DMM Input Capacitance

[mawyatt]

This was precipitated from a measurement here:

https://www.eevblog.com/forum/testgear/sds800x-hd-actual-use-cases/

Curious what others find when measuring bench DMM +- input terminal capacitance wrt to ground.

Best,

[KungFuJosh]

What's your test setup for this? How are you referencing ground?

Thanks,
Josh

[2N3055]

Meter used Uni-T 612 LCR meter.
100kHz, C mode
Negative terminal connected to PSU ground terminal (it was just convenient)
Banana cables, calibrated in approximate position of measurement.
Grounded metal plate was simply top of other grounded instrument.

Absolute values are far from being accurate, but relative values should be good for relative comparisons.

Rigol DM3068 (T+ T- are positive and negative input terminals)

+ T 30pF
 - T 180pf

This is one spec where high end handheld will be "more" differential than benchtop meter for small AC signals...

Metrix MTX3293 :

sitting on top of isolated desk is 5pF both
siting on antistatic mat 15pF
sitting on grounded metal plate 19pF both.

Brymen is a bit worse, but still better than benchtop and also if kept isolated (by distance to ground and NOT holding it in hand) absolute values are order of magnitude better:

BM869S :
+T 4,8pf    -T 6,4pf
+T 10,4pF  -T 14,4pF
+T 12pF    -T 16pF

Best,

[mawyatt]

@ Josh,

Just used a handheld LCR meter (DE-5000) @ 10KHz to stay away from mains frequency influence. Connect handheld LCR meter from DMM + or - terminal to chassis ground (earth mains ground in our setup).

Best,

[mawyatt]

@ 2N3055,

Important point you made, measurements are relative not absolute...thanks for pointing that out, we should have done so in the 1st post!!

The SDM3065X is the worst of our bench DMMs with ~500pF from - terminal and ~93pF from + terminal at 10KHz.

Ironically the old HP and AG 34401A are best we have wrt this!!

Another interesting observation while doing this capacitance test, note the difference in direct DMM +- input capacitance when in DCV and ACV modes!!

Best,

[floobydust]

I measure 10-40pF across the jacks of 5 multimeters on DCV. Most are around 35pF except 34401a measured 680pF quite high. Adding test leads does not add much.
34401a 177pF and 112pF to PE.

I find DMM input capacitance an issue if measuring low impedance HVDC, say 500V I've seen multimeters blank out or crash when connecting the (+) probe and even a tiny spark can be seen. It's a bit scary. A few product recalls over this (Fluke included) where the MCU reboots due to the EMI of that.

Also important is the common-mode capacitance, the other electrode being the multimeter's shield and your hand, or your workbench, or the grounded metal the multimeter is resting on. The input capacitance from each input (+), (-) to shield is not symmetrical. Mains-powered bench multimeters can have a Y-cap to PE.
This is a problem at RF especially if the DMM is cheaper and has no foil shield.

[2N3055]

I measure 10-40pF across the jacks of 5 multimeters on DCV. Most are around 35pF except 34401a measured 680pF quite high. Adding test leads does not add much.
34401a 177pF and 112pF to PE.

I find DMM input capacitance an issue if measuring low impedance HVDC, say 500V I've seen multimeters blank out or crash when connecting the (+) probe and even a tiny spark can be seen. It's a bit scary. A few product recalls over this (Fluke included) where the MCU reboots due to the EMI of that.

Also important is the common-mode capacitance, the other electrode being the multimeter's shield and your hand, or your workbench, or the grounded metal the multimeter is resting on. The input capacitance from each input (+), (-) to shield is not symmetrical. Mains-powered bench multimeters can have a Y-cap to PE.
This is a problem at RF especially if the DMM is cheaper and has no foil shield.

This is about common mode capacitance, i.e.  measuring from T+ of the meter to ground (P.E.) and then comparing it to T- to ground (P.E.) in your lab.

[floobydust]

I wasn't sure about interpreting "when measuring bench DMM +- input terminal capacitance wrt to ground."
CM wrt to earth-ground it varies widely with handheld meters, depending on where they are, in your hand, on the floor, foil shield etc.
It's always imbalanced, (-) input is higher capacitance I believe because it goes all over the circuit board and not like (+) going to the divider resistor.

Bench multimeters you are basically measuring the power transformer winding capacitance or their Y-cap.

[2N3055]

I wasn't sure about interpreting "when measuring bench DMM +- input terminal capacitance wrt to ground."
CM wrt to earth-ground it varies widely with handheld meters, depending on where they are, in your hand, on the floor, foil shield etc.
It's always imbalanced, (-) input is higher capacitance I believe because it goes all over the circuit board and not like (+) going to the divider resistor.

Bench multimeters you are basically measuring the power transformer winding capacitance or their Y-cap.

That is why i wrote:

sitting on top of isolated desk is 5pF both
siting on antistatic mat 15pF
sitting on grounded metal plate 19pF both.

And made a comment of not holding it in a hand.

Yes we know why benchtop meters are asymmetric in this regard.
Mike wanted to make this fact popularized by discussing it.

[mawyatt]

I wasn't sure about interpreting "when measuring bench DMM +- input terminal capacitance wrt to ground."
CM wrt to earth-ground it varies widely with handheld meters, depending on where they are, in your hand, on the floor, foil shield etc.
It's always imbalanced, (-) input is higher capacitance I believe because it goes all over the circuit board and not like (+) going to the divider resistor.

Bench multimeters you are basically measuring the power transformer winding capacitance or their Y-cap.

That is why i wrote:

sitting on top of isolated desk is 5pF both
siting on antistatic mat 15pF
sitting on grounded metal plate 19pF both.

And made a comment of not holding it in a hand.

Yes we know why benchtop meters are asymmetric in this regard.
Mike wanted to make this fact popularized by discussing it.

And just how asymmetrical they can be!!

Also, how they can influence certain measurements in peculiar subtle ways as shown over here:

https://www.eevblog.com/forum/testgear/sds800x-hd-actual-use-cases/

Best,

[KungFuJosh]

I have a convenient outlet on my bench, so I used a hex key inserted as my ground reference.

I tested with my ST2832 (setup image attached - Cs-Rs @ 100kHz) and I will likely also test with my DE5000.

SDM3055X-E:
DCV:
Negative Terminal: 1.47615pF
Positive Terminal: 0.55323pF

ACV:
Negative Terminal: 7.49180pF
Positive Terminal: 0.44843pF

SDM3065X:
DCV:
Negative Terminal: 1.60443pF
Positive Terminal: 0.38473pF

ACV:
Negative Terminal: 1.31416pF
Positive Terminal: 0.31178pF

I felt it was better to reference ground in the socket, rather than letting the chassis interfere. Also no inconsistencies between ground reference if switching devices ground ref point. And I suppose the Probe Master probes increase the capacitance too. 🤷

Thanks,
Josh

[mawyatt]

Josh,

Those measurements seem very low. Try using your DE-5000 instead at 10KHz like we did. You can use short banana plugs & cables directly with DE-5000, and plug the negative DE-5000 end into your power supply Green ground terminal like 2N3055 did, or just connect to something that has earth ground.

No need to calibrate the DE-5000, we are just looking for relative capacitance, cable capacitance and such shouldn't matter.

Best,

[KungFuJosh]

What's wrong with referencing earth at the outlet?

[mawyatt]

What's wrong with referencing earth at the outlet?

Nothing if you get the Allen Wrench into correct terminal 

Best,

[floobydust]

I wasn't sure about interpreting "when measuring bench DMM +- input terminal capacitance wrt to ground."
CM wrt to earth-ground it varies widely with handheld meters, depending on where they are, in your hand, on the floor, foil shield etc.
It's always imbalanced, (-) input is higher capacitance I believe because it goes all over the circuit board and not like (+) going to the divider resistor.

Bench multimeters you are basically measuring the power transformer winding capacitance or their Y-cap.

That is why i wrote:

sitting on top of isolated desk is 5pF both
siting on antistatic mat 15pF
sitting on grounded metal plate 19pF both.

And made a comment of not holding it in a hand.

Yes we know why benchtop meters are asymmetric in this regard.
Mike wanted to make this fact popularized by discussing it.

My interest is when the (differential) input or (common-mode) stray capacitance causes problems while making measurements out in the real world.

After a few decades of using many different multimeters and seeing a few malfunction over this capacitance/effect I know when it's a problem. But it's not discussed and people have difficulty understanding it. Very few multimeters have formal EMC immunity test certs or are subjected to EMI in a lab.

Another thing to mention is using a mains-powered LCR meter to measure capacitance to PE does not always work.
Either you get CM noise on the in-guard section from the isolator/SMPS, or it's swamped by a Y-cap, something to look out for.
With any mains-powered LCR meter or the mains-powered multimeter, people think their inputs are floating in outer space yet some gear has bad parasitics. You have to know your test equipment in this undocumented regard.

[Sariel]

Why is a DMM with very low input capacitance favoured?

[2N3055]

Josh,

Those measurements seem very low. Try using your DE-5000 instead at 10KHz like we did. You can use short banana plugs & cables directly with DE-5000, and plug the negative DE-5000 end into your power supply Green ground terminal like 2N3055 did, or just connect to something that has earth ground.

No need to calibrate the DE-5000, we are just looking for relative capacitance, cable capacitance and such shouldn't matter.

Best,

I agree these are very low. It might be because LCR meter is mains powered.
I also want to see DE-5000, which is handheld like my UNI-T.

Floobydust made some good points..

[KungFuJosh]

Josh,

Those measurements seem very low. Try using your DE-5000 instead at 10KHz like we did. You can use short banana plugs & cables directly with DE-5000, and plug the negative DE-5000 end into your power supply Green ground terminal like 2N3055 did, or just connect to something that has earth ground.

No need to calibrate the DE-5000, we are just looking for relative capacitance, cable capacitance and such shouldn't matter.

Best,

I agree these are very low. It might be because LCR meter is mains powered.
I also want to see DE-5000, which is handheld like my UNI-T.

Floobydust made some good points..

I'll test with the DE5000 next. I'm curious, you think I should dig out my isolation transformer and power the LCR from that to see if it changes anything?

What's wrong with referencing earth at the outlet?

Nothing if you get the Allen Wrench into correct terminal 

🤯 How is it that there's no electrocuted emojis on this site??

[KungFuJosh]

Results with DE5000 (using kelvin clips) at 10kHz Cp, all else the same (ground ref in power outlet, Probe Master probes):

SDM3055X-E:
DCV:
Negative Terminal: 560.5pF
Positive Terminal: 66.26pF

ACV:
Negative Terminal: 559.9pF
Positive Terminal: 57.25pF

SDM3065X:
DCV:
Negative Terminal: 192.04pF
Positive Terminal: 52.36pF

ACV:
Negative Terminal: 190.88pF
Positive Terminal: 40.63pF

[alm]

It makes sense that on a guarded multimeter (which I'd expect any halfway-decent bench meter to be), the impedance of the low terminal to ground is much lower than the high terminal. Generally the guard is connected to the low terminal except on meters with a separate guard terminal where you generally want to connect guard to the low terminal unless you are connecting it separately. And the guard is a big box (or at least a trace) between the input circuit and ground. So it should have a lower impedance to ground to do its job of shunting common mode current away to ground. In terms of impedance to ground or noise, you should treat your DMM a bit like a scope in that you connect the high lead to the high impedance point of the circuit and the low terminal to a low impedance point of the circuit. Connecting the low terminal to a high impedance point (like a high-impedance voltage divider) can lead to increased noise and unexpected circuit loading.

SDM3055X-E:
DCV:
Negative Terminal: 1.47615pF
Positive Terminal: 0.55323pF
...

These figures look off by at least two orders of magnitude. Figures like this would be good for an active FET scope probe, and unthinkable for a DMM. It may be either that the ground-referenced LCR meter was shorting it to ground, or maybe there was no connection between the Allen key and the DMM's ground.

[Alex Nikitin]

Measuring parasitics of multimeters is a very healthy exercise! 

The capacitance would also depend on the voltage range setting, when the input divider is connected the capacitance from the "+" input terminal might drop significantly. My measurements with a hand held capacitance meter (1kHz) for the DC voltage mode only:

Keithley 2015

Positive input to mains ground - 10V range and below ~100pF, 100V and 1000V ~25pF
Negative input to mains ground ~ 180pF
Between input terminals  ~300pF for 10V range and below, ~20pF for 100V and 1000V ranges

HP3456A

Positive input to mains ground - 10V range and below ~350pF, 100V and 1000V ~85pF
Negative input to mains ground ~ 1nF with the guard open, ~2.2nF with the guard to negative
Between input terminals  ~600pF for 10V range and below, ~60pF for 100V and 1000V ranges

Hioki DM7275-03

Positive input to mains ground - 10V range and below <30pF, 100V and 1000V <10pF
Negative input to mains ground  <30pF
Between input terminals  ~1500pF for 10V range and below*, <10pF for 100V and 1000V ranges.

Cheers

Alex

* - possibly something to do with the contact checking feature, even if it is OFF. The cure is simple (if the input C is a problem) - to put a 10K resistor in series with the positive lead and use Hi-Z, on 10V and below the meter has ~20pA input current or ~0.2uV added error with the resistor.

[KungFuJosh]

The capacitance would also depend on the voltage range setting, when the input divider is connected the capacitance from the "+" input terminal might drop significantly. My measurements with a hand held capacitance meter (1kHz) for the DC voltage mode only:

Interesting. I tried that on my SDM3055X-E, and the DCV ranges did have a small impact. 20V to 1000V was 10pF higher than 200mV to 2V.

No effect at all for ACV.

[mawyatt]

The capacitance would also depend on the voltage range setting, when the input divider is connected the capacitance from the "+" input terminal might drop significantly. My measurements with a hand held capacitance meter (1kHz) for the DC voltage mode only:

Interesting. I tried that on my SDM3055X-E, and the DCV ranges did have a small impact. 20V to 1000V was 10pF higher than 200mV to 2V.

No effect at all for ACV.

What do you see when switching from DCV to ACV wrt the capacitance between terminals and + or - terminal to ground?

Best,

[KungFuJosh]

Interesting. I tried that on my SDM3055X-E, and the DCV ranges did have a small impact. 20V to 1000V was 10pF higher than 200mV to 2V.

No effect at all for ACV.

What do you see when switching from DCV to ACV wrt the capacitance between terminals and + or - terminal to ground?

All previous tests were terminal to ground.

SDM3055X-E between terminals:
Auto (200mV):
DCV: 66.27pF
ACV: 58.54pF


SDM3055X-E Positive Terminal to Ground:
Auto (200mV):
DCV: 62.61pF
ACV: 54.18pF

Interesting that there's 4pF less ref to ground vs. terminal to terminal.


SDM3055X-E Negative Terminal to Ground:
DCV: 559.2pF
ACV: 559.3pF

[ebastler]

I have a convenient outlet on my bench, so I used a hex key inserted as my ground reference.

I kind of like the idea of having an outlet right in the benchtop for testing. (My main power strip is below the benchtop, hence a bit inconvenient to reach when I want to plug/unplug a DUT often.) In practice, how often do you drop clipped wire ends and stuff into the mains and USB outlets?