off topic: multimeter performance

[Calamar]

I bought a multimeter (40.000 counts true RMS) a year ago,
at the very beging I notice a large fluctuation over ohm
range, 50 counts, large nosie lecture over a 100K resistor (AC mnode),
200mV , and a clock signal contecting the tester
leads over osciloscope leads, after a lots repley the following
is the repley from product manager like a correct spect
performance, they stolen me right ?

------------------------------------------------------------------------


APP-207 Fax Reply


Tested on a APPA-207 :
1Mohm Resistance Test:
Settings:
1. AC 4V Range (1Mohm resistance division)
2. DC Powered

Readings:
1. Test Leads removed from Meter:
0.0023 (2.3mV) stable reading
2. Test Leads connected to Meter :
0.0928 (92.8mV) stable reading
3. Test Leads connected to Meter with hands holding on probes:
0.1657 (165.7mV) stable reading


100Kohm Resistance Test:
Settings:
1. AC 40V Range (100Kohm resistance division)
2. DC Powered

Readings:
1. Test Leads removed from Meter:
0.003 (3mV) stable reading
2. Test Leads connected to Meter :
0.097 (97mV) stable reading
3. Test Leads connected to Meter with hands holding on probes:
0.197 (197mV) stable reading

Suggestions :
1. According to the test results above, it proves the meter is not able to conduct an internal clock
which signaling out to ambient environment.

2. The noisy readings are environment signals when the test leads are connected to a meter

3. The resistance of input terminals on ACV is approx. 10Mohm , so the readings are at the same
testing conditions with different ranges, which means both readings are almost the same.

4. Noisy rejection measurement methods is required for high impedance measurements sometimes in order to eliminate this problem!
_________________
Best Regards
Daniel H. Sagarra
La Plata (Argentina)

[rwyoung]

There is a bit of a language barrier here but I think you are describing the following:

Resistance range - drifting
AC voltage mode - noise pickup
any mode, clock injection via meter leads.

It is not uncommon for DMMs to inject clock noise into the circuit they are measuring. I've seen this on cheap meters. I've seen this on multi-thousand dollar meters. The meter itself should reject the clock noise as it is synchronous to the measurement. If it proves to be an issue with testing, I have modified probes in the past to include a passive low-pass filter in-line with the "red" lead.

Noise pickup is also very common. The simplest thing you can do is twist your test leads together to eliminate any large loops. Most test leads are not sheilded (they are available) and you can get EMI pickup. Typically power line frequencies or multiples from flourescent lights. I have also seen pickup from low frequency DC/DC supplies. AC powered meters should have zero-crossing detection circuts to let them integrate readings over one or more power supply line cycles. Battery powered meters may be attempting this by simply integrating for some "arbirary" period of time.

Resistance mode may just be flakey on your particular meter. The best results would be a true four-wire "Kelvin" connection. Or a bridge. But in your case check the following:

1) Internal resistance of meter. Use shortest possible length of wire to bridge the ohmmeter inputs. Hope that the reading is very low and reasonably steady.
2) Use good quality test leads and short together to estimate their contribution. Twist leads together to minimize loop induced noise pickup.
3) If possible, don't use autoranging on the meter. For values that fall close to the transition points of the ranging algorithm, you can get very poor readings in one range but a better (stable) reading in the next range up or down as appropriate.

Your meter may just have a poor resistance measurement circuit. While DMMs are nice, they can't be all things to all people all the time. Some just work better in DC or AC or voltage or current mode than in every mode on the dial.

If your meter is AC powered, try testing it in a different location that you know is using a different power substation. Also, test the meter in an electrically "quiet" area if possible. The meter will generally work better with a clean power supply. Other equipment on the same branch flanging the line around and injecting noise may be causing some trouble. This is where a good battery powered meter can shine.

Most DMMs have relatively low input impedance of 10Mohm. It is a compromise between not loading the DUT and not picking up noise and minimizing Johnson (thermal) noise. Of course in current mode, the input impedance would be much lower, fractions of an Ohm in general. Some DMMs have the option to switch to a much higher input impedance like 10GOhm for more sensitivity at the expense of a bit more noise pickup.

By holding the probes in your hands, you are acting as an antenna for pickup of radiated noise. Some simple fixes are:

1) Don't touch the exposed metal (seems obvious)
2) Wash your hands to remove skin oils. Dry well.
3) Scrub down the insulation on your meter leads to remove residual oils (amazing how nasty the they can be).
4) Change brands of leads. Pomona leads are generally high quality but look for ones intended for use with 6 1/2 digit or better DMMs
5) See above comments about keeping the leads twisted together to eliminate loops for induced current noise.
6) Look around your bench area for DC/DC supplies or flourescent lights that are temendous noise sources.
_________________
Rob Young
The Screw-Up Fairy may just visit you but he has crashed on my couch for the last month!

[Calamar]

[rwyoung]

One more thing, I limit my DMM useage to HP (Agilent), Keithley and Fluke brands. I have some other "cheap" ones but I don't trust any number they report. They are simply used for GO/NO GO testing (is it near 0V or near 5V, etc).

Also have a few old Simpson and HP analog voltmeters. VERY LOW NOISE!

Good luck with your DMM
_________________
Rob Young
The Screw-Up Fairy may just visit you but he has crashed on my couch for the last month!

[SherpaDoug]

I agree with rwyoung that there is a lot to a quality multimeter that is not in the datasheet. I suggest:

1) Stick with a brand you trust. For me that is Fluke, Simpson, Triplett, or Beckman.

2) Avoid autoranging meters.

3) Use short high quality leads. I have many specialized meter leads I have made for specific uses. Remember the old VTVMs (vacuum tube volt meters) had a 10Meg input impedance plus a 1Meg resistor in the probe tip. The 1Meg in the probe tip reduced noise from the lead being injected into the curcuit, but required the meter to be calibrated to see 10/11 of the applied voltage.

4) Understand how your meter works and what its weaknesses are. What does a True RMS meter read when given DC? I had a nice Beckman DVM 20 years ago that read the RMS value of whatever it was given. Give it 9VDC and it read 9V. Give it 6VAC and it would read 6V. Give it both and it would read the proper sum. My newer Beckman 4410 strips the DC out of its RMS readings, which I do not like. I have a Triplett 310 (analog) that is not very true RMS but at least it reads AC + DC about right.

No tool is perfect. You have to learn to deal with it.

FYI my battery powered Beckman 4410 reads reads about 0.3mVAC with no leads sitting on my bench, 0.05mVAC shorted. My AC powered Fluke 8810A reads 0.003mVAC no leads/ 0.001mVAC shorted.
_________________
The search for better is endless. Instead simply find very good and get the job done.

[Calamar]

I've also a fluke 87 serie IV, using it in the same place
and, with the same tester leads, over 100k resistor
it have SNR almost of 60 dB better than my APPA-207,

I usually had to calibrate a pick-up circuit sensibility, the scheme is the
following:


SenoidalFunctionGenerator-->coil*magnetic coupled*coil->pick-up circuit
^
|
|
---tester leads.(APPA-207, Fluke 87,... )

when I use the APPA-207, and only with it, the amount of noise
coupled to the pick-up circuit case a several KHz triggers,
just like some clock of the APPA-2007 were present at tester leads...
_________________
Best Regards
Daniel H. Sagarra
La Plata (Argentina)

[Calamar]

The tester leads are over the first coil not the generator
_________________
Best Regards
Daniel H. Sagarra
La Plata (Argentina)

[Calamar]

Sorry, the second coil
_________________
Best Regards
Daniel H. Sagarra
La Plata (Argentina)