Handheld meter robustness testing

[joeqsmith]

My response to fungus's questions about the free Harbor Freight meters. 

Those free meters have gone through several hardware changes over the years to reduce cost.  Even if we saw a difference between the two meters I show, I couldn't tell you if it had anything to do with the label or not. 

Is 1100V the limit? Seems to me like a 2000 count meter should be able to display 2000V... 

Original posts may be found here:
https://www.eevblog.com/forum/testgear/trashy-meters-redux/msg4817894/#msg4817894

[joeqsmith]

Maybe it's worth the time to look at.  At least they don't claim the counter can read 200MHz this time!

[NoMoreMagicSmoke]

This time they are just claiming that Bluetooth will work through a closed ebox and a heavy steel door. Lol.

From their video it looked neat, then I looked at the specs. 0.3% 60000 count meter for $140. WTF

[Grandchuck]

About $110: https://www.aliexpress.us/item/3256805466430596.html?spm=a2g0o.productlist.main.5.48385619g9zuNa&algo_pvid=3b398086-d824-47a4-8d7d-ec0838872b5f&aem_p4p_detail=2023062605551514660048629148000000567842&algo_exp_id=3b398086-d824-47a4-8d7d-ec0838872b5f-2&pdp_npi=3%40dis%21USD%21213.32%21108.79%21%21%21%21%21%40211bea7b16877841157241682d07d7%2112000033895704502%21sea%21US%214108842368&curPageLogUid=ehSqclrOkexN&search_p4p_id=2023062605551514660048629148000000567842_3

[joeqsmith]

From their video it looked neat, then I looked at the specs. 0.3% 60000 count meter for $140. WTF

Well the ad does state "high precision" not great accuracy.   Still, is 60000 counts really high precision.  I think that Gossen I have is 300000 count and the Brymen BM869s supports 500000 counts in a few modes. 

I wonder what that PCB is that they show.  It looks like some sort of meter with the large shunt and HRC fuse.  Maybe a development tool for firmware.   

[NoMoreMagicSmoke]

From their video it looked neat, then I looked at the specs. 0.3% 60000 count meter for $140. WTF

Well the ad does state "high precision" not great accuracy.   Still, is 60000 counts really high precision.  I think that Gossen I have is 300000 count and the Brymen BM869s supports 500000 counts in a few modes. 

I wonder what that PCB is that they show.  It looks like some sort of meter with the large shunt and HRC fuse.  Maybe a development tool for firmware.

I wonder if that's not a dev board for the meter? Same number of input jacks populated, close on the buttons.

[joeqsmith]

Very possible.

[glarsson]

I wonder if that's not a dev board for the meter? Same number of input jacks populated, close on the buttons.

The display has too few digits. The board must be for a different meter.

[rsjsouza]

Interesting to see how Uni-T is targeting Fluke's 117 meter and its form factor. For about half the price you can get an interesting electrician's kit with clamp and magnetic hanger accessories. At this market, though, electrical and mechanical robustness are kings as well as track record, but for the latter there is nothing that can be done at this time. For the other two, Joe to the rescue! 

It would be an interesting test indeed, but it is quite an investment.

[joeqsmith]

At this market, though, electrical and mechanical robustness are kings as well as track record, but for the latter there is nothing that can be done at this time.

I had just posted a comment to kissanalog on their best pick of an electrician's meter and basically repeated your comment.   

[Chance92]

Found a teardown of the UNI-T UT117C (in Chinese):
https://www.mydigit.cn/thread-401149-1-1.html

[joeqsmith]

Once again, single PTC with nothing else to limit the current.  With what they are starting cost now, I'm surprised they aren't using proven designs.   They look to be about the same package as the parts that came in the 61E+. 

[Chance92]

Once again, single PTC with nothing else to limit the current.  With what they are starting cost now, I'm surprised they aren't using proven designs.   They look to be about the same package as the parts that came in the 61E+.

Do you mean it lacks a surge resistor?

[joeqsmith]

That would be common.   Still, even if they had added some other limiting device,  they would need to make sure they didn't fuck up anything else.  Tall order for UNI-T.   

[rsjsouza]

Once again, single PTC with nothing else to limit the current.  With what they are starting cost now, I'm surprised they aren't using proven designs.   They look to be about the same package as the parts that came in the 61E+.

Well, in all fairness it seems they might be learning a thing or two from your previous failure analysis. If my eyes are not failing me, the HV path seems to have a pair of PTCs, two pairs of clamping transistors and a larger resistor R57 they think might hold back a serious surge.

However, it is a hefty price for a somewhat basic Uni-T.

[Fungus]

"product accuracy can still be guaranteed after a 2-meter drop" 

https://meters.uni-trend.com/product/ut117c/#Specifications

[joeqsmith]

Once again, single PTC with nothing else to limit the current.  With what they are starting cost now, I'm surprised they aren't using proven designs.   They look to be about the same package as the parts that came in the 61E+.

Well, in all fairness it seems they might be learning a thing or two from your previous failure analysis. If my eyes are not failing me, the HV path seems to have a pair of PTCs, two pairs of clamping transistors and a larger resistor R57 they think might hold back a serious surge.

However, it is a hefty price for a somewhat basic Uni-T.

Assume R67.  I would need to trace it out.    Normally you would basically have two MOVs in series (shared on the return leg).   Ignore the second leg for now, you end up with one PTC in series with a MOV.  Say the MOV is 1kV.   If I apply 2kV, thats 1kV across the PTC.  Maybe it can handle it.  Those stupid piss ant 5mm parts they use in the cheap meters come apart and arc over.  Then there is nothing to limit the current.  In this case, let's say we have that SMT resistor in series with the clamp.  The clamp is basically a 0V and the 1kV is now across that resistor.  Maybe it survives.   Does it at 6kV?       I like the tried and proven methods some of these companies have adopted.  Still, it doesn't mean they can't still screw it up.   The 121GW for example has the basic parts but the simple grill starter damaged the production meter with ease, just like so many UNI-T products.   Designers have to to their job.     

[Fungus]

The worst part is that many people will never see past the fuse.

They'll see that fuse and think "It has the same protection as a Fluke!" even though the fuse only protects the 10A range and all the other ranges aren't any better than an Aneng.

[floobydust]

Once again, single PTC with nothing else to limit the current.  With what they are starting cost now, I'm surprised they aren't using proven designs.   They look to be about the same package as the parts that came in the 61E+.

Well, in all fairness it seems they might be learning a thing or two from your previous failure analysis. If my eyes are not failing me, the HV path seems to have a pair of PTCs, two pairs of clamping transistors and a larger resistor R57 they think might hold back a serious surge.

However, it is a hefty price for a somewhat basic Uni-T.

Assume R67.  I would need to trace it out.    Normally you would basically have two MOVs in series (shared on the return leg).   Ignore the second leg for now, you end up with one PTC in series with a MOV.  Say the MOV is 1kV.   If I apply 2kV, thats 1kV across the PTC.  Maybe it can handle it.  Those stupid piss ant 5mm parts they use in the cheap meters come apart and arc over.  Then there is nothing to limit the current.  In this case, let's say we have that SMT resistor in series with the clamp.  The clamp is basically a 0V and the 1kV is now across that resistor.  Maybe it survives.   Does it at 6kV?       I like the tried and proven methods some of these companies have adopted.  Still, it doesn't mean they can't still screw it up.   The 121GW for example has the basic parts but the simple grill starter damaged the production meter with ease, just like so many UNI-T products.   Designers have to to their job.   

I see a (2512?) SMT 1.5k surge resistor R67, three dinky MOV's and two dinky PTC's. UniT claims it has 61010 certificates but you have to wonder given their fraudulant track record and the quite small parts. They charge way too much for basic protection as a "feature" given the $2 max. in extra parts.
The PTC's there is only one in china used in all their multimeters at 1,000V rated and as you say, it just blows up.

ACA, DCA it only displays to 0.1A resolution. ICK. I want my 60,000 counts back.

[joeqsmith]

As I said, I it would need to be traced out.  Normally we would expect the to MOVs for the two legs to share a common node that would tie back to that third MOV.  I am not sure if this is what they have done or not.   The surge rated resistors are then in series with the two PTCs.  Same old same old.  If R67 is rated for surge, odd they place it after the MOV where it only acts to limit the current to the low voltage clamp.   With it all routing to the switch, hard to say.    The 121GW for example will flash over the switch in my normal testing.  The just didn't consider it.   

I doubt this particular UNI-T would withstand the transients that I consider to make a meter robust but you never know.   I've looked at more of their products than any other and I've yet to see them get it right.   I doubt very much they all of a sudden have their shit together.   Now with the high costs, why would you ever do this route.  At least with the 61E, you could say, it's only $50 but it sorta gives you some higher resolution and a PC interface.  Sure, you may look at it funny and it dies but it's $50.   

[Fungus]

So we're back to the big question: Does the meter have to survive the transient to get a safety certificate or does it only have to protect the user?

I'd say that a big bang in your hand isn't "protecting the user" - even if the blast is contained they might step back from fright, put their foot in a bucket and fall off the scaffolding.

[joeqsmith]

So we're back to the big question: Does the meter have to survive the transient to get a safety certificate or does it only have to protect the user?

I'd say that a big bang in your hand isn't "protecting the user" - even if the blast is contained they might step back from fright, put their foot in a bucket and fall off the scaffolding.

I would say YOU are back to ...  Don't wrap every one else into your safety chatter.   

I was mucking around a few weeks ago with some quartz crystals and was curious if I could tune their resonance frequency by adding a DC bias voltage across them.    I was using a meter to make these measurements and was commenting how this is a typical case where I would be concerned about using a crap meter.    Almost lost a cheap VNA that day.   Like cheap meters, they are somewhat disposable.   

I could care less about safety but I sure do care about a meters ability to survive some basic transients. 

[Veteran68]

I was hoping Joe, Dave, Kerry Wong, or another of my favorite YT'ers would look at this meter soon. I've been seeing it go for around $85 alone or $100 with the kit (clamp leads and magnetic hanger) for a 60K count meter and thought it looked like it had interesting features for the money. Sure, Dave's BM786 is a far better 60K count meter I'm sure (I do love mine), but the cost is twice as much. Not sure I've seen another 60K meter in this price range, at least from a brand you've actually heard of.

Note: I have an unashamed addiction to buying and collecting handheld multimeters of all stripes. I have some of everything, and multiples of several, including Fluke, Hioki, Brymen, Aneng/Zoyi/Zotec, Greenlee, Voltcraft, Craftsman, Uni-T, Owon, Hantek... you name it and I probably have at least one. I definitely don't need another meter. Just curious as to whether this one might scratch the itch to add to the collection, lol.

[NoMoreMagicSmoke]

I was hoping Joe, Dave, Kerry Wong, or another of my favorite YT'ers would look at this meter soon. I've been seeing it go for around $85 alone or $100 with the kit (clamp leads and magnetic hanger) for a 60K count meter and thought it looked like it had interesting features for the money. Sure, Dave's BM786 is a far better 60K count meter I'm sure (I do love mine), but the cost is twice as much. Not sure I've seen another 60K meter in this price range, at least from a brand you've actually heard of.

Note: I have an unashamed addiction to buying and collecting handheld multimeters of all stripes. I have some of everything, and multiples of several, including Fluke, Hioki, Brymen, Aneng/Zoyi/Zotec, Greenlee, Voltcraft, Craftsman, Uni-T, Owon, Hantek... you name it and I probably have at least one. I definitely don't need another meter. Just curious as to whether this one might scratch the itch to add to the collection, lol.

Take a look at the specs though. This meter is literally an order of magnitude less accurate than any other 60k count meters. What good does that extra digit do if it's not accurate??? To me, it's close enough in price to the BM786 or 789 that if I want a 60k meter I would still prefer one of them. If I wanted cheap I would get the 61E which has better accuracy than this meter.

I am VERY confused by that teardown pic and the specs. That looks like a HY3131 chip, which is definitely capable (with the correct other hardware of course) of much better accuracy than this meter.

[Veteran68]

Take a look at the specs though. This meter is literally an order of magnitude less accurate than any other 60k count meters. What good does that extra digit do if it's not accurate??? To me, it's close enough in price to the BM786 or 789 that if I want a 60k meter I would still prefer one of them. If I wanted cheap I would get the 61E which has better accuracy than this meter.

I am VERY confused by that teardown pic and the specs. That looks like a HY3131 chip, which is definitely capable (with the correct other hardware of course) of much better accuracy than this meter.

Interesting. I didn't really look into the specs, I just saw a new 60K meter from Uni-T at an attractive price point. I just assumed it would be at least as accurate as a 61E+ (which I also happen to have). I don't understand why they wouldn't take a fairly proven (if not perfect) design like the 61E+ and add the precision to it for $25 more.