Earth Ground Resistance & Resistivity Testers

[ocw]

There has been some discussion about earth electrode resistance recently.  See:
https://www.eevblog.com/forum/beginners/electrical-test-insulation-continuity-rccb-polarity-and-earth-electrode/msg1045847/?topicseen#msg1045847

Due to that I thought that I'd relay some accuracy data which I've measured on several low to moderate cost earth resistance and resistivity testers this year.  See the attached spreadsheet.  I should be adding data on another clamp tester in another week or two.

Earth resistance measurements are most commonly made via the three point method.  An AC current is injected between the earth electrode being tested and a short test probe 10 - 50+ meters away from the electrode.  For the most accurate readings a third probe is placed 62% of the way from the earth electrode and the distant probe.  The voltage is then measured between the earth electrode and the middle probe.  The earth resistance is then calculated based on that voltage and the mentioned injection current.  A better detailed explanation of this is in AEMC's "Understanding Ground Resistance Testing."  See:
http://www.aemc.com/techinfo/techworkbooks/Ground_Resistance_Testers/950-WKBK-GROUND-WEB.pdf

A common problem with the above measurements is that the electrode being tested is connected to the ground wire on the commercial power feeding the building.  That places the earth electrode in parallel with the extensive grounding on the commercial power lines.  That includes the power entry ground and all of the ground rods connected as the base of the tower poles (or at regular distances underground).  Often that commercial grounding is then the main influence on the reading--not the earth electrode being tested.  That electrode needs to be disconnected from commercial power for an accurate reading to be made on just it.  Due the fast rise time of a lightning strike, the connecting wire inductance to the commercial power ground makes that lower resistance ground not very effective in dealing with a lightning surge.

Often disconnecting the earth electrode from the commercial ground is not wise unless the commercial power is completely shut down.  A way around this is to use a clamp earth resistance meter.  It can measure the earth resistance without disconnected the commercial power ground when it is clamped around the wire just going to the earth electrode being tested.  The commercial ground is still connected in the other direction on the same wire.  Because of that, what is actually being measured is the earth electrode's earth resistance in series with the earth resistance of the commercial power ground.  Since the commercial power's earth resistance is generally so much lower than that of the tested electrode, that can be ignored.  Or you can get even closer by subtracting the earth resistance obtained by a three point measurement of all grounds (the single tested electrode generally doesn't have a significant influence on that). 

Four point testers can also measure earth resistivity/ground conductivity.  Four equally spaced short probes are used and the voltage between the center two probes is used along with the AC current to calculate a resistance.  That resistance is used in a formula along with the probe separation distance to calculate the earth resistivity or its inverse, the ground conductivity.

Regarding the attached spreadsheet, it is based on measuring precision resistors with the meter.  Readings obtained on my Agilent 34401A are included for reference.  The extra resistance needed for the test wiring affects the precision of the low resistance measurements.  But, since the accuracy shown on the specification sheets for the earth resistance meters is so much higher I did not try and eliminate that.

The accuracy of my home made earth ground resistance meter is also included.  It was intentionally made with higher current measurement capability (up to one amp).  That needs to be wisely used so that you are not boiling off water and affecting the measurement.  Most of the other meters have AC currents under 10 mA and have frequencies other than those of the commercial power so that leakage currents don't combine with those low measurement currents and reduce the accuracy.

All of the meters had pretty good accuracy.  Higher price meters have better logging and computer interface than those tested.  The KEW4106 meter by Kyoritsu Electrical (not tested) looks a lot like the Duoyi DY4300B (not the DY4300A tested) with a different face plate.

The Uni-T UT276A clamp meter had stable readings for everything other than when measuring resistances above 200 ohms.  Due to the low currents during those test measurements (under 0.2 mA) the reading displayed might be a bit off if the clamp is not closed exactly right.  The lower readings (which are my prime concern) don't have that limitation.

The Analog Devices CN0359 was designed for conductivity measurements of fluids and has problems with the inductance from the long wires used in earth resistance and resistivity measurements.  It should not be purchased for earth testing.  It was included due to its similar measurement method.

I'll add another meter when it becomes available and perhaps fill in the empty boxes.  See my 23 October message for two added meters and an updated  spreadsheet.

[Mechatrommer]

Thanks mate i'll read up when i got home..i have slightly better view now of why the need of 2,3,4 stakes earth measurement.

[Mechatrommer]

Earth Ground Resistance Testers.xls

this one really help on device selection, thanks alot!

[ocw]

I added on to my accuracy verification spreadsheet covering earth ground resistance and resistivity testers.  I added the ETCR2000A+ and the Simpson 444 micro-ohmmeter.  The ETCR2000A+ is pretty much identical to the UT276A with different features enabled.  The Simpson meter isn't made for earth ground measurements but measures four wire resistance via a low AC (or DC) voltage and current, and I thought that it would be useful for comparison.  The Simpson's low test voltage is helpful in that it won't arc over a corroded connection which may be destined to later have more serious arcing.

The UT276A and ETCR2000A+ were both made by the same manufacturer as shown by their outward appearance and their circuit boards shown on the attached pictures.  It seems more likely that ETCR made both.  They both have the same IC's with their numbers scratched off.  They are still fairly legible in the Uni-T meter.  Those IC's include the Exar SP3232EEN RS232 transceiver IC and the NXP HEF4066BT switch in just the UT276A since only it has the RS232 interface.  Both have TI IC's labeled 72500 14T CZ90 and 22521 08M AZN2.  I haven't been able to translate those numbers yet.

Both meters have good accuracy.  The ETCR meter can only measure up to 200 ohms while the UT276A can measure up to 1,200 ohms.  The lower limit on the ETCR meter isn't important for my use.  Plus, for measurements above 200 ohms the UT276A accuracy falls off and the precise contact of the clamping jaws becomes much more critical with high resistance readings due to the low measurement current for that range.

Only the UT276A has a computer interface.  That is just via RS232 and the current software isn't the best.  Attached are pictures of the sampling and memory screens.  It doesn't tell you much more than the meter's front display and the graphing part of it doesn't work with my computer.  The memory shown on the computer isn't too useful other than seeing all entries at once.  There isn't any time information stored or anything else to identify the specifics of the reading.

The UT521 takes 2 and 3 point readings.  While I've never had a Yokogawa EY200 meter, they both look like they are very similar and probably made by the same manufacturer.  The Uni-T meter is typically much less expensive.  The UT521 has a STC 12C5410AD microcontroller, ATMEL 24C04N 512x8 EEPROM, 3201-BI ADC, HEF4053BT and CD4052BM multiplexers, OP07C, OP97FS and LM324DG op amps, MAX660CSA regulator and more IC's.

I found that the test figures on my prior spreadsheet were a bit inaccurate for the UT521, particularly for low resistance tests, due to my not noticing extra resistance on a test cable used.  Those figures were updated on the attached spreadsheet..  The UT521 had pretty good accuracy, generally meeting its specs.  As with all of the resistance/resistivity meters, due to their limited resolution, a one or two digit error can cause a significant percentage change.  And while all spec sheets shows accuracy per cent +/- so many least significant digits, I haven't included those digit figures in my accuracy figures.  And with the readings dependent on soil moisture and temperature and well as composition, the earth ground resistance readings are not constant.

The DUOYI DY4300A is a couple of steps above the UT521.  I can also take 4 point readings, adding resistivity measurements to the earth ground resistance measurement which the UT521 can make.  It also has greater microcontroller power and includes the individual resistances between the adjacent points to its front display.  It also measures keeps track of any interference and displays the frequency and level of the interference.  Its memory includes time and location information.  The DY4300A appears to be the same thing as the Kyoritsu KEW 4106, with different features activated.

The main IC's in the DY4300A include the STM32F103ZET6 microcontroller, AMIC LP62S16256FV-70LLTF 256k x 16 CMOS SRAM, 24LC028 EEPROM, TI TL1453 pulse width modulation controller, DG409CY+ multiplexer, TI LM2575S and PS767D318 voltage regulators and probable op amps labeled 2904, LT521, LT121 and LT423.

The DY4300A has good accuracy and has the choice of four test frequencies.  Or, the frequency can automatically selected to that having the least noise level.

The accuracy of my personally built unit is also included.  It was designed to be most accurate on the lower frequencies.  The resistance/resistivity values need to be calculated based on the five digit voltage and current displays.  While in one sense that's inconvenient, knowing those values is useful when evaluating the best probe separation distance.

The accuracy of the Simpson 444 was good.  I included some additional accuracy measurements of 10, 25 and 100  milliohms.  Those extra resistors all had 1% accuracy, so the meter should have been more accurate than the resistor.  I could have used my 34401A measured resistance as the standard for all equipment rather than the resistors value.  However, due to most meters having just 2.5 to 3 digit displays that appeared to be overkill.

While still accurate, the Simpson 444 is an old design.  It uses a MAX7129CPL ADC, some 4000 series logic IC's and LT1001ACN8, LF412ACN, KA1458 and LTC1052CN8 op amps.

As said before, while the CN0359 has good accuracy, it's not made for earth ground resistance or resistivity testing.  You will have problems using it for that.  It's made for fluid conductivity testing.  Or, it can be used with completely isolated components (like my test resistors).

So, here is addition information, with two extra meters being covered.  This is already too long with too many attachments, so I'll leave it at that.