Introduction
Following in from the Sonel MIC-30, which turned out to be a very respectable and capable insulation tester, I have moved over to a manufacturer well known to me in Gossen Metrawatt. Whilst they offer numerous insulation testers, I was particularly attracted to the MetraHit Coil unit to add to my list of reviews. Previous testers reviewed are;
Keysight Insulation Multimeter Failure
Chauvin Arnoux CA6526 Insulation Tester Review
Flir IM75 Insulation Tester Review
Extech MG302 Insulation Tester Review
Di-LOG DL9307 Insulation Tester Review
RS Pro RS-9985 Insulation Tester Review
RS Pro IIT1500 Insulation Tester Review
Uni-T UT505A Insulation Tester Review
Sonel MIC-30 Insulation Tester Review
The MetraHit Coil is a relatively new instrument from Gossen Metrawatt and appears to have a twin in the form of the MetraHit ISO multimeter. The MetraHit ISO is a straight forward Insulation Multimeter but the MetraHit Coil has an additional Interturn Short Circuit measurement function, which I have not seen before on an instrument of this nature. The range also consists of the MetraHit 27I that offers a 4 terminal milli-ohms function as well as insulation testing.
As with the Sonel insulation tester previous to this, the MetraHit Coil was not easy to get hold of in the UK, and eventually it was purchased direct from Germany. Besides the MetraHit range, Gossen Metrawatt also offer three insulation testers in the form of the MetrISO range. These are of a different styling to the MetraHit range and aimed more at insulation testing rather than additional multimeter functions.
This blog has become crazily big as I have tried to get to grips with the operation of this meter. I have also included a number of extracts from the manufacturer's data sheet and meter manual. As these are unrestricted access on the web, I have attached them to this blog, for anyone that wants a little more information.
Unboxing and Overview
The MetraHit Coil arrives in a fairly plain white box and seemed to lack the some of the accessories included with a lot of other insulation testers. The meter itself is included along with a rubber holster to protect it, red and black test leads with built in probes and the coil adapter, that forms the hardware supplied.
Paperwork wise, there is a CD ROM with the full manual, a quick start guide and a calibration certificate. The calibration certificate shows test results and I believe is equivalent to a UKAS / NIST traceable calibration.
Immediately noticeable is that a case is not included and neither are any crocodile clips. Looking a bit further, it is then realised that a thermocouple, and both the IR adapter and the software are also extras. These items can of course be purchased to supplement your brand new meter. These items would set you back another £414 before tax in the UK, on top of a meter that already costs £740 not including VAT.
This would put the MetraHit at the top of the price chart, a full £357 more than the Keysight U1461A before tax. Bearing in mind that you get a hard carry case with the Keysight, a couple of thermocouples and adapters and wireless connection module.
The quick start guide, seen below, is multilingual and is quite comprehensive and easy to follow.
As already stated, the calibration certificate comes with measurement data and is in German and English. The calibration date was March 2018, with a recommended recalibration date of November 2019. The meter was purchased in May 2019, so has sat on the shelf for its first year of calibration and another fee is due soon to keep it in calibration, adding more expense.
In comparison to my two main bench marks, the MetraHit unit is of a slightly smaller size. Ergonomics are reasonable, the multifunction dial has a similar level of functionality to the Keysight unit with 5 push buttons offering the rest of the control. Whilst the insulation and coil test functions are from the same location on the rotary switch, they are activated by different push buttons on the front of the meter.
The display on the MetraHit is a little smaller than the Megger unit display, so the auxiliary elements of the display were found to be very small and sometimes hard to read in comparison. The backlight is blue in colour and enhances the display, but for me, the Megger unit still has the edge over this meter in terms of the displays.
The tilt stand is part of the rubber boot and its hinge is part of the rubber, this is completely different to the majority of other units, that will have either a steel or plastic hinge pin. I am not sure how robust this hinge on the MetraHit would prove to be over a longer period of time. The holster also offers a metal hanging loop, but no other attachments come with this to make use of it. Probe holders are also moulded into the holster, to allow for probe storage or holding a probe during measurements.
Lifting up the tilt stand gives access to the fuse compartment lid, which can then be removed to gain access to the fuse. The battery compartment was found further up and the holster has to be fully removed to gain access to it. My preference would be to have this the other way around, as I would hope to be replacing batteries more than fuses.
The leads themselves have moulded probes. There are GS38 style caps over the probe tips that are removable to give access to a 4 mm connector that would accept any compatible accessory. This isn't my favoured approach, as I find the combined weight and length of probe and accessory makes them harder to use and more likely to fall off during a test.
I utilised a crocodile clip from another meter with the MetraHit test lead, and you can see the length of the combination of the probe and crocodile clip below.
The GS38 caps were found to restrict access to the smaller terminals. SAK6 and above, along with standard MCBs all accepted the probes with the caps fitted. With the caps removed, access to the smaller terminals was achieved.
The probe leads seem like they are silicone insulated and do feel like good quality. However, I had issues when testing them at 1000 V. Held together by the clip supplied with them the insulation resistance was around 176 GOhm.
This dropped to 14 GOhm when hand pressure was applied to the leads. This is the worst results I have had with silicone insulated leads. They would be adequate for the MetraHit instrument, as it only reads up to 3.1 GOhms, but would create issues if used with instruments with higher reading capabilities over DAR and PI test periods. The leads are also new and will only get worse as they get older and dirtier. This certainly wasn't what I was expecting from a top tier manufacturer.
Interestingly, the operators manual contains a little note regarding insulation testing and keeping the test leads apart.
Whilst on the subject of the probes, Gossen Metrawatt instruments do have a rather unique feature to protect against incorrect jack selection on the meter. A red door is moved around by the rotary switch to block of the voltage input jack when a current measurement function is selected, and block off the current input jack when any other function is selected. For me, other than not having a current measurement function, this is the best possible solution to preventing incidents when measuring voltages on power circuits.
The other thing to note here is that the input rating of the MetraHit is CAT II / CAT III. Given that they have aimed this at relatively high powered motors in heavy industry, a CAT III / CAT IV rating would be preferred. The majority of the competitors instruments have a CAT III / CAT IV rating, so the MetraHit is a little behind in this aspect.
Comparing the instrument with the others reviewed puts it in as a mid-range instrument with a very high price tag. Whilst it has its unique coil test function and offers more multimeter functionality than a standard insulation tester, it lacks quite important functionality when carrying out insulation tests, not having a variable or ramp voltage function and no DAR or PI calculation capability. There is no remote probe option with this meter, although personally, I would not find that to be much of an issue for motor testing.
Looking at the technical aspects, it has a somewhat limited insulation test range, reaching 3.1 GOhm for the two higher test voltages and 310 MOhm for the lower test voltages when reading the data sheet, although this proved not to be the case when tested. Both the Keysight and Megger units along with the Chauvin Arnoux tester offer up to 200 GOhms on a 1000 Volt test.
Voltage and current functions on the MetraHit are comparable to the other insulation multimeters and better than the Fluke. The complete current range on the MetraHit is from one input jack and not split across two, as it is with the Keysight. The MetraHit does have a good capacitance range and comparable temperature range with a thermocouple.
The resistance range goes up to 30 MOhms, which is a little lower than the majority of the other units. It also lacks a bond test function, which is surprising for an instrument aimed at testing motors.
Performance Tests and Comparison
The MetraHit is yet again, another instrument with reduced insulation resistance ranges, so the full range of accuracy tests could not be completed. There was an anomaly with the technical data from the data sheet and the actual test results. The data sheet gave me the impression that only the 500 V and 1000 V ranges measure above the 300 MOhm, as seen below, but in reality the 100 V and 250 V ranges also measured up to 3 GOhms, and only the 50V range was restricted to 300 MOhms.
The deviation plot for the accuracy shows a consistent trend across the voltage ranges, with the most inaccuracy occurring above 300 MOhms.
Compared to the other instruments, the MetraHit lacks accuracy against the Keysight and Megger units over their full range. When compared to a reduced number of tests, the MetraHit performs a little better, seemingly more accurate than some of the cheaper units, but not as accurate as the more renowned brands such as Fluke.
In terms of voltage regulation, the MetraHit Coil is the closest competitor yet to the Keysight and Megger insulation testers. It was noted that the output voltage on the metraHit dropped off quicker when the load on the output was increased, dropping to around 350 V in comparison to an output of 450 V for the worst of the other units.
A mixed bag of results was seen for the open circuit voltage measurements. On the lower ranges the MetraHit had higher outputs, that were seen to drop off for the higher test voltages. More importantly, at 1000 V test voltage, the output measured only reached 991 V.
Looking at the entry in the MetraHit specifications, the voltage outputs have maximum values but no minimum values, so this does not really help. The IEC 61557-2 standard specifies a 0 to +25% tolerance for the output voltage, so the MetraHit does not appear to be compliant with this aspect.
Reading the specifications introduces two other interesting values. The Nominal current for the 1000 V range is specified as 0.5 mA. This is also non-compliant with IEC 61557, that requires a minimum of 1mA for the nominal current. I am also intrigued by the 'Capacity Time' column, which indicates that the insulation tests are limited to 10s duration.
Measurement of the 1mA load tests and short circuit currents confirms this aspect of the table in the data sheet.
The MetraHit Coil has the largest spread of current output seen, due solely to the 0.052 load and short circuit currents that are not compliant to the IEC standard.
The initial trace of the output curve on the 500V range showed a very fast rise and fall of time, the fast 25ms rise time leading to a little overshoot on the output voltage. This is a trait I often associate with the cheaper insulation testers, that have a poor voltage regulation in comparison to the more expensive units that have a slower more controlled voltage rise in the region of 200 to 300 ms with little or no overshoot.
Applying a 10Hz filter to the signal allowed a fall time to be captured, coming out as 17.5ms. This parameter is not as important as the rise time and overshoot as the type of electrical items tested are not susceptible to fast voltage drops.
Changing the test voltage on the instrument for an insulation test was problematic for me. There are two methods for changing the test voltage described in the manual.
The first method requires navigating through the setup menus. This is a time consuming method and still only allows the selection of the five standard test voltages. I would only advocate this kind of procedure, if it gives an enhanced function to select a variable test voltage. The second method is to use the central control key. This is obviously a short cut and much faster, the problem is that this key is also used to enable the data function, and I found it very easy to change the test voltage inadvertently.
The next plot shows the discharge resistance of the instrument after an insulation test. This was found to be at the lower end of the range of instruments tested, indicating a relatively fast discharge of any capacitive elements under test.
The battery pack for the MetraHit Coil consists of two AA batteries, which is rather small in comparison to the majority of the competition. This is likely to lead to higher current drain on the cells for each of the functions. Also of interest was the soft start of the instrument, similar to that of the Sonel MIC-30 previously reviewed, that had a high inrush current.
The highest draw on the battery was for the Coil test, note though that this is a pulsed test and the value shown is the maximum seen. The next highest loading was from an insulation test supplying 1mA load current.
The inrush when switching on the instrument showed a short pulse when captured with an oscilloscope. Measured with a power analyser, the current peaked at 679mA, considerably lower then the Sonel instrument. It is also comparable to the inrush measured on the Megger unit, that has a more traditional off switch, totally disconnecting power from the instrument.
When switched off, the current draw on the battery pack was 422uA. This is considerably more than the Sonel, and may impact on battery life, especially considering the small nature of the battery pack in comparison to other instruments.
The MetraHit has a permanent 3 bar symbol to indicate the battery level. It quickly drops down to two bars with a 0.13V drop in battery voltage. Cutout of the meter occurs at just over 2V but at 0 Bars, (2.26 V) the backlight does not stay on, it will light up but then goes out immediately.
In terms of battery usage, the MetraHit seems to fall into the mid-range in comparison to the rest of the instruments.
Winding Simulator Tests
Winding resistance and polarisation index tests were carried out on the winding simulator. As the simulator is more capacitive in nature than inductive, the coil tests had to be left until I was away on a site with a motor store available.
Winding resistance measurements are to two decimal places. The test current for the resistance range was measured as 276uA, and despite the lack of a 200mA test current, the results obtained were very good in comparison to the nominal values expected.
The MetraHit does have a zero / relative function to compensate for the lead resistance which was utilised during the tests. Whilst the meter remains switched on, the zero value is remembered even after changing functions and switching back to the resistance function. After the meter has been switched off, the zero function is lost and has to be set again.
On to the insulation resistance tests and a major issue immediately arose. The MetraHit does not have any lock function to carry out an insulation test with. The button must be held down to keep the insulation test ongoing. This is no good for DAR and PI tests on motors. However, in the good tradition of engineering, there are always alternatives.
I was curious if the lack of a lock / timing function for the insulation test was due to the 'Capacity Time' of 10 seconds in the data sheet for the meter. The MetraHit however, did not show any issue with being coerced into a 10 minute insulation test and produced a reasonable test curve, comparable to the other insulation testers reviewed.
The resistance plot diverges from the other two towards the end of the test, but the PI ratio of 7.88 recorded is inline with the 7.89 ratio recorded on the Megger insulation tester.
All the data had to be recorded manually using paper and pen. Whilst the MetraHit has a data recording function, this is limited to just one value on the insulation test ranges. The voltage is measured with an external meter, as the MetraHit only shows the test voltage range and not the actual voltage applied. There is also no timer function on the instrument, so an external clock is used to identify the time stamps.
All of this manual interaction seems woefully inadequate for an instrument that is sold as a 'Motor Testing Multimeter' and is also well below the functionality of any other insulation tester I have reviewed, or used to date. Even the old analogue insulation tester, I have used have a push button that can be twisted to lock the insulation test on.
More reading of the manual and experimenting with the buttons, and I believe that the 10s capacity relates to a semi-automatic insulation test function built into the instrument. By pressing the data key in insulation test mode, the insulation test can be started and then runs until the reading is stable or for 10s, if stability is not reached. The screen then does hold the insulation test result.
The video goes through this test and the issues I had with it.
Motor Tests
More motor tests were carried out with the MetraHit, mainly to investigate the functionality of the Coil Test, rather than the insulation test, as it was already apparent that this element was substandard.
The coil test is another test carried out with the motor isolated from its energy source. It utilises a small extra test box connected between the MetraHit and the motor under test. As the box comes with its own test leads that plug into the MetraHit, the test leads from the MetraHit are used from the test box to the motor.
Another slightly bemused scenario occurs. The operators manual for the MetraHit includes the warning below, that the tests should be carried out using crocodile clips and not to hold the test probes onto the motor - shame they didn't supply some crocodile clips with the package then.
I have a variety of motors to use the coil test facility on. The manual indicates that the standard 'Coil Test' adapter supplied is suitable for use on motors with an inductance between 10 uH and 50 mH at 100 Hz. I do not find this much use, as the inductance of a motor winding is very rarely known. It could be measured of course, but that would mean another meter. A look at the data sheet provided, (extract below) provides a bit more information stating that the 10 uH to 50 mH corresponds to motors in the region of 15 kVA to 80 MVA rating. This is obviously, much easier for engineers to relate to. It would be beneficial to include this information in the subsection of the manual detailing the coil test function.
The coil test function is easy enough to use. Unlike an insulation test where the connections are on the winding and the motor case, the coil test requires both connections onto the windings. There is no stipulation that the winding must be in star or delta, so I tested each motor in both configurations.
The test is specifically designed for three phase windings and provides a comparison of the three phases tested. Each phase test is marked by a progression bar on the left hand side of the screen. The comparison is seen in the analogue display at the top of the screen, there is no digital confirmation of this. The actual reading is shown on the main screen, but as with the insulation testing it is lost as soon as the test button is released.
The manual gives a basic interpretation guide that a difference greater then 10% indicates the presence of shorted turns in the stator winding. The motor tested above shows a difference of circa 17%, which is annoying, because as far as I am aware, the motor is new and never been in service.
Overall, I tested 7 motors in the store, I did test another 15kW motor but unfortunately, the links provided were too large for the terminal posts and I could not configure the windings into star or delta. I also went back at a later date to measure the inductance of the windings with my Voltcraft LCR-300 instrument. A 1 minute insulation test was carried out using the Chauvin Arnoux insulation tester, as it could measure higher values of resistance.
A summary of the results can be seen in the table below.
As you can see, I struggled with making the coil tests on some of the motors, predominantly this seemed to be more when the windings were in a 'star' configuration. This increases the inductance of the winding. However, only the 22 kW motor measured over the 50 mH inductance limit specified by Gossen Metrawatt, and only for the 'star' configuration. Yet, I could not get any coil test reading for this motor either in 'star', delta, or split winding configuration.
I could only get a coil test reading for the 15 kW motor whilst it was connected in 'delta'. The reading obtained was close to the 250 us limit of the MetraHit instrument, and the inductance was around 10 mH. In contrast, the 37 kW and 90 kW motors had around 8 mH and 7 mH inductances in 'star' configuration, but I could not obtain a coil test reading, despite these inductance values being below 10 mH, when a coil reading could be obtained.
I could only get a coil test reading for all 3 winding configurations for the 30 kW motor. When in 'star' configuration, it is possible to measure individual coils by placing one connection onto the star point and then moving the other connection to each winding in turn. Using this methodology, a coil test reading could be obtained for the 18.5 kW, 37 kW, 55 kW and 90 kW motors, leaving only the 22 kW motor unable to be tested.
However, other inconsistencies are also observed within individual motors, particularly the 3 larger sized motors. The table below, shows the differential for the winding tests for the delta configuration and individual windings.
The most concern is over the 55 kW motor, that shows the presence of an inter-turn fault in 'delta' configuration, that is not present when each individual winding is tested. The 90 kW motor also shows a drop in the differential going from 'delta' to individual windings. The 37 kW motor shows the opposite. The four smaller motors showed much more stability across their coil test readings.
The video goes through the tests on the 90kW motor for each winding configuration in turn.
As far as I am aware, all of these motors tested are of the random wound variety and I wondered if this could be the source of the inconsistent results. I do have a spare 3.3 kV motor available, so decided to carry out some tests on this with the MetraHit.
| {gallery} 1090kW Motor Tests |
|---|
Inductance measurement on motor |
Winding inductance value measured |
Winding resistance value measured |
Coil test in progress on motor |
Coil test finished |
This motor can only be connected in 'delta' configuration due to the spacing of the motor terminals and the links. The inductance measurement indicated a values in the region of 14mH, so based on the readings from the low voltage motors, I wasn't expecting to see a result for the coil test.
First, I tested the motor winding resistance. For a motor this size, this test would usually be carried out with a micro-ohmmeter, but the results obtained were balanced at 0.19 Ohms across all three phases.
The victory was short lived though, as the coil test merely showed over-range on all three phases.
This is quite a disappointing result and I am unsure as to why the instrument is failing to provide readings on so many motors. To investigate the issues a little more, I set up tests using an oscilloscope and Rogowski coil style clamp meters to record the output voltages and currents during the coil test.
The first picture shows the open circuit output voltage of the coil test function measured through a 500x differential probe. You can see the output gets up over 1 kV with a rapid rise, that the scope could not measure and then drops away over 100 us, but never appears to go back down to zero.
I set up a number of tests using the 40 and 100 turn coils I had wound for the current amplifier to provide a load. These have been measured as having 19 uH and 135 uH inductance when tested. They read 11.12 us and 27.37 us respectively when tested with the MetraHit coil test. Interestingly, with the two coils wired in series, the MetraHit read 29.5 us, which is less that the sum of the individual tests. With the coil test having a load, the currents were measured with a Flux Flex clamp wired to the scope and the Flir CM55 set to inrush mode.
This did not work to well, so to build up the inductance, I utilised the spare coils of wire I had from winding the coils and linked them together. The table below shows the measurements and the coil test results. You can see as the number of coils in the circuit it increased, the coil test value also increases. I still did not have enough wire to reach the limit of the coil test function.
I captured the voltage and current for some of the tests. Below are the plots for 1 coil and then 4 coils under test.
| {gallery} Voltage and Current Plots |
|---|
Coil Test Pulses with Coil 1 under test on own |
Single Test Pulse for Coil 1 under test on own |
Coil Test Pulses with Coils 1, 2, 3 and Part Coil under test |
Single Test Pulse with Coils 1, 2, 3 and Part Coil under test |
It can be seen that as the inductance is increased, the current supplied during the test drops, the individual pulses are also seen to expand, but the length of the pulses did not correlate with the coil test value produced on the MetraHit. A video of some of the coil tests carried out.
I did try to measure both the meter and coil test adapter output voltages at the same time, but I only have one high voltage differential probe. The output channel was measured with a high voltage divider probe, but this is configured to work with the 10 MOhm input impedance of a multimeter and not the input of a scope channel, so I am not sure of the accuracy of the result. I therefore repeated the test with just a single channel monitoring the meter output to the coil test adapter.
| {gallery} Meter and coil adapter voltage outputs |
|---|
Meter output (channel 1) and coil adapter output (channel 2) |
Meter output during single coil test |
Single output pulse from meter during single coil test |
Meter output during two coil test |
Single output pulse from meter during two coil test |
Armed with this knowledge, I returned to the motor store to capture the pulses for one of the motors that I could not obtain readings for. The motor of interest to me was the 18.5kW motor that I could measure in 'delta' configuration but not in 'star'.
| {gallery} Voltage and current plots during motor test |
|---|
Motor testing |
Pulse output during delta tests |
Pulse output during delta tests |
Pulse output during star tests |
Pulse output during star tests |
Not much difference was seen in the output between the tests in 'delta' and 'star' configuration. A current was still observed to flow, although slightly lower in 'star' despite the voltage amplitude being seen to be higher during the 'star' tests. The pulse length was certainly longer in 'star' configuration compared to 'delta', presumably, this is longer than the meter can detect. Again. the measured length of the pulses, did not correlate with the readings obtained on the meter of the 'delta' configuration.
DMM CheckPlus Tests
With the MetraHit Coil having multimeter functions, I decided to carry out a quick check of them against a DMM CheckPlus calibration source. The DMM CheckPlus offers single values of AC and DC Voltage and Current test points, along with four resistance values. The AC frequency can be switched between 50 Hz and 1 kHz. Below the MetraHit can be seen measuring the 1 kOhm resistor.
The results obtained were all within the manufacturers limits.
The CheckPlus does not have a capacitance test, so to test this function a 300 uF test capacitor I built up was used to test the MetraHit to its maximum capacitance range. The reading obtained showed the highest deviation of any of the multimeter functions tested, but it was still within specification and measurement of high value capacitance usually comes with high tolerances for multimeters.
Whilst carrying out the current checks, the 10A fuse was removed to simulate the effect of a blown fuse. The meter detects this and prompts the user to check the fuse, thesis another excellent safety feature, particularly useful when a meter is used to measure current transformer secondary currents.
Software
The MetraHit does have wireless communications to both download saved data and live readings from the instrument into Gossen Metrawatt's bespoke software, Metrawin 10. This would also require the purchase of the communications module that is an option and not included in the MetraHit Coil package.
The manual is unclear as to how the software operates with the insulation testing function, so it seems it would have to be purchased to find out and I am not sure that I want to go down that route with this instrument.
The software is initially available as a trial product, which will lessen the financial burden. There is also an alternative to the IR adapter on Youtube and EEVBlog that would reduce the expense further for some initial trials.
Youtube DIY MetraHit USB Adapter.
EEVBlog MetraHit USB adapter thread.
There also appears to be a LabView driver available for the MetraHit range, but currently the MetraHit Coil is not listed as a compatible instrument, but this may just be because of its relative youth in comparison to the other MetraHit instruments in the range.
At this moment in time, I have not progressed any of these options any further, I have just included a brief section to let readers know of its existence and alternatives.
Build Quality
Eventually, I decided to take a look inside the MetraHit. The external rubber boot, is actually very easy to remove, although it still fits snugly around the meter. It is much better than some off the other meters I have had to have a good wrestle with to free them from their protective boots. On the back of the meter, there are separate compartments for the double AA battery pack and 10A fuse, each compartment has a single securing screw, that is retained in the lid, and clips on the opposite side to secure them. A neoprene type seal protects against dirt and moisture ingress from the lids.
Four torx head screws hold the two halves of the meter together. Once removed the meter halves are still held together by two internal securing clips at the top of the meter.
With the rear cover removed, it can be seen that both the positive and negative battery terminal contacts for each battery are soldered directly onto the PCB and extend into the battery compartment via slots cut into the meter case. The two words from the power supply jack in the rear case are soldered at both ends, so the back cover stays connected at this point.
The input protection of the meter is rated as CAT III and is formed by two PTCs, two varistors and a gas discharge tube. For the current side, the 10A fuse can be seen in its PCB mounted holders.
The high voltage aspect for the insulation and coil test, appears to be mounted on a separate PCB that unplugs from the main PCB. This PCB has its own micro controller in the form of a MC9S08GT16 8-bit controller from NXP as well as the high voltage transformer.
The HV PCB sits over another micro controller and the rotary function switch, amongst the passive components. The micro controller is the FS9704B from Fortune Semiconductors and provides the analogue front end for the multimeter functions.
Further up towards the top of the PCB is a third micro controller, this time the M430F149 from Texas Instruments. Also seen are the diodes for the IR communication port next to the continuity buzzer. A ribbon cable connects this main PCB to a third PCB below that contains the control buttons and display.
To get to this, the three input jacks have to be unsoldered from the PCB, leaving the jacks pressed into the front half of the case.
Flipping over the main PCB, reveals an EMI shield, that appears to have predominantly passive components beneath it. It also reveals a second part to the multifunction switch, with more contact pads and what appears to be a two pole moving contact.
The detent system for the rotary switch is made up from a plastic housing and two spring loaded ball bearings. The mechanism for the input jack blocking, is sealed into the plastic moulding. I believe the solder residue was from me desoldering the inout jacks. Up to this point, I had not found any other debris inside the meter.
The mechanical parts of the rotary switch can be split apart. A red plate can then be seen on the front cover side (arrowed), that is actuated by pins on the rotary actuator. to flip the cover over the relevant input jack.
The display PCB is held in by four small screws immediately surrounding the push buttons. The push buttons are sealed units.
Overall, I found this to be a good high quality build. The soldering all looked to be good, with no signs of residue of flux. There were no signs of any rework having been carried out and no post design or 'calibration' modifications.
Conclusions
This is a bit of a Jekyll and Hyde instrument for me. On one hand there is good quality and innovation, but this is mixed in with some intermittent operation and some glaring short-comings.
The DMM functions are all good, they seem to fit the needs of most electricians and are accurate against the DMM CheckPlus. The input jack blocking safety feature is brilliant and my preferred method against warning bleeps or messages. The blown fuse warning is also another excellent safety feature, that is missing from a lot of instruments I have looked at.
The Coil Test function is intriguing and something I have not seen before. In principle, it appears to work, but I think it needs more refinement. I certainly struggled to get it to work across the range of motors suggested by Gossen Metrawatt. I think the 80 MVA rating is a typo, unless they are implying that an alternative Coil Adapter is required. The Coil Test was intermittent between different winding configurations, that I was not expecting and leads to uncertainty about its reliability. There was no recording of the individual readings, just the end comparison, that, for me, also detracts from its usefulness.
Then we come to the insulation testing function. Oh dear. For me, this is one of the worst implementations of an insulation tester I have ever come across. It seems as though the inclusion of the Coil Test function has been done at the expense of the insulation testing function. However, this is marketed as a 'Motor Testing Multimeter'. Surely a basic test of any motor is an insulation test, probably even more important than testing of inter-turn faults.
There is no DAR or PI ratio and not even a way to lock on the insulation test, so that a standard 1 minute test can be achieved. The actual output voltage is not displayed, there is no option to display the leakage current. The result can only be retained on the screen when using the semi-automated test function.
Looking through the literature and meter markings, it is not obvious that the meter has been built to the IEC 61557-2 standard, the 1000V test function did not test as compliant with this standard. This, I just do not understand, especially when coming from such a renowned manufacturer as Gossen Metrawatt. Testing a customer's apparatus needs assurances and compliance with IEC and other standards gives my customers the assurance that the test apparatus is designed to function the right way.
On the good side, the insulation test does range up to 3.1 GOhms for test voltages of 100 V and above. For my specific tests, I would want a higher range, in the region of 40 GOhms, but for general motor testing the 3.1 GOhms, should be fine.
There is no 200mA continuity function. Whilst the resistance function worked very well and was very accurate, IEC requirements for testing earth bonds requires at least a 200mA test current. That means I need another meter with me to carry out this function.
Since the meter is aimed at motor testing, and according to Gossen Metrawatt, has the capability to test large motors, these types of motors are likely to be in installations requiring CAT IV protection, so the CAT III level is also a disappointing feature.
The lack of crocodile clips and a case just adds to the misery really. This is a premium manufacturer with a very expensive offering. I appreciate the design that has gone into the Coil Test facility and I am all for innovation, but not at the expense of other aspects of the test apparatus.
Overall, the MetraHit Coil is a disappointment for me and I am struggling to justify the high price just for an inter-turn test facility. The design of the insulation tester means that it just isn't comparable to either the Keysight or Megger units and it just isn't worth me taking it up to the generator rotor to see how it would perform.
I am not sure what to do with this next. The Coil Test facility does intrigue me and the documentation suggests that other adapters for different sized motors are available to order, so I might try and engage direct with Gossen Metrawatt. I do have a section of redundant plant at one site, so I may look to see what motors are available within that, with the intention to strip it down and deliberately place inter-turn faults in the stator winding to see how the MetraHit Coil performs.
I have thought of giving Gossen Metrawatt another go to see if they can redeem themselves, but looking more closely through their insulation tester range, I can only find the METRISO Xtra that has DAR and PI functionality, but that is an awfully expensive instrument and does not seem to be readily available in the UK.
Whilst I have a think about how to progress this and possibly engage with Gossen Metrawatt, I will move on to another tester that is waiting in the wings ready to make an entrance.
 | metrahit-iso-coil-datasheet.pdf |
 | metrahit-iso-coil-manual.pdf |
Top Comments