms ranges. For LO, maximum voltage is stated to be 0.2V.
The Siemens B1043 does it differently: when selecting a Resistance range, a diode symbol is shown if the voltage is above 0.5V. If none is shown, it is a 'diode safe' range.
The Metrawatt M2036 just gives a diagram in the manual from which it can be seen that the open-circuit voltage on all ranges except 30MOhms should be <400mV.

[David Hess]

Tektronix DMM916:

Ohms   860 millivolts   0.41 milliamps
Low Ohms   280 millivolts   0.03 milliamps
Continuity Test   3.2 volts   0.41 milliamps   Reads out ohms
Diode Test   3.2 volts   0.41 milliamps   Reads out volts

Beckman RMS225:

Ohms   650 millivolts   0.78 milliamps
Continuity and Diode Test   3.2 volts   3.2 milliamps   Reads out volts

Tektronix DM501:

Ohms   10.6 volts   1.0 milliamps   2k ohms range

Tektronix DM502:

Ohms   11.2 volts   1.0 milliamps   200 and 2k ohms ranges

Another very good reason for a low test voltage and limited energy would be the presence of squibs/detonators.

Electrical squibs and detonators are deliberately designed to be as difficult to trigger as possible.  They require high voltage at high current.

[Neomys Sapiens]

The DM501 and DM502 don't have that switch. The 'A' models are completely different.
-> TEKWiki

[AVGresponding]

Well, that was as much work as doing the whole test before! The settling time is very long, due of course to the low currents involved. Two meters refused to settle, rising to a peak over a number of seconds, then falling continuously (but very slowly) beyond the point of my patience and in the case of the Metrix, beyond the limit of the auto power off.
There is a broad pattern, but with exceptions, as you'd expect.

[ermionesrl]

AVGresponding:
Interesting: it seems that many multimeters use the same way as 34401A to measure >10Mohm, by inserting a parallel known 10Mohm resistor and do calculations. It's evident when the open circuit voltage is the result of the test current over 10Mohm, and when your previous measurement is the result of the test current over 5Mohm.
About the fact that some didn't settle, I think that the voltage source is not very well stabilized and thus reflect the battery voltage, that fall continuously during the test. If they use a ratiometric method against a known resistor (like original Fluke 87), both the current and the voltage need not to be stable over a timespan longer than the 2 adc samples.

[Neutrion]

If the beeping function is as fast as on the Brymen, you can just start to probe around any circuit for shorts, to see wether you get any short false beeps where there are no shorts. But again, this only happens if the beeper function is fast enough. But if you did not design your circuit to avoid it, it would be a wonder if it would work.

the bip on the shorty has an audio tone proportionnal to the value of the short. so if there is a charging capacitor, the tone is changing, and you can hear it.

So there are no sudden beeps with your gear at all where there is no continuity, like with fast meters?
Is it latched?

Well, that was as much work as doing the whole test before! The settling time is very long, due of course to the low currents involved. Two meters refused to settle, rising to a peak over a number of seconds, then falling continuously (but very slowly) beyond the point of my patience and in the case of the Metrix, beyond the limit of the auto power off.
There is a broad pattern, but with exceptions, as you'd expect.

Nice spreadsheet, thanks for it!
I also measured the BM789, but I don't have a high Z meter so I measured with the scope instead.

Vcont (with 10MOhm meter) 2,38V  p-p (with scope) 2,56V    A in cont (with 10Ohm meter) 0,35mA   V High Ohm (with 10MOhm meter) 0,221V  p-p(with scope)  3,52V   V in Low Ohm (with 10MOhm meter) 2,38V   p-p(with scope) 2,56V   A in low Ohm 0,35mA.

Edited to make it more clear.

It seems to differ from its specs as the open circuit volts in Ohms supposed to be less than 1,3V and less than 1,5 V for the 600 Ohms range.
And there is no nifference in manual range.

So only NTCs and PTCs are problematic?

[AVGresponding]

AVGresponding:
Interesting: it seems that many multimeters use the same way as 34401A to measure >10Mohm, by inserting a parallel known 10Mohm resistor and do calculations. It's evident when the open circuit voltage is the result of the test current over 10Mohm, and when your previous measurement is the result of the test current over 5Mohm.
About the fact that some didn't settle, I think that the voltage source is not very well stabilized and thus reflect the battery voltage, that fall continuously during the test. If they use a ratiometric method against a known resistor (like original Fluke 87), both the current and the voltage need not to be stable over a timespan longer than the 2 adc samples.

Given the two meters that were the most difficult in this regard were the Metrix MX 57EX and the Agilent U1401B, I think this unlikely. The Agilent in particular is a process calibrator; its outputs are in my experience very stable. There must be another explanation, perhaps the meters are cutting the current back once the ohms over-range is detected for an extended period, in order to extend battery life a fraction?

[Neutrion]

When you measure the volts with the BM869 in high Ohms range, is there any difference, if you measure the peak value with a scope?

Should not the name of the thread (and any future video) include the word power or compliance voltage?
As I read it now it sounds like a meter that can show low resistances, like a 4-wire measurement or the dedicated low ohms mode of the Fluke 289?

Good point. And the whole topic started because of the false beeps on continuity mode, which we forgot.
Hovewer it is also an interesting topic, and that is what can be annoying with many meters.

[Kleinstein]

Ideally you want the choice:
High current is wanted to measure small resistors.
High voltage is needed to test higher voltage diodes, maybe even zeners and also high resistors work better this way.
A high voltage (e.g. >5 V) may however damage sensitive parts (e.g. LEDs, Transistors with BE in reveerse).
Not too high a current may be preferred for very small, temperature sensitve resistors.
A very limited voltage is needed for testing contacts in dry ohms, not braking through oxide films.
So no one setting is good for all. Only a few meters have different setting for the different jobs. For the usually limited resolution of hand held DMM the power from the test current is rarely a problem. It is more with some bench meters where the power could go over the top, like 10 mA with 12 V open circuit (and maybe 8 V with current ?) for the 3458 in the 10 Ohms range.

The beeps with a capacitor are a 2 sided thing: they can be confusing if you don't know about it. The can even help to detect capacitors. With a fast reacting beeper, there is no good way around it - waiting for a stable voltage would be a bit over the top and make the reaction slow.

[AVGresponding]

When you measure the volts with the BM869 in high Ohms range, is there any difference, if you measure the peak value with a scope?

I'll give it a poke. I'd expect the relatively low impedance of a scope/probe to have a marked effect.

[Neutrion]

Ideally you want the choice:
High current is wanted to measure small resistors.
High voltage is needed to test higher voltage diodes, maybe even zeners and also high resistors work better this way.
A high voltage (e.g. >5 V) may however damage sensitive parts (e.g. LEDs, Transistors with BE in reveerse).
Not too high a current may be preferred for very small, temperature sensitve resistors.
A very limited voltage is needed for testing contacts in dry ohms, not braking through oxide films.
So no one setting is good for all. Only a few meters have different setting for the different jobs. For the usually limited resolution of hand held DMM the power from the test current is rarely a problem. It is more with some bench meters where the power could go over the top, like 10 mA with 12 V open circuit (and maybe 8 V with current ?) for the 3458 in the 10 Ohms range.

I hope if Dave does the video, he will cover also some practical and theoretical cases for damage.

But from the cited document above, with the 100uA "for less damage" so far we only got  PTC and NTCs if I understand right. Or are there any other  "temperature sensitive resistors" which we can encounter?
And I don't mean the 12V with 10mA, but the average handheld meters current and voltage.

The beeps with a capacitor are a 2 sided thing: they can be confusing if you don't know about it. The can even help to detect capacitors. With a fast reacting beeper, there is no good way around it - waiting for a stable voltage would be a bit over the top and make the reaction slow.

I was thinking about doing two really fast measurement. If the second shows less current, than there would be  no beep because of capacitance. If they are equal, than comes the beep. Obviously it would need a dedicated circuit.

[floobydust]

With low power ohms, I find it's mainly because you need the ability to differentiate between presence of a semiconductor junction or ohmic resistance - in-circuit. For the compliance-voltage limiter.
Damaged semiconductors frequently test as if they're a leaky diode-junction. With multimeters, there is a resistance (say 500R) on diode-test that looks like a good junction. There is a Vf that looks like a resistor. You have to know these values.

Transistor has three leads, six possible readings (both polarities) and double that if you are using both ohms and diode-test. If repairing gear with many BJT's like audio power amplifiers, missing one damaged transistor can be very costly. They will almost test OK, especially some non-symmetrical VBE and VCB junctions which are usually quite close in Vf but not with damaged parts. Or resistive leakage (carbon) is present. Or one junction is low or high Vf. Have to be really thorough sometimes.

[ermionesrl]

AVGresponding:
This is very puzzling then. I have not been able to find the schematic of either meter to check, yet it would be very strange if this was intentional. In this case, you would almost surely see a step change.

Neutrion:
If you used a regular x10 probe with your scope, its input Z should also be 10Mohm so your results shouldn't differ. In general scope readings are less accurate than any meter, in particular if your scope has a 8bit adc.

About variable tone beeping, in Art of Electronics II edition, page 954, there is a circuit useful to discover shorts without desoldering parts, measuring millivolts (caused by an external current source). This circuit uses a meter, but I remember somewhere I saw a more interesting version that emitted a proportional tone instead and had an internal current source.

[Wallace Gasiewicz]

Just for attempting to be complete:
My old Fordee PSM 37 is analog and has the low voltage ohms. Manual states it is .094 Volts for Low Ohms.
It has a 20 ohm mid scale reading so it is not that great for anything below 5 ohms.
This unit is essentially a JFET input device, so it has characteristics of both the usual modern multimeters with an analog meter.

Wally

[AVGresponding]

AVGresponding:
This is very puzzling then. I have not been able to find the schematic of either meter to check, yet it would be very strange if this was intentional. In this case, you would almost surely see a step change.

Neutrion:
If you used a regular x10 probe with your scope, its input Z should also be 10Mohm so your results shouldn't differ. In general scope readings are less accurate than any meter, in particular if your scope has a 8bit adc.

About variable tone beeping, in Art of Electronics II edition, page 954, there is a circuit useful to discover shorts without desoldering parts, measuring millivolts (caused by an external current source). This circuit uses a meter, but I remember somewhere I saw a more interesting version that emitted a proportional tone instead and had an internal current source.

You'd be unlikely to see a step change because of the very low currents involved, 41 nanoamps and 55 nanoamps for the Agilent and Metrix respectively, and the very high impedance (≥10GΩ) means that even with very low input capacitance it takes a while for the indicated voltage to drop.

[Kleinstein]

The high Z meters have a high input resitance, but often also some input capacitance that may start in some not so well defined state. So one can still see an initial change in the voltage from the capacitance and just a possible slow setting to a final voltage. For the open circuit voltage the ouput impedance of the meter under test may no longer be that high in resistance even with a low current setting.

[AVGresponding]

Here's a couple of pics of the BM869S high ohms output as seen by an oscilloscope.

The high Z meters have a high input resitance, but often also some input capacitance that may start in some not so well defined state. So one can still see an initial change in the voltage from the capacitance and just a possible slow setting to a final voltage. For the open circuit voltage the ouput impedance of the meter under test may no longer be that high in resistance even with a low current setting.

Yes, I might have a go at measuring the input capacitance, since it's not specified in the manual (for DC anyway, AC specifies <100pF).

[ermionesrl]

Kleinstein, AVGresponding:

What I expect to be inside the multimeters is a voltage limited, step controlled current source. The current is controlled (influence measurement) and different on many ranges, while the voltage is just limited (by giving a limited voltage input to the current source) and has not any requirement for accuracy or stability (in the 34401A the voltage is derived from +18V power with a simple 5.1V zener, I don't think in the U1401B took care of deriving it from a reference).

Unless there is a reason (like not to push semiconductors in conduction on some ranges, or the opposite) I expect it to be the same for all ranges even if might seem to have changed, in case multimeter insert a parallel resistor in highest range.

About falling time, if total capacitance is 10nF and total "leakage" resistance is 10Gohm, we would have a tau of 100s. Long, in effect, but I suspect both values being lower, in particular capacitance, so you should see voltage to settle in a reasonable time, if it is really stabilized.

It would be interesting to test again the U1401B with different condition batteries, to see if and how much the testing voltage change, as well as in the last but one range.

[Kleinstein]

For the input capacitance I would expect maybe 100 pF to 1 nF. When at the limit the current source would also be no longer very high impedance. This is what limits the voltage.
depending on how the current shouce is made, the voltage may change. A common configuration is to start the current source from the positive supply and than get a defined drop on a switchable resistor. The voltage drop can change with ranges, with possibly less voltage lost in the lower current ranges in a way of switching rangs.  The 34401 uses such a circuit. So there can be a slightly different maximum voltage with the ranges.

Most handheld meters have a fixed voltage (like 200 mV with 2000 count) for the full scale reading on the ranges.
The fixed range corresponds to the configuration with a fixed votlage range that can be measured without a divider in front and thus with very high impedance.
Better bench DMMs have a higher voltage with the higher resistance ranges (e.g. 10 V for 20 M). This avoids very small current sources for the high ohms part.

[David Hess]

A common configuration is to start the current source from the positive supply and than get a defined drop on a switchable resistor. The voltage drop can change with ranges, with possibly less voltage lost in the lower current ranges in a way of switching rangs.  The 34401 uses such a circuit. So there can be a slightly different maximum voltage with the ranges.

In older multimeters, the configuration I regularly see bootstraps the switchable input divider (1k, 10k, 100k, etc.) to apply a fixed voltage across it.  So the operational amplifier buffers the voltage at the high input, and applies that voltage plus a fixed voltage to the switchable input divider and applies it back to the high input.  The example below is comparable.

Autoranging multimeters which use a grounded input divider must be doing it in a different way.

[AVGresponding]

Well, I haven't charged the battery in the U1401B for some time, so I'll top it off, but I doubt it'll make any measurable difference.

I'm warming up my main LCR meter for a couple of hours, then I'll get a decent reading on the Fluke 8840A input capacitance.

Tried to get a nS reading on the Fluke 8840A to work out the input impedance, but both the Fluke 289 and Fluke 87V gave a settled reading of 0.00nS, and the Brymen 869S wouldn't settle, fluctuating between 0.02 and 0.07nS. Either way, It looks like the ≥10GΩ impedance is more > than =.

Also had to edit the spreadsheet, as I realised the Tek 912 had a LV ohms mode...   



EDIT: The Agilent U1401B with freshly charged battery peaked at 4.2788V, with an identical 41nA current. The voltage drops off at two or three hundred microvolts per minute, much more slowly than if the Fluke 8840A had no input.

EDIT 2: It looks like I should have had more patience the first time around with the Agilent!     It stabilised at 4.2767V after around 10 minutes, then climbed back to the original figure over the next 20 minutes.
I'll leave it to run for a while and see where it ends up. Looks like it's just drifting around a bit under warm-up.

EDIT 3: It peaked at 4.3001V after about 3 hours and 40 minutes, then started a very slow (10-20uV/minute) drift down again. Not really a practical warm up time for a battery DMM! The current was still 41nA.
Naturally I had to unplug from the Fluke 8840A measuring voltage to plug into the Keithley 197 measuring current to check that, and when I plugged back into the 8840, the voltage went to 4.2995 pretty quickly. I'm confident to call it warm up drift and nothing else at this point.

EDIT 4: Hopefully the last edit for this post, but well...  Tried measuring the input capacitance of the Fluke 8840A; much more problematical than I expected. My main LCR meter (Iso-Tech LCR819) says slightly under 51nF, however checking with other meters to look for a mean value threw up various non-responses. Fluke 87V and Fluke 289 both went over-range,  Metrix 57 said about 430nF, Brymen 869S and Tek 912 both cycled through capacitance ranges without giving a reading, and the 328-based component tester said 150uF...
I have recently acquired a GenRad 1657 DigiBridge but I haven't yet had a chance to fettle it and check its reliability, so haven't tried it with that.

[ermionesrl]

AVGresponding:
I don't think you will ever succeed with a LCR meter. I would try to use the oscilloscope. Knowing the Z of a x10 probe is 10Mohm, put any voltage source on the meter, disconnect it abruptly (even by hand, we aren't on the picoseconds, but a reed relay, or a series of them not to add further capacitance, could help ) measure the decay time and calculate C from it.

I'll try later. It should work even with 10Mohm input multimeters, just knowing the resulting resistance will be 5Mohm. To measure effective input capacitance of the ohm function could be much more tricky.

[AVGresponding]

I'll give it a try with the 289 in parallel with the 8840, and use the data logger function of the 289, should give broadly similar results.

[Kleinstein]

I just saw one more case where a relatively low test current in the ohms mode is helping / needed: The on-resistance of JFETs is somewhat nonlinear and thus changes with current if the votlage grop gets high. I just got one such example reading 1.02 K and 1.14 K depending on the direction - so the effect is visible, though still good enough. In this example the test current was around 0.19 mA.

[Neutrion]

Here's a couple of pics of the BM869S high ohms output as seen by an oscilloscope.

The high Z meters have a high input resitance, but often also some input capacitance that may start in some not so well defined state. So one can still see an initial change in the voltage from the capacitance and just a possible slow setting to a final voltage. For the open circuit voltage the ouput impedance of the meter under test may no longer be that high in resistance even with a low current setting.

Yes, I might have a go at measuring the input capacitance, since it's not specified in the manual (for DC anyway, AC specifies <100pF).

Thanks for checking! It seems that the scope reading is about the same as the high Z multimeter reading.
So maybe my scope reading is correct as well.

I just saw one more case where a relatively low test current in the ohms mode is helping / needed: The on-resistance of JFETs is somewhat nonlinear and thus changes with current if the votlage grop gets high. I just got one such example reading 1.02 K and 1.14 K depending on the direction - so the effect is visible, though still good enough. In this example the test current was around 0.19 mA.

So we are having PTC/NTC and JFET so far. Never any problems with micros? With microscopic components I would think the effect should be more dramatic.

[ermionesrl]

AVGresponding, Kleinstein:
I finally took the time to do the measurement as promised. I tested the 34401A input in >10G DCV mode with oscilloscope and 10V source. Fall time seems consistent enough and average to 16.5ms, so input capacitance should be 750pF (fall time defined as 90% to 10% and tau~=fall time/2.2). It's a reasonable value, so perhaps the experiment is correct. I tried 10M DCV mode and fall time got to 8.2ms as expected, I tried ACV mode and (this was expected but not to this extent) 34401A is awful here with a fall time of 5.14s, that translates to an effective 240nF, not according to 34401A specs that say 1Mohm in parallel with 100pf. In the schematic (page 158 of the document 34401-90013) I see C301, a 220nF capacitor in series with a 1Mohm resistor going to a OPAMP virtual GND, thus we are reading the 220nF discharging to an effective R of 11Mohm, measuring it through a resistance divider 11M/10M. Taking everything in account, effective measured C is 230nF. Good, it's C301.

I also tested the Fluke 177. DCV: wonderful 590us resulting to a 53pF (effective Zin 5Mohm). Here the oscilloscope probe should be taken in a account with its 15pF, so we have 38pF, something less because also the cables count in this case. ACV: the circuit looks similar to the 34401A, but with a 10Mohm in series to a C that looks to be 12nF.

[AVGresponding]

You had better luck than me! Trying it with the F289 data logger yielded no usable results. I might try it with a scope this weekend if I get time.

[Alex Nikitin]

FWIW, my MetraHit 28S has the o/c voltage on all Ohms ranges ~0.62V at 23C room temperature, dropping to 0.3V with 10M connected on 30M range. The current on 300 Ohm range is ~0.2mA . The diode test mode has ~2.9V o/c (which I suspect is just the battery voltage at the time of measurement) and <1mA current (looks like a 3K resistor in series).

Cheers

Alex

[BlackFX]

Hioki 3238
Ohms 6v Max, 20nA (100meg range) - 1mA (200ohm range)
Low Ohms 0.45v Max, 100nA (2000k range) - 100uA (2000ohm range)

[alm]

I tried ACV mode and (this was expected but not to this extent) 34401A is awful here with a fall time of 5.14s, that translates to an effective 240nF, not according to 34401A specs that say 1Mohm in parallel with 100pf. In the schematic (page 158 of the document 34401-90013) I see C301, a 220nF capacitor in series with a 1Mohm resistor going to a OPAMP virtual GND, thus we are reading the 220nF discharging to an effective R of 11Mohm, measuring it through a resistance divider 11M/10M. Taking everything in account, effective measured C is 230nF. Good, it's C301.

That's the AC coupling capacitor in series that has forms a high-pass filter together with the input impedance. The ACV range is specified from 3 Hz - 300 kHz. At 3 Hz, the 220 nF capacitor has an impedance of about 240 Ohms. So I'm sure this won't affect the 1 MOhm +/- 2% spec. A fall time of 5.14s is well within the spec for a 7 second settling time in AC mode with the filter set to slow (see attached table from page 51 of the 34401-90004 user manual). Although I'm not sure how this relates to Ohms measurement?

[ermionesrl]

alm:
Yes, the 220nF capacitor is not a problem in real ACV measurements. It has nothing to do with ohm function but I found interesting it's possible to infer much about the input circuit with this simple setup.

It was interesting to know multimeter input capacitance to understand a supposed anomalous settling time in a measurement done by AVGresponding. Of course ideal would be to measure capacitance in ohm function, but I think it's very difficult to do it, so I did in DCV just to have at least an idea of the order of magnitude of the value.

[HKJ]

It was interesting to know multimeter input capacitance to understand a supposed anomalous settling time in a measurement done by AVGresponding. Of course ideal would be to measure capacitance in ohm function, but I think it's very difficult to do it, so I did in DCV just to have at least an idea of the order of magnitude of the value.

Capacity in DCV basically has nothing to do with ohm ranges.
I explain a bit about multimeters here: https://lygte-info.dk/info/DMMDesign%20UK.html
and here: https://lygte-info.dk/info/DMMDesignProtection%20UK.html#Ohm/Continuity/Diode/Capacity

The second link shows the typical input circuit for ohms (and other ranges), the first link shows the typical measurement circuit.

[AVGresponding]

It was interesting to know multimeter input capacitance to understand a supposed anomalous settling time in a measurement done by AVGresponding. Of course ideal would be to measure capacitance in ohm function, but I think it's very difficult to do it, so I did in DCV just to have at least an idea of the order of magnitude of the value.

Capacity in DCV basically has nothing to do with ohm ranges.
I explain a bit about multimeters here: https://lygte-info.dk/info/DMMDesign%20UK.html
and here: https://lygte-info.dk/info/DMMDesignProtection%20UK.html#Ohm/Continuity/Diode/Capacity

The second link shows the typical input circuit for ohms (and other ranges), the first link shows the typical measurement circuit.

Given the varied responses by different meters, I think it's fair to say that some at least may be atypical, in implementation if not broad theory of operation.

[HKJ]

Given the varied responses by different meters, I think it's fair to say that some at least may be atypical, in implementation if not broad theory of operation.

That is correct, but all DMM's has a high-z voltage input and at least one low-z path (current output) in ohms (Together with a very high z input path), i.e. very different.

[jasonRF]

I recently picked up a Fluke 8012a off of ebay for this feature, but unfortunately it doesn't work as well as the listing indicated...

Next time Dave collaborates with a company for a new DMM, he should include it.  I would buy one!

jason

[J-R]

I recently picked up a Fluke 8012a off of ebay for this feature, but unfortunately it doesn't work as well as the listing indicated...

Did you already try blasting the switches with contact cleaner and working them back and forth a bunch?