This review is a little different from the insulation tester series, as it is an extension of the reviews on the Extech MG302 and the Di-LOG DL9307, that are essentially the same meter as the RS Pro RS-9985. Ultimately, all of these meter are a clone of the DT-9985 from CEM instruments.
There were initially three objectives for the review of this RS Pro unit.
1) Further investigation into the failure of the wireless communications on the Extech MG302
2) Further assessment of the short-circuit current that exceeds the IEC specification on the Di-LOG DL9307
3) Further assessment of the build quality of these clone units from CEM.
Unfortunately, item 1 was a no go from the start - when I got hold of the unit, I actually found that it doesn't have any down load capability, only memory storage.
As can be seen, the construction of the RS-9985 is almost the same as the other two units. Whilst the case colour is different to the Di-LOG unit, which is the blue one in the middle, the markings are the same as the Di-LOG. My personal preference is that the additional colour on the MG302 markings add better clarity for its use. If you look closely, you can see the missing 'Wireless USB' marking above the backlight button missing on the RS-9985, that is present on the other two units. That brings me to the first question on the RS-9985.
Why make a unit that can up to 2000 records to memory, without the ability to then download all those saved readings to a computer, so they can be analysed?
It makes the memory capability almost useless. I do not mind copying a few readings from memory into a spreadsheet, but that kind of quantity would be soul destroying to copy manually, and probably highly likely to suffer human error.
The DT-9985 is boxed in the same way as the Di-LOG unit and comes with the same accessories apart from the wireless receiver. Naturally it has the same specifications as the other two units, but comes in at half the price of either of the other two units. This seems like quite a price difference to me just for wireless communications.
I was hoping that may be RS Pro would put some better test leads and accessories in, instead of the PVC ones offered in the MG302 and DL9307 units, but this was not the case. Also to note is the lack of a carry case to protect the unit and keep everything together, same as the DL9307, unlike the MG302 where a hard-case is supplied with the unit.
The expectation was that the unit should perform similar to either the MG302 or the DL9307. A similar set of tests were conducted, although the RS9985 was not utilised on any motors as this was not the direct driver for reviewing this unit.
In terms of insulation test accuracy, the RS9985 came out close to the MG302 and significantly better than the DL9307. The failing of the DL9307 was due to inaccuracies across the whole range of tests, whereas the MG302 had most of the inaccuracy on the 125V range. The RS9985 was similar to the DL9307, with smaller inaccuracies across the whole range.
The RS9985 also had the same issue as the DL9307 on the 100kOhm test on the 500V range, returning a value nearly twice the required value. Whilst the MG302 was inaccurate on this particular test, it was only 11% out against the 97% of the RS9985 and 100% of the DL9307.
Plotting the full range of values for the RS9985 leads to an unreadable graph, and so another plot is made with the test values below 500kOhm excluded.
Plotting the data in this manner shows reasonably consistent behaviour of the RS9985 across the rest of the test results. The fact that this one test value at 500V seems to cause an issue just for these meters is a little puzzling. The video below shows the 100kOhm test being conducted on the RS9985 and then the Fluke 1587FC verifying the test resistor value.
Open circuit voltage regulation at 500V was the same between the MG302 and the RS9985. The DL9307 had a slightly better voltage regulation than the other two, but it was not really significant enough to be noted as better performance.
The test of interest to me was the short-circuit and 1mA load tests on the RS9985.
Output voltage across the test ranges was comparable across all of the units. The MG302 tended to have the higher open circuit voltages out of the three units. A significant difference is seen for the short-circuit current tests, with the DL9307 showing the high current that exceeds the IEC specification. The RS9985 is more comparable to the MG302 at around 13mA. Whilst this is very high for a unit of the nature, it does not exceed the IEC recommendations of 15mA maximum. At the 1mA load test there is very little correlation between any of the meters.
This would indicate to me that the DL9307 is either faulty or has been programmed / constructed in a different manner, but without input from Di-LOG or CEM, I am unlikely to identify what the actual issue is. I have been through the circuits as much as I can comparing component values and I have not found any major discrepancies between the three units.
The output waveform at 500V can be seen above with a 20MHz filter applied. As with the other two instruments, the rise time is extremely sharp and cannot be measured without using a much faster timebase.
On a faster timebase, the rise time was measured as 2ms, comparable to the MG302 and DL9307, but significantly faster than a lot of insulation testers that are usually around 300ms. An overshoot of around 40V can be seen on the full plot, likely to be due to the fast ramp rate of the output. This is the same though for all three of these units.
Power drawn from the battery pack for the various functions was comparable across all three units.
Motor Simulator Tests
Whilst I did not take the RS9985 out to any motors, I did test it against my winding simulator.
Winding resistance tests were consistent with the nominal values and the readings obtained from the other two instruments.
Polarisation Index test on the winding simulator was consistent with the MG302. The RS9985 did not display the reading spike of the DL9307.
Overall tests on the motor simulator were good and did not give rise to any issues or concerns.
The final concern was the build quality that had been found to be quite more on the DL9307, but a bit more reasonable on the MG302. Based upon Extech serial numbers, I believe that the MG302 is quite a bit older than both of the other two instruments.
This isn't a full set of pictures of the RS9985, but more of a comparison against the other two units and identification of potential build quality issues.
The rear covers removed shows the same build for all three of the units. The wireless communication module at the top go the PCB on the MG302 and DL9307 can seen to be missing on the RS9985. The beginnings of quality issues on the RS9985 become visible with evidence of rework around the HV transformer and the clamping transistor arrangement behind the rotary function switch.
There is one of the mechanical restraint clips missing from the transformer and a lot of solder residue left on the surrounding PCB and components. This re-work was also evident on the DL9307, but not on the MG302. It seems like a strange component to have issues with, they could be poor quality components to start with, or may be they are being damaged during the manufacturing process of the meter.
A globule of solder was also found stuck to the electrolytic capacitor, close to the HV transformer. It would not be good for that to come loose during operation and cause a short on some part of the PCB.
At the bottom end of the board, there looks to be a pair of transistors in a clamping circuit arrangement. Both of these transistors look like they have been replaced at some points. Again, there is evidence of solder residue around the board.
A closer look around the input circuit varistors, shows that they have been squashed and the case around the leads is cracked and starting to fall away. This appears to be just down to the lack of internal space within the meter and they are not the only components that have been squashed down.
The high voltage board is removed and remains connected to the rear case via the battery pack leads that are soldered at both ends. I was interested to see what the shunt resistor arrangement was, as this had looked to be different between the MG302 and DL9307 units.
Immediately noticeable is that there is no EMF shield over the top of the DAC chip as there is on the other two units. Presumably, this is only fitted if there is a wireless communications module installed and so the RS9985 is not in need of it. This does however, allow the main DAC to be identified as ES51966A from CyrusTek. The main micro-controller being an ATMEL ATMEGA644. Also notice the mounting of the film capacitor, with the leads very close together and liable for shorting.
To remove the lower board, the 6 securing screws, rotary switch contact block and remaining input jack connections are all removed. This is where the next major issue arose. On undoing the securing nut for the 10A jack connection, the connection strip sprung off from the board. revealing that the soldering had not really worked for this connection. Looking closely at the spring clip for the 10A fuse above the solder pad, some signs of arcing are seen to have occurred.
The shunt resistors, look to follow the MG302 version rather than the DL9307. However, in-circuit measurement of the resistor values hows that the overall resistances are comparable between all three units, so this is not likely to be the source of the higher short-circuit current of the DL9307.
The rest of the build is inline with the other two units, and whilst not comparable to the more expensive units from the likes of Chauvin Arnoux, Keysight and Megger is fairly respectable and adequate for an insulation tester in this price range.
The manual supplied with the RS9985 is more in keeping with the manual from the Di-LOG. The manual is supplied in multiple languages and there does not appear to have been much attempt to re-write any of the sections. Some of the instructions are understandable and can be worked through, some are more confusing and some are just wrong.
The instructions for erasing memory have never worked for me, they are similar to the instructions in the Di-LOG manual, and they do not work on that meter either. The correct way to erase memory seems to be to hold down the 'Range' button whilst switching on the meter as per the video below. Does make me wonder how a manufacturer can get the instructions wrong for their own meter.
This has been intriguing to compare these three units together and provided more insight into rebadged units from CEM. Unfortunately it hasn't provided much help in resolving the issues I have with the unit.
In some ways it is quite a shame, as I find the meter very useful as a general electricians tool. Combining the multimeter and insulation tester functions onto the same box, reduces the number of tools being carried around. It has a useful 4-20mA function built in alongside the current ranges. Giving the insulation function a boost up to 40GOhm reading capability would greatly enhance this unit and coupled with the variable recording function can carry out PI and DAR tests, just leaving the user to calculate the ratios manually.
Specifically for the RS9985, I do not understand the reasoning behind not having the download function. Sure the meter is cheaper this way, but the memory function is severely limited without data download capability.
The short circuit unit current from the RS-9985 matches that from the MG302 unit. This would indicate that either there is a fault on the Di-LOG unit causing 23mA short circuit current, or Di-LOG have a different built or program in their unit. Checking what components I could, did not reveal any disparages between the three units. As Di-LOG don't seem to want to comment, it is probably something I will never know. 13mA for a short circuit current is still too high though, but it would seem that Extech do not know how to calibrate this element of the meter.
The build quality is not up to what I would expect for a unit from RS Pro. I am surprised that of the three units, two appear to have had replacement transformers, of which refitting the mechanical securing clips seem to be an issue for the rework station. Failing to clean the board after rework is poor quality standards, and leaving globules of solder to bounce around inside leaves room for a failure at a later date. Poor solder joints on the 10A current circuit could lead to heat damage, as the unit is aimed at industrial electricians, measuring 10A is likely to occur on a repeated basis.
A lot of work also needs to be done to the manual to correct the errors and make the sections more readable. It is evident that neither Di-LOG or RS have decided to do this, but Extech have put work into their version of the manual and made it significantly better.