Handheld meter robustness testing

[joeqsmith]

I added a few MOVs along with some resistors, similar to all the other meters I have looked at the hold up.  I hadn't thought about the fuse when I made the video and talked about cutting the trace and moving it.  This meter detects a blown or missing fuse with this circuit.   So I left it.   

Using a spare plug, then wrapped the inductive pickup's cable around the wire and ran it like this for about 15 minutes without a reset and the RPM is stable (as long as the RPM is high enough to disable the multi-spark).   The OEM inductive pickup that was left out of the box when I received it, may not do as well as this Fluke clamp.

I've been asking Brymen about their 319s.  They have confirmed that while they support both 2 and 4 stroke engines but they only support 2-stroke wasted spark, not 4 like I run. So this will cut the RPM by half.  This meter is also a little more generic with a mV and uA range.  The bar graph updates 40 times/sec rather than at the normal display rate like the All-Sun.  It also has both a + and - trigger rather than just +.   Sadly, it does not appear to be TRMS.

They watched the video where I had the two Brymen's on the ignition jig and Dave's went through the reset.   It may be interesting to see how the BM319s would handle this.  The 319s is also marked for CAT III 600V.  None of this CAT II stuff.  So if I run one, I will transient test it as well (after I run all the other tests). 

If the 121GW is released with an open source, maybe we can role an automotive version of that meter. 

[joeqsmith]

Testing the final mods to the ALLOSUN.  Nitrogen tank was almost empty but it had enough to provide some additional stress to the meter.   

https://youtu.be/Xk6mPHkxs1Y

[floobydust]

Huge difference in EMI between resistance spark plug cables and plugs compared to solid copper-core.
I know some magneto nitro dragsters run all solid-core wire and murder to do any datalogging without interference. But they have a distributor too, which also radiates a lot of EMI.

[joeqsmith]

I have continued to look at ways to run some sort of life cycle testing on a few of the meter's rotary switches.  I want to have a better way to determine how the contacts are degrading rather than just a visual inspection. 

I have considered using a camera and just using the resistance mode but the meter all have different specs and may not even read down low enough to detect a problem.    Using an external meter to measure the meter under test's current source may be an option but again, as the meter's batteries degrade and such, it may not be good enough.   Another option I considered was using an external meter to measure the resistance of the switched current inputs.   The values used the mA ranges are somewhat of a standard.  The problem again is trying to measure sub ohms of change on possibly 100s of ohm of shunt.   

The problem being obviously there is no standard why to beyond just doing the measurement directly.  What I don't like about this is the meter will have to come apart and test point added to the board and again use an external meter to make the measurement.   

I setup a test like this using one of the meters I had damaged.  In the plot shown, the vertical is the resistance and the horizontal is the number of switch cycles.  Again, this was a damaged meter so I would hope we would not see this poor of results with this few cycles.   

Obviously, the plan will be to run this destructive test on new meters at some point but until I sort out the details there is not much reason to take this step.  Similar to running the transient tests early on, it took a while to sort out how to run them.  For those of you who have been around since the beginning, you may remember I start out running damaged meters for that testing as well.   I doubt I will look at very many meters.   

[HKJ]

How bad is the result?

For a cheap meter I do not see it as that bad:
The mA range is usual located so you do not need to pass through it except when measuring A or mA. I.e. it will not see that many passes.
For the other ranges on the meter I would not expect a ohm or two in the range switch is a serious problem.

For the more expensive meters I hope the range switch will lasts considerable longer.

[joeqsmith]

How bad is the result?

For a cheap meter I do not see it as that bad:
The mA range is usual located so you do not need to pass through it except when measuring A or mA. I.e. it will not see that many passes.
For the other ranges on the meter I would not expect a ohm or two in the range switch is a serious problem.

For the more expensive meters I hope the range switch will lasts considerable longer.

The switch is commonly used to control the power to the meter.  Of the meters I purchased to run, two of them would not power up because of problems with the switch.   The switch may not just hard fail but instead start to become intermittent.  I have ran into this with some of my old meters.  As a matter of fact, when I first started working on the transient testing, one of the meters I tested was my old BK Precision which the switch was going bad on. 

I really have no idea if its a common problem with handheld meters or not. It's not really my objective to determine this.  Like the transient testing, I am really just looking at how the different meters compare against some standard way of testing them. 

As to how bad this particular switch is, it's really bad.   Your feeling that it is not so bad is partly my fault.  I have zoomed into the sub 2 ohms but the resistance is off the graph.   

The switch continues to cycle and these are up to date graphs with it zoomed out to 150 ohms and 10M.  The switch has started to open up at 3000 cycles.   In this case, I am calling one cycle switching from off to on then back to off.  The cycle rate is currently fixed  for one cycle every 2 seconds.   I am running it slower to minimize heat buildup. 

It is very possible that the meter would have had trouble powering up long before the switch started to go open.   And again, maybe this is good enough for many people.  And again, to be clear this is a damaged meter.  The data I have collected so far is really just to get an idea how to run the test, not to try and suggest a meter has a bad switch design or not. 

[HKJ]

We are talking more than 2000 cycles before it fails, for occasionally usage that would last many years. For daily use it would probably not be acceptable.
I would not see the 2000 cycles as a problem for <$30 meters, but for >$100 meters I would like considerable more cycles.

[joeqsmith]

There were a few people who commented early on during the transient testing about how I should not be showing $50 meters in the same light as $200 meters.  When I finally do start running them, they will all still be treated the same. 

I am not sure what a good target number is.  I tossed out some thoughts when I started working on this project.  In the end, I plan to run them to failure and let each viewer decide for themselves. 

Even with a 2 second cycle, there is still some small amount of heat built up in the switch. 

[tautech]

For a full time DMM user 5-10 On-Off cycles/day would not be unreasonable IMO.
Consider max and 200 days use/year and we might think for a 1 year warranty 2000 cycles is a fair amount.

Quality meters, you might expect 4000 cycles IMO but would they actually do it ? 

Interesting test Joe. 

[bitseeker]

With all the data in a spreadsheet, one can choose their budget range and see what performs to their requirements. Pre-defining price tiers is probably unnecessary unless you do summaries/round-ups to choose winners.

[joeqsmith]

Consider that some may actually use their meters in CAT III and IV environments.  If the parts are wearing, that material is going somewhere and it may not be in a place you want it.      

If the basic Fluke is around for 30 years, using your 2000 a year we are talking about 60,000 cycles.   It seems like if we really want to know how well meters like the Brymen are going to hold up long term under normal use conditions, this may be one indicator.   I think that BK I mentioned was about 13 years old when I retired it with a bad switch.     

I'm not sure what metrics I will place in the spreadsheet.  Maybe the starting resistance, cycles until contact reaches 1 ohm, cycles until contact reaches 10M.   That or I may just keep the data on the PC like I have done with the test leads so I can compare them directly as they are ran.   I am also thinking I would like to run all of them some number of cycles regardless if they open prematurely. 

Attached, our dead meter is now at over 25,000 cycles.   Notice how the contact resistance continues to loft around but it continues to stay below 2 ohms now. 

[stj]

most meters have grease on the switch-pads,
although if that gets contaminated with metal-dust it could maybe create an unexpected path.

[joeqsmith]

most meters have grease on the switch-pads,
although if that gets contaminated with metal-dust it could maybe create an unexpected path.

I think I have looked at maybe 60 different meters now which I will give you is not much of a sample.  Most of these have not appeared to have any sort of lubrication on the contacts.  Looking through pictures, I can't even say it is related to cost.  Have you seen a lot of meter reviews where they had grease in them?  Just curious why our experience with seeing it is so different. 

I could see the grease trapping contaminates as you suggest, but it may also prevent it in the first place.  When I run the meters, they really don't get enough cycles on them to see any problems.   

[Muttley Snickers]

There are plenty of other meters around which do not incorporate a rotary selector switch and in turn the necessary lubrication, I'm not sure if you have tested any of these types of meters as yet, maybe you did and I simply missed it. Additionally there are also few other testers and devices about such as CCTV testers and the like which incorporate multimeter features and functionality and from what I have seen they generally do not carry CAT ratings at all, I'm not sure how good the protection is on these types of meters or how they would hold up to scrutiny.

A few of the button type meters listed on the page below in addition to some pictures of those other testers.
https://www.eevblog.com/forum/reviews/wanted-the-ultimate-multimeter-for-electricians-and-technicians/

[joeqsmith]

There are plenty of other meters around which do not incorporate a rotary selector switch and in turn the necessary lubrication, I'm not sure if you have tested any of these types of meters as yet, maybe you did and I simply missed it. Additionally there are also few other testers and devices about such as CCTV testers and the like which incorporate multimeter features and functionality and from what I have seen they generally do not carry CAT ratings at all, I'm not sure how good the protection is on these types of meters or how they would hold up to scrutiny.

Please explain why you feel that it is necessary to add lubrication to rotary selector switches.  I am also curious why you feel the vast majority of meters I have looked at do not have lubrication.           

One of the very first meters I looked at did not have a rotary switch.  This was the only one I ever looked at like this.  I don't believe anyone has ever asked me to run another meter like this.  I suspect they are just not popular. 

At this time, I don't have any interest in looking at other equipment outside the basic CAT III 600V and up rated handheld meters. 

[Muttley Snickers]

Please explain why you feel that it is necessary to add lubrication to rotary selector switches. 

I did not at any point state that lubrication was necessary and only referred to meters which incorporate a rotary selector and in turn lubrication which was the context of the prior posts, as you well know there are discussions on which lubricant is best suited for this application and if I remember correctly Fluke themselves recommend a particular product, I don't recall ever having to apply lubricant on any of my meters other than a slight smear of silicone grease to a Fluke 83 which was not working and in a bad way to start with anyway. The question I submitted was entirely in relation to button type meters and their robustness and nothing more than that, sorry if you interpreted my post in the wrong manner.

I am also curious why you feel the vast majority of meters I have looked at do not have lubrication.

I made no such suggestion, again with all due respect Joe I think you are reading too much into my post.

[stj]

i was the one mentioning lubrication.

you obviously opened more meters than me, so i assume it was just more common in the past - although there was a very fine layer of lube on an aneng 8002 i recently stripped.
i think it was done in the past to stop the copper contacts oxidising or to prevent tracking in damp enviroments.

these days meters arent expected to have such a long life i suspect.

[SeanB]

About the only place I would put a grease of sorts in a multimeter switch is on the detent balls, and the central shaft area, the contacts would probably be best left dry, as they hopefully are a thick selective gold plate running on a selectively plated wiping contact, and this is pretty corrosion resistant in any case. A lubricant on the plastic parts to keep them from wearing each other out is fine, but all switches aside from those designed to run in transformer oil are best left dry.

[joeqsmith]

I did not at any point state that lubrication was necessary ....

There are plenty of other meters around which do not incorporate a rotary selector switch and in turn the necessary lubrication,...

i was the one mentioning lubrication.

you obviously opened more meters than me, so i assume it was just more common in the past - although there was a very fine layer of lube on an aneng 8002 i recently stripped.
i think it was done in the past to stop the copper contacts oxidising or to prevent tracking in damp enviroments.

these days meters arent expected to have such a long life i suspect.

I had a ZT102 (AN8002) that had something on the contacts but it did not appear to be a lubrication.

Most of the old test equipment I have appears to use silver plating.  That old 70's IBM meter I recently rebuilt did not appear to have anything in the contacts, only the detent area.  I don't have any other old meters to compare it with.   I have seen a few new meters where they really put it on thick like this Victor pocket meter.   

I would guess if you took a pole, people buying Fluke brand today expect them to last as long as a Fluke they bought in the 80s.  I doubt you would find many who expect a Kasuntest ZT102 to last this same amount of time with the same usage.  I really have no idea about the switch life of these meters but I have seen a few posts where people have asked about the life of the Brymen BM869s.   

About the only place I would put a grease of sorts in a multimeter switch is on the detent balls, and the central shaft area, the contacts would probably be best left dry, as they hopefully are a thick selective gold plate running on a selectively plated wiping contact, and this is pretty corrosion resistant in any case. A lubricant on the plastic parts to keep them from wearing each other out is fine, but all switches aside from those designed to run in transformer oil are best left dry.

I tend to agree but I am not a metallurgist or a lubrication / switch expert.  As Mutley mentioned, Fluke does call out what type of grease to use in some of their meters.   I have no plans to try and investigate the merits of using one type over another or dry vs lubed.  It would seem like the manufacture's would have studied this but there may not be any published papers on it.   

[kalel]

How does the grease affect conductivity? I have absolutely no idea about contact lubrication. I heard that potentiometers also have it.

[joeqsmith]

I have seen a few pots with some sort of lubrication but again, it seems rare from my experience.   I have seen it in a WW pots but I don't think I have seen it in any other types. 

If the grease used were a good conductor, you can imagine the havoc it would play in a meter.  Then again, if it prevents wear it may help in the long run.  Then again, the switch designer may need the bit of friction to keep the contacts clean...  I have no idea.   

When I do start cycle testing the switches, I plan to take some before and after pictures.  This should give people some idea on how much grease, if any was used.   

[joeqsmith]

I have completed my initial testing.  The external meter (HP34401A) is measuring the contact resistance using two wires that I soldered to the back side of the board, more than an inch away from the contacts to ensure that they were not disturbed during the setup.   This first curve, you can see the vertical is resistance and horizontal is the number of cycles.  I was planning to run the switch until it was a full open but I was getting close to 60,000 cycles and decided to end the test.   

[joeqsmith]

This next curve, I am zoomed in to 10 ohms to give you a better idea how the contact resistance is changing.

[joeqsmith]

This is looking at a histogram (or distribution) of the resistance from 0 to 5 ohms.   Note that we really would expect this to have a tail as the contacts wear.

[joeqsmith]

Overall, the contact areas look pretty good.  There are some metal particulates but the vast majority is made up of the solder mask.   I zoomed into one pad to give you an idea how the very edge of the pads is wearing.