Handheld meter robustness testing

[joeqsmith]

Just wanted to add thanks for all your work testing these meters Joe, really interesting to watch and learning a lot. Also great when looking to purchase some of these meters - especially the ESD type tests.

How much will the transient tests be wearing out the MOVs on these meters? It'd be interesting to see a test on a meter where the MOVs have worn out, and to see whether it changes the way the meter responds to transients.

No problem. Glad to share the little bit I have done.  I learn something from each meter I look at and some of them still impress me.  There was no way I would have thought that the BM235 would survive at these levels.  I know the fan boys (Fluke 87V, UNI-T, etc)  hate seeing their meters so low in the standings and will deny it, suggest the test is bad, suggest I had it in for their meter...  It's pretty funny really that people can be so attached to their meters.  I just run the tests and present the data.   

About MOVs wearing out, it's a good question and I don't think it's something I can answer.  The MOVs will be rated for some level, some number of hits at some standard.  These are normally a direct hit.  Now you put a PTC and maybe a resistor in front of it...   I have some experience with them being damaged in real life, catching fire, splitting, etc.

Let's consider the EEVBLOG rebranded BM235.   I tested five of the function selections (LoZ, V, ohms, mV and diode).  I then ran eleven different levels (1KV, 1.5, 2, 2.5, 3, 4, 5, 5.8ish, 6, 8 and 10KV).   Each function is tested with five transients at each level in both positive and negative polarity.  So 5 * 11 * 5 * 2 or 550 transients total!
But they are not at the same energy levels....    Again, the MOVs are normally taking a direct hit, with a very low impedance source.   The small MOV in the BM235 is rated
for 40J with the 8/20us.   But in the meter, we have a 1.5K PTC and a 1K resistor limiting the current into the device.  We can pretty much ignore the 2ohm source of the generator.   Even if the MOV was taking a direct hit from my generators, I test with under 20J.  Most meters were damaged with less than 10J available! 

So, where does that leave us.  What I can tell you is that I have yet to see an MOV fail as part of my testing.  What fails are the PTCs, input resistors, high speed clamps, ICs, PCB traces.

Consider the Fluke 107 that 5KY's sent me.  That meter has seen the worst of any meter I have tested.   You would really have to be stupid and hook it to a MOT or something to damage it.  15KV 100us FWHH to damage it!  And what failed, not the MOVs.  The high speed clamps and the PCB.     

[SeanB]

MOV units degrade slowly. The energy rating is not an indication of life, just how much they can take in a single shot. I have had them connected across the mains, and find that the large or small ones fail only after over 5 years of being abused, generally by going either low resistance, and thus blowing apart, or going dead short and either vaporising the one lead or turning into a skid mark on the mounting and a powder all over the inside of the equipment. Nice failure mode, generally takes out the fuse or breaker doing so.

[Muttley Snickers]

Just on the subject of MOVs, I was recently called out by a client to check and rectify their existing security system which I had installed years earlier, the premises have been hit twice by direct lightning strikes destroying everything in the past but this time it appears to be a transient or surge up the phone line, Telstra did attend to the site and confirmed that they do have a line surge device on the line at the entry point but obviously something got past it. Here is a picture below of the damaged MOV measuring 26 ohms in situ, the other two appear ok and are each around 17 megohms.

The rest of the existing control panel was in good order and operational apart from failing to communicate so I simply replaced the main PCB with a new one as I didn’t want any further issues, the site was 500 kms round trip so I stayed for a few days and did some other maintenance around the place to make it a worth while journey, we also ate rabbit for a week. 

[SeanB]

Just on the subject of MOVs, I was recently called out by a client to check and rectify their existing security system which I had installed years earlier, the premises have been hit twice by direct lightning strikes destroying everything in the past but this time it appears to be a transient or surge up the phone line, Telstra did attend to the site and confirmed that they do have a line surge device on the line at the entry point but obviously something got past it. Here is a picture below of the damaged MOV measuring 26 ohms in situ, the other two appear ok and are each around 17 megohms.

The rest of the existing control panel was in good order and operational apart from failing to communicate so I simply replaced the main PCB with a new one as I didn’t want any further issues, the site was 500 kms round trip so I stayed for a few days and did some other maintenance around the place to make it a worth while journey, we also ate rabbit for a week.

Just how does the response time SLA work there, 500km is going to eat badly into the 5 minute response time for an alarm activation.

[Muttley Snickers]

Hi Sean, the system just reports via SMS message to the clients mobile over PSTN for mains outages, alarm/ fire events and also notifies of low water tank levels in which case we can remotely pump water up to the header tanks from either the dams or the bore, on an event he can readily contact neighbours who will rapidly attend to check things out in addition to remote camera surveillance, all pretty nifty stuff but mobile phone coverage in that particular remote location is ordinary at best and the latency of his satellite internet connection makes it a bit tricky to do stuff in real time reliably.

[joeqsmith]

MOV units degrade slowly.

Not always the case. 

The energy rating is not an indication of life, just how much they can take in a single shot.

I don't think anyone made this claim.  However if I put 10J into a 1J part or a 100J part, which do you think will survive longer?  This is a question of derate and it's not something simple that could be answered.    OP's question was  "How much will the transient tests be wearing out the MOVs on these meters?" and my claim again is that I have yet to see a MOV fail during my testing.  And I would go so far as to say if I took for example the EEVBLOG rebranded Brymen BM235 and attempted to run a test using my transient generator, I could not get MOVs to fail.  The generator is just not capable and the PTC with the added resistor will really go a long way to prevent any sort of overheating in the MOV. 

If you are interested in reading a little on MOV's MTBF, Kenneth Brown published a fairly basic article several years back.   Again, how you would model it, not so simple.   
http://iaeimagazine.org/magazine/2004/03/16/metal-oxide-varistor-degradation/

[saturation]

Yes, on a historical note, it tooks decades before MOV makers put in-lines fuses with MOV [ you can still get them without it], as when they finally fail they fail catastrophically; although a 'used' one can be taken out of service by testing its pass through voltage, that's practically impossible once imbedded in devices. 

In good DMM designs the casing holds in the debris and fire, but in prior years particularly in surge strips when higher joules were all the rage, the high joule capacity MOV would burn with enough energy to eat through the strip housing; in the past, the solution of the designers was too use a metal container  .  A simple in line fuse or better yet in-line and a thermally activated one as double fail safe, saves the day.

MOV units degrade slowly. The energy rating is not an indication of life, just how much they can take in a single shot. I have had them connected across the mains, and find that the large or small ones fail only after over 5 years of being abused, generally by going either low resistance, and thus blowing apart, or going dead short and either vaporising the one lead or turning into a skid mark on the mounting and a powder all over the inside of the equipment. Nice failure mode, generally takes out the fuse or breaker doing so.

[saturation]

Muttley just on the off chance, do you know what part number of MOV that is or do you know if a fuse is inline with it?  It failed fairly gently so I'm presuming it has an in line fuse.

Just on the subject of MOVs, I was recently called out by a client to check and rectify their existing security system which I had installed years earlier, the premises have been hit twice by direct lightning strikes destroying everything in the past but this time it appears to be a transient or surge up the phone line, Telstra did attend to the site and confirmed that they do have a line surge device on the line at the entry point but obviously something got past it. Here is a picture below of the damaged MOV measuring 26 ohms in situ, the other two appear ok and are each around 17 megohms.

The rest of the existing control panel was in good order and operational apart from failing to communicate so I simply replaced the main PCB with a new one as I didn’t want any further issues, the site was 500 kms round trip so I stayed for a few days and did some other maintenance around the place to make it a worth while journey, we also ate rabbit for a week. 

[SeanB]

Well, the MOV's on Mutley's boards are on a phone line, so the typical value for these is 250VAC rating, and generally there is also a upstream dual spark arrestor, typically a 470VAC device. Limited energy on a phone line, you will find the few kilometres of fuse wire, otherwise known as the twisted pair cable, are a very effective energy limiter to reduce the power into the board.

However most of the time the board fails from having an inadequate ground connection, as typically the installers either use the mains wiring, with a 0.75mm earthing conductor, which is very inductive as it is typically quite long and only connected to a poor local grounding system. Ideally you need a 4mm cable (minimum) direct to a good ground rod array, directly into the ground and with minimum length, with the telco side spark gap connected there as well.

Attached is an old failed electric fence load unit, which has to absorb the 4J of power in each pulse from the energiser, and which typically gets a belt every second. The VDR elements there rarely fail, unless hit by lightning, and the typical failure is the high voltage transformer suffering from insulation breakdown. 6 1kV MOV units in series, and the IR led is the touch detector, measuring the voltage pulse on each cycle and communicating it as a light intensity pulse to a remote phototransistor on the main board. Cheap way to get a 20kV optoisolator.

[joeqsmith]

Most applications, like line cords, the MOVs will be directly across the source.  I doubt that would ever be allowed in a hand held meter (certified).  There was one meter I looked at that had the footprint for a MOV right across the input.     It was a not populated.   Every meter I have looked at has something to limit the current to the MOV (combination of resistors and PTCs).   Then the MOVs are always clamping above the rating of the meter.  Even if one were to short, we have a second device in play to keep the MOV from coming apart (or catching fire, etc). 

The meters on nice high impedance devices (except for the current inputs) so they can get away with the PTCs / resistors.   Tough to add a few Kohms in line with your AC outlet and have it still be useful.         As mentioned, a fuse or CB is an option.   

There was a pretty good article I read once about the marketing side of line cord surge protectors.   Marketing kept pushing for lower clamping voltages as a sales pitch.   Of course, this led to all sorts of problems.  Pretty funny really.  Again, nothing to do with hand held meters. 

[Muttley Snickers]

Muttley just on the off chance, do you know what part number of MOV that is or do you know if a fuse is inline with it?  It failed fairly gently so I'm presuming it has an in line fuse.

Greetings Saturation, there are no fuses that I am aware of in line with the phone connection on any security system that I ever worked on or installed and I have done quite a few, all three of the MOVs are 14D471K and the relay for the mode 3 is still working and tested as good, I just had a look at the earlier original iteration of this series (Bosch Solution 16) and they also used similar MOVs, this was the first time I had seen this occur on any panel.

Anyway the reason for the post was just to point out that large transients can be present on systems and equipment that most of us take for granted, the next time I hear a thunderstorm on the other side of the hill instead of fitting off equipment particularly phone related I will instead go to the fridge and sit it out.

Data sheet and picture below.

[SeanB]

Tested a cheap meter on 7500V neon sign transformer. it survived .....for a while at least.



Then I went a little Photonic on it, and cooked it to death.



Battery expired in 2011, but still has life in it ( Yay Energiser) so put it in another cheap meter to eke the last life out of it ( Ok, I am out of 9V batteries, and only going to buy more later the month if I can get past the one shop with them pretty cheap).

[joeqsmith]

Sean, that's not a neon sign transformer... This is a neon sign transformer!! 

https://youtu.be/RBkjr3b5hQo?t=379

[PedroDaGr8]

You should rename your channel on YouTube to MeterAbuse.

[rsjsouza]

Tested a cheap meter on 7500V neon sign transformer. it survived .....for a while at least.

Then I went a little Photonic on it, and cooked it to death.

Battery expired in 2011, but still has life in it ( Yay Energiser) so put it in another cheap meter to eke the last life out of it ( Ok, I am out of 9V batteries, and only going to buy more later the month if I can get past the one shop with them pretty cheap).

Sean, I don't think you were going Photonic on your last video. To me you were simply using a non-orthodox method of reflow soldering to bring it back to life. Sure, a few cooked plastics here and there, but that happened even with Dave and his attempts to restore a TV in one of his episodes. 

[joeqsmith]

Some time ago someone had offered to give me a pen type Mastech MS8211.  Another person had asked about running one.   Fairly inexpensive so I picked up a MS8211D.  It's marked CAT III 600V, has some sort of current, a funky retractable tip and a strange logic sort of LED indicator.   I have not had it apart yet but I wonder if it even uses a fuse as t looks like it can only read into the mAs.   

Hope to have some time in the next week or so to do something with it.   

[rsjsouza]

Some time ago someone had offered to give me a pen type Mastech MS8211.  Another person had asked about running one.   Fairly inexpensive so I picked up a MS8211D.  It's marked CAT III 600V, has some sort of current, a funky retractable tip and a strange logic sort of LED indicator.   I have not had it apart yet but I wonder if it even uses a fuse as t looks like it can only read into the mAs.   

Hope to have some time in the next week or so to do something with it.

Joe, that was me, although mine is safer as it does not do current measurements. I have hi res pictures of its guts if you are interested (I am away from my computer now)

Keep in mind the retractable mechanism is flimsy and the thread on the tip started to miss steps until the point of not pushing it out anymore.

Otherwise, it is an interesting form factor.

[joeqsmith]

You must be the one who offered to send the defective one.   I had a few people write me about it.  Not sure why it would be popular.  The first thing I noticed was how poor that retractable tip was constructed.  I hope to have a some time this week to run it.

If there is anything you want to see with it while it is still in working order, let me know.   

[SeanB]

Sean, that's not a neon sign transformer... This is a neon sign transformer!! 

https://youtu.be/RBkjr3b5hQo?t=379

I said it was the baby neon sign transformer, the other 14kV ones are at home, where they can make a really respectable snap crackle and pop. They also deliver 30mA, and the one electronic one is running my neon collection.

[Muttley Snickers]

Hi Joe, If you haven't already blown up the Mastech 8211 can you verify the maximum current it will cope with in regards to current draw, the manual states 200mA and I've been looking at these things for years but the two main concerns were current limitations and a lack of a backlight.   

Also Mastech's web site appears to have recently expired, shorted out, blown a fuse or something.   

[joeqsmith]

Hi Joe, If you haven't already blown up the Mastech 8211 can you verify the maximum current it will cope with in regards to current draw, the manual states 200mA and I've been looking at these things for years but the two main concerns were current limitations and a lack of a backlight.   

Also Mastech's web site appears to have recently expired, shorted out, blown a fuse or something.   

You are in luck!  I have not yet done anything with it.  You want to know what is the maximum current it can measure for both DC and AC?  The manual does show 200mA with a 100uA res.  The max display is listed at 1999, so I would assume 199.9mA.  The manual does mention the resettable fuse.  They also show a frequency range of 40-200Hz.   

[Muttley Snickers]

Thanks Joe,   

For my needs just DC is fine, I carry a couple of larger meters in the truck anyway and was only looking at these things after having a short play with one a while back, I thought that for around thirty bucks one could live permanently in the tool box, some devices we test pull around 150-180 mA but occasionally they can draw slightly more, many thanks.   

[Muttley Snickers]

Haven't seen Joe about the place in a while, hope he is ok and the Mastech didn't get the better of him.   

He is also about to hit 1000 subscribers so a big thank you and congratulations are probably in order, Thank you Joe and well done.   

Joe's Videos.
https://www.youtube.com/channel/UCsK99WXk9VhcghnAauTBsbg/videos

[joeqsmith]

Haven't seen Joe about the place in a while, hope he is ok and the Mastech didn't get the better of him.   

He is also about to hit 1000 subscribers so a big thank you and congratulations are probably in order, Thank you Joe and well done.   

Joe's Videos.
https://www.youtube.com/channel/UCsK99WXk9VhcghnAauTBsbg/videos

I have moved the spreadsheet to GoogleDocs.   The latest may be found here.   

https://docs.google.com/spreadsheets/d/1cXzYpIoyVm9QJUju4KXqM22CEQZP3_xwWvDyeVwxTy4/edit#gid=400910915

I would say I got the better of the Mastech MS8211D.   Not real impressed with the design or quality.   

[joeqsmith]

From youtube

I'm surprised you had all the features and were able to function test this unit with that wire and solder blob.

Note that the two diodes in question are shorted so the solder bridge would have no effect if it stayed in place.   Also the clipped lead is not making contact with anything.  It was held in place with the insulating tape that was used to prevent the crystal from shorting against the IC's pads.