Aneng SZ20 25,000 count, 1uV, 10nA, 20A

[Fungus]

Question is, WHY do they do this, is obvious that is a very simple issue to fix, you don't need special technologies or design tools to make sure that the spacing constraints are met,

There's no rational explanation except mismanagement. The person who did the PCB obviously doesn't know that stuff.

The meter works so that's it.

[Fungus]

Anyway:

Now you have one can you tell us what you think of the wrought iron screen and thunder monitoring feature?

[Kleinstein]

There are other areas with stupid small spacing, e.g. with the 2nd PTC that is rather close to the main chip set, but electrically more or less connected to the input, at least in the Ohm mode.

The spacing at the switch is a but more tricky. So this may need more space there. The switch part is quite often a weak point.

I can totally understand the rant: I can accept using cheaper fuses (though not the stupid 5x16 size), especially if the lable would be accordingly.
A sloppy layout does not reduce the BOM costs very much. No real excuse for this.

[Fungus]

A sloppy layout does not reduce the BOM costs very much. No real excuse for this.

It doesn't reduce it at all.

If they want to reduce the BOM they could get rid of a few of those drill-holes-to-nowhere in the oversized pads under the fuses. Drilling costs money.

[floobydust]

Thanks for the pics, my concern was with the HV input jack trace doing a long drive across the board.
Informal armchair looksee, the PCB layout is an ongoing weakness. Either the creepage/clearance are not consistent, or lots of dead copper, ground-pour over regions that should or shouldn't have any, and scarce via's etc. It's a layout limitation due to inexperience. Wisdom is gained through testing.

5x20mm specialty fuses can be rated almost 500VAC so the fuse PCB gap is meh, it could only arc after fuse clearing.
The HV divider MELF's are way up the board and a few traces too close. MOV traces too skinny. Three via hole sizes? I could go on.
For the (IEC 61010) 600V Cat. IV 1,000V Cat. III label it bears, well that is an 8kV impulse test and no that won't be achievable. In practice what you do is take a PCB and give it even half that and see where it arcs and crackles in a dark closet. I marked up the top side pic with a few regions I didn't like, the rotary switch is strange having solder-mask on the track, which wears off... for a design review V1.0 needs polish.

It might do fine with household use to ?V but... always test instead of speculate. It takes just one spacings error to fail. Or a component - here there is no 1kΩ surge resistor so a tiny PTC rated for harmonious 1,000V usually just blows up without that anyway. I just want to know if it's OK for mains use- to what voltage and with no certificates/approvals hard to trust.

It's an improvement over previous models. 3-LED backlight. It has an LED at the 20A jack?  Pt1.7 could have serial data 
I'm wrong about the missing NCV electrode, they are using a long trace from side to side for that.

Hopefully the LCD contrast and update rate is good. $50 for 25,000 counts I wonder how it does with fine measurements. Would scope power and upsize caps as required.

[DC1MC]

Sometimes I have these crazy post-Covid dreams  , let me share one of of them:

As far as I can see, the main weakness of this devices is not the parts, they seem to be half decent, not even the case, but the layout of the PCB that for one reason or the other, seem to not improve over time.
So the idea is: keep the case and redesign the PCB properly, and then transfer the main parts from one PCB to the other, where it makes sense (MCU, LCD & controller, some resistor, etc) and have the community designed device.
Of course is not that there is a lack of highly performant and expensive multimeters all over the place, but hear me out: if the PCB is highly performant and the result is really good, we make it public and it will be copied only with small adaptations that I hope will not destroy its purpose and suddenly will will have an abundance of cheap but safe and performant multimeters  .
Myself I can do measurements, transfer parts and do testing, hey I may even sponsor the PCB ordering and send them to the interested, if it works with cars tuning and other things, why not with the multimeters, then we send the result to Dave and count "meeehs" compared wit the original  .

What do you say about this crazy Easter idea ?

 Cheers,
 DC1MC

[Kleinstein]

The parts are also not that great, at least not for the claimed 1000 V cat 3 rating.  As far as I can see there is no series protection (e.g. fusible resistor) in case the first set of MOVs fails. There are MOVs that include a thermal link, but I somewhat doubt these as such types. For a 1000 V CAT 3 rating the 2 MELF resistors for the voltage path also look a bit small, though better than many cheap meters.

The tricky part can be the spacing at the switch, though it does not look that bad. It looks like the the wiper next to the critical switch part (outer 2 traces towards the terminals) is not populated. So the solder mask covered part in the switch just looks strange - though a little close to the input voltage. The 2 small uncovered spots may be for testing - though still odd.

For the low voltages it may be interesting how good the resolution really is - not just at 0 V (in case there is a dead band). Having a dead band and thus crippling the low voltage range would be a show stopper in investing more time in the PCB.
For the µA range, I don't see a big issue. The current is not super low and it is just a simple shunt. No special problems expected there.

[Fungus]

$50 for 25,000 counts...

But basic accuracy is only (0.2%+10)

Aneng AN870 is 20,000 counts and (0.05%+3) for less money.

https://www.aliexpress.com/item/33052395019.html

[indman]

$50 for 25,000 counts...

But basic accuracy is only (0.2%+10)

And here is a screenshot with characteristics for a similar multimeter, only with a different marking!
NEWRUIKE XCRK61E. Specified 0.05%+3 basic accuracy. Which characteristics do you suggest to trust more? Or does it have a different filling inside?

[Fungus]

Which characteristics do you suggest to trust more? Or does it have a different filling inside?

Looks like it came out of the same factory to me. 

Aneng are rebadgers, I don't know if they make any meters themselves. They're sold a lot of Zotek meters in the past. Now it looks like they're using somebody else.

FWIW many Aneng meters have "ZT" part numbers on the PCB, eg. the AN870 shown on this page:



I can't see any markings on the SZ20 PCB pics posted by DC1MC though.

(maybe they're under the LCD...  )

[mwb1100]

Aneng are rebadgers, I don't know if they make any meters themselves. They're sold a lot of Zotek meters in the past. Now it looks like they're using somebody else.

FWIW many Aneng meters have "ZT" part numbers on the PCB ...

I can't see any markings on the SZ20 PCB pics posted by DC1MC though.

The Aneng SZ20 is a Zotek (aka Zoyi) ZT-225

The Aneng PCB mentions "225" without the "ZT".  Maybe a weak attempt to obscure the link?  Probably nothing so deliberate - just what the board layout guy typed in.

[indman]

Don't you understand yet that all cheap counters are made by Chinese friends on their knees in a dim basement, and then different labels are stuck on the same case - ZOTEK, ANENG, NEWRUIKE and others. And the characteristics are also what you want, they will draw them in the table!

[DC1MC]

Don't you understand yet that all cheap counters are made by Chinese friends on their knees in a dim basement, and then different labels are stuck on the same case - ZOTEK, ANENG, NEWRUIKE and others. And the characteristics are also what you want, they will draw them in the table!

What about the highly expensive counters, are they made by Americanians in highly advanced SF laboratories  ?

[indman]

What about the highly expensive counters, are they made by Americanians in highly advanced SF laboratories  ?

No, I think they are also made in China, only in contrast to the conditions of the basement, Chinese friends wash their hands there more often.

[Fungus]

The Aneng SZ20 is a Zotek (aka Zoyi) ZT-225

Hah! They fooled me by not putting it on their web site.

I wonder what the true specs are.

different labels are stuck on the same case - ZOTEK, ANENG, NEWRUIKE and others.

Zotek is the manufacturer.

"Zoyi" is the official Zotek brand, everyone else is a rebadger.

 

[indman]

Zotek is the manufacturer.

Zotek is the manufacturer advanced SF laboratories in China?! You made my house slippers laugh

[Fungus]

What's the diode test voltage on this meter?

(and why do they never publish that number?  )

If the 0.05%spec is the correct one then this could be a meter worth owning. Maybe the ZT219 (AN870) has a successor...

I like that it has a separate dial position for each function.

[tooki]

BS1362 fuses (6.3mmx 25mm) are cheap and readily available in the UK. They open safely with a dead short on 230V mains.
Rated at breaking 264V AC at 6kA.
So much better than most meters and probably all most people would ever need.

Yep, I don't know why they aren't used in everything.

Well, some users (electricians, especially industrial electricians, though almost none of us electronics people) need safety ratings beyond CAT II, which is where the BS1362 fuses operate.

I’d actually love to see BS1362 fuses used in cheap multimeters, combined with an honest CAT I or II rating, instead of little glass fuses or none at all.

[Fungus]

I’d actually love to see BS1362 fuses used in cheap multimeters, combined with an honest CAT I or II rating, instead of little glass fuses or none at all.

Don't expect it from China. The precedent has been set, the horse has bolted, the ship has sailed.

[DX1]

Could someone please do a test of the uV range? For example measuring current shunts or finding shorts on ground planes. Looking for noise, monotonic increase of measured uV with increase of input, thermal effects and accuracy. A power supply with voltage divider (using good resistors)  is probably a good enough source of microvolts, though would be useful to check against a proper meter.

[DX1]

I am amazed nobody has access to surge testing equipment. It would be really helpful to do some actual tests on these cheap meters. This meter seems better than most, but the PCB does not seem to be laid out that well and the fuses and MOVs could be bigger. MOVs have a limited surge life and can only take a few of the rated surges, after which they can be prone to thermal runaway, which could be a real risk if you are trying to fix a fault. You can now get MOVs with an integrated thermal fuse.
What we need is honest ratings, preferably with a test report from a credible test house. You are not going to sell these meters to someone who actually needs 600V Cat IV. Even if it passes the tests I would still buy a reputable make for this.

[Kleinstein]

I am amazed nobody has access to surge testing equipment. It would be really helpful to do some actual tests on these cheap meters. This meter seems better than most, but the PCB does not seem to be laid out that well and the fuses and MOVs could be bigger. MOVs have a limited surge life and can only take a few of the rated surges, after which they can be prone to thermal runaway, which could be a real risk if you are trying to fix a fault. You can now get MOVs with an integrated thermal fuse.
What we need is honest ratings, preferably with a test report from a credible test house. You are not going to sell these meters to someone who actually needs 600V Cat IV. Even if it passes the tests I would still buy a reputable make for this.

There is no real need to destroy a meter that is not yet bad. It is obvious that the protection is not that good to be used for high voltages and even if the meter survives the test the MOVs may show leakage.

The real interesting test would be the performance with low voltages. That is what this meter and few other new ones offer than many other DMMs don't.

[Fungus]

Joe's done some Anengs:

[floobydust]

Realistically, true Cat. III 1,000V/600V IV ratings to 8kV impulse require: around 8mm PC board spacings, 10x38mm fuses, tough MOV's, bigger PTC, surge resistor etc. ... which make a multimeter huge and expensive. It's not realistic for these low cost products, you cannot have your cake here. People here know the label, claim is a deception. This is all about price and we'll compromise anything for getting it cheap.

Real surge testing equipment is very expensive and destructive, you'd need a pallet of these multimeters to test. Nobody here would do it for three reasons - one, it's gonna fail. Two, you can't "test in" safety, you have to design it in first. Think about it, will the plane go to 500km/h? Blows up at 200km/h, now what? Rate the plane to 199km/h? but a windy day...
Three, you don't want to endorse a test-pass for the legal repercussions and the politics of safety approvals.

If this meter can withstand 1,000V safely (and that is with no mains transients to 8kV claimed or expected), like multimeters of old, then that is good enough as Cat. II use.
I've worked in Cat. III, IV and would never use anything without formal 61010 testing and certifications there. That's why I still have my hands.

It's interesting the spiral around fake 61010 claims and an unknown truth about the usage safe limits.
The user manual and prerequisite front pull-off sticker on the amps jack kind of tell you not to use this up there yet the jack Cat. xx label is commonplace malarkey.

[bastl_r]

I am amazed nobody has access to surge testing equipment.

Hi
Do you mean surgetests like this? 

by the way: The Aneng 870 is a good cheap Multimeter. I paid for my item 25$ on a black friday sale.