OwenDuffy.net 


Mastech MS6514 review

This article reviews the Mastech MS6514 dual thermocouple thermometer with USB interface and software.

Important features of the instrument were:

An instrument was purchased for A$66 including post from Goodluckbuy, an online seller located in Shenzhen, China.

Goodluckbuy's service in shipping the goods was prompt, they arrived in just a couple of weeks over our Christmas period. In fairness, this is outstanding service for sellers in mainland China who appear for the most part to offer goods for sale without any stock on hand (though they almost never declare that).

Initial inspection

On unpacking and inspecting, two issues were immediately apparent:

The instrument had some interesting buttons, but operation was not intuitive, an English manual was needed.

The manual though does not have any description of the serial protocol if the user wants to drive the instrument from software other than that supplied.

English manual

The seller was aware that the product was shipping to Australia, and English speaking destination.

On checking Mastech's web site for an English language manual, none was found. Mastech's website doesn't offer many manuals, so that was no surprise.

The seller was asked for a manual, and a couple of days later, Mastech was asked directly for a manual which was forthcoming. The manual revealed that the instrument was probably a bit better than I had expected from the online advertisements.

USB socket defect

Fig 1: 

Fig 1 shows the USB socket. The metal shroud is distorted, shows tool marks, has a projection into the space where a plug is inserted, and the insulator and contacts are offset to one side of the connector more than should be the case.

The seller was contacted and offered to fix the problem if it was returned to them by tracked post. The minimum cost of tracked post from Australia was A$54. This reduces the value of the damaged item to me to A$12, not a good outcome and a statement of the worth of warranty.

If these things don't work out of the box, the cost of rectifying them can be as much as the goods cost. Quality of manufactured product is critical to low risk transactions.

Software

The software worked on WinXP, but often froze on Win7. The port range available is COM1 to COM16, which is very limiting when port assignments have been made to lots of similar devices.

The software aborts if it finds the last used port is busy, crashes instead of getting on with searching for an available port that has the instrument connected.

The serial protocol is not described and description sought from Mastech was not forthcoming. It is a binary protocol, and not of obvious structure so it is a challenge to crack it... cheaper to buy another instrument that does provide protocol documentation.

The software does not properly indicate the units of measurement, except for the case of Kelvins. It logs about two measurements per second to a maximum of 50,000 entries, so less than 7 hours of data logging endurance.

The software is riddled with problems.

Options

Several options were considered:

Notwithstanding the software problems, given the high cost of return, the uncertainty of repair, and risk of loss or damage of the replacement (should it be replaced or repaired), it was decided that the unit would be repaired locally.

Repair

The instrument was dismantled for repair. The USB socket is a quite standard part, selling here for around $1.

Fig 2: 

Fig 2 shows the USB socket after removal. The damage was evident from outside the box, and once the PCB was removed from the enclosure it was evident before removal of the socket from the PCB. Removal of the socket did not do the damage seen in Fig 3.

Note that the metal surround is distorted, it does not join properly in the middle, the tab is bent inwards and prevents entry of a USB plug, and there are tool marks from a pair of pliers.

The tool marks on this connector could only have been made with the printed circuit board removed from the case, so unless the seller pulled the thing apart and did this or sold me a used item, it has come out of your factory like this.

Replacement of the socket reveals a possible reason for the damage. The replacement socket was a very tight fit in the hole in the plastic escutcheon and it took patient perseverance to assemble the escutcheon to the PCB assembly. It is likely that the assembler dealt with a difficult socket by squeezing it up with pliers!

It speaks volumes for the quality of Mastech product!

The instrument has a procedure for recalibration which requires an accurate thermometer and accurate 40mV voltage source. The procedure was followed and both probes read with 0.1°C of each other and the external temperature reference.

Experience

To some extent, buying from a supplier in China has some risks and it is part of the territory, but obviously one makes decisions based on experience.

I have purchased six Mastech instruments in the last two years, ranging in price from $40 to $400.

The MS8910 was perfect, has given faultless service and is an excellent product.

The MS2203 was perfect, has given faultless service and is an excellent product.

The MS8226T was perfect (though the software has not been used / tested), has given faultless service and is an excellent product.

The MS8050 was supplied loose in an attache case, and the front escutcheon came of in transit and was slightly damaged. It was re-glued in place. During service, one of the conductive sleeves in the safety banana sockets pushed inside the instrument. It was simply glues in place and insufficient adhesive used in manufacture. It was re-glued with CA glue and has given good service since. So, two faults.

The MS5308 LCR meter sometimes fails to initialise at power on, however on recycling the power once or twice it usually starts.

The MS6514 has a faulty USB connector and software that freezes on my Win7 machine.

So, of six instruments, two have been faultless and 3 have had faults.

Faults in 3 out of 6 instruments indicates an underlying quality problem that is not a fault of the retailer but of the manufacturer.

Fig 3: 

Fig 3 shows the quality stickers and certificate from the faulty MS6514. It shows that Mastech think that quality comes from plastering the thing with stickers and inspection certificates! How did an inspection certificate get signed off when the USB connector was so damaged?

Mastech's final word was to suggest it is a counterfeit product (ie not their problem). The Chinese would appear to have no regard for intellectual property, yet cry foul of counterfeiting when it suits. The seller assures me (as they would) that it is genuine product from the Mastech factory, they are probably right. If Mastech do have a quality management program, they should be able to track down the certificate above.

In review, my experience has been that the probability of faults in Mastech product is high, too high to stand the risk of warranty service available from an overseas supplier at high transport cost.

If retailers experience the same problem rate that I have, they surely would not sell the product. Sadly, though this is not the direct fault of Goodluckbuy, the Goodluckbuy experience is soured.

By contrast, I have purchased four Rigol instruments and three Uni-T instruments and they have been faultless, zero failures out of seven. I have purchased many non-Chinese instruments (eg Kyoritsu, Fluke) and never had a failure.

Today, almost all of this stuff is made in China. The strategy is to avoid low quality manufacturers such as Mastech has proven to be for me.

Other observations

The instrument is labeled with a warning to not exceed 30V from the inputs to ground. This might suggest that the inputs are isolated from ground, and hopefully from each other.

Note than many thermocouples have a conductive path between exposed metal (the bead itself or metal sheath) and the electrical connections.

 Experimentation suggests that the inputs are not isolated from each other or from the USB ground.

Fig 4: 

Fig 4 shows the top view of the PCB. It is quite nicely made, soldering is excellent and flux residues have been cleaned from the board. The clips for the input leads and battery terminals are non-magnetic, so quite probably stainless steel and corrosion resistant.

Fig 5: 

Fig 5 shows the input circuit area. The inputs connect more or less directly to a SGM8552 dual op amp which runs on a 3V supply. The specifications of the SGM8552 limit input common mode voltage range to -ve to +ve supply, 0 and 3V in this case. It is difficult to understand the maker's 30V rating.

Note that one terminal of each input connects to a common track, so the inputs are not isolated from each other. The thermocouples use must not make an electrical connection between them, eg if two naked beads were connected to a conductive heat sink, measurements would be disrupted. Likewise, if either thermocouple has an electrical path back to the USB ground, measurements would be disrupted.

The aperture in the escutcheon for the thermocouple plugs is a bit small. The 'standard' pin thickness seems to be 0.8mm, and common Chinese plugs for 0.95mm or greater do not fit into the socket (though they are accepted by all other instruments I own). Most plugs that are more than 0.8mm thick have deformed pins using material that is about 0.8mm thick, so if they are flattened in a press, the pin thickness may be acceptable... but you may need a 2t press!

 Overall

The instrument has some nice features, and once repaired is more useful than with the defective USB , but the failure of the software, its lack of capacity and flexibility and the lack of documentation of the serial protocol prevent full exploitation of its potential.

Like most inexpensive instruments, cold junction compensation range is limited (18-28°C) but can be disabled for more accurate differential measurements outside that range.

The lack of isolation between probes and probes and USB bus imposes significant limitations in use.

The internal data log uses a 25LC256 EEPROM which has an endurance of 1,000,000 write cycles. Depending on how the memory is used, this might result in significantly less data writes.

The instrument has been put to effective use, but not before time spent in reverse engineering the serial protocol and writing some custom data logging software.

This is yet another Mastech instrument which I have purchased with faults out of the box, quality is an issue.

Links

Changes

Version Date Description
1.01 21/01/2013 Initial.
1.02    
1.03    

© Copyright: Owen Duffy 1995, 2017. All rights reserved. Disclaimer.