Poll

How many cycles will the KeySight U1281A's detent spring last?

0-2000
7 (17.9%)
2k-4k
5 (12.8%)
4k-8k
14 (35.9%)
8k-16k
7 (17.9%)
>16k (most rubust meter ever made)
6 (15.4%)

Total Members Voted: 36

Author Topic: Handheld meter robustness testing  (Read 1149759 times)

0 Members and 3 Guests are viewing this topic.

Offline MacMeter

  • Frequent Contributor
  • **
  • Posts: 321
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1875 on: August 27, 2017, 05:54:59 pm »
Amazing amount of custom modding. While I understand little of it, still find it entertaining. I would be heartbroken when it failed after all that time and effort, so I appreciate how far you go in your experiments!
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1876 on: August 27, 2017, 06:51:18 pm »
Amazing amount of custom modding. While I understand little of it, still find it entertaining. I would be heartbroken when it failed after all that time and effort, so I appreciate how far you go in your experiments!

Don't feel too bad.  To make all these mods, run the tests, film it, edit it was basically a day.  This was a lot less time than I spend on many of the videos I make.  The goal was not really to show how to try and improve a $15 meter but provide some education.  And while the meter dies in the end, and we know the root cause of failure.   It's not like this was the first piss ant PTC I have seen fail.  I doubt the added cost of using better parts is all that much.   I added a few pictures of the ZT102 next to the Fluke 107.  The Fluke 107 was tested at the highest levels I have generated to date.  Also note that series resistor.  They know... 
 
The following users thanked this post: floobydust

Offline MacMeter

  • Frequent Contributor
  • **
  • Posts: 321
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1877 on: August 27, 2017, 07:24:37 pm »
How many here or anywhere would take a full day to mod a $20 meter, just to test it to failure, for NO financial compensation? You are far too modest, and way UNDERPAID! :)
 
The following users thanked this post: joseph nicholas, drumbum

Online floobydust

  • Super Contributor
  • ***
  • Posts: 6926
  • Country: ca
Re: Handheld meter electrical robustness testing.
« Reply #1878 on: August 27, 2017, 08:33:06 pm »
It's for science  :-+

P.S. I think the big WW power resistor is needed to help the PTC take the hit.
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 6451
  • Country: hr
Re: Handheld meter electrical robustness testing.
« Reply #1879 on: August 27, 2017, 09:27:32 pm »
It's for science  :-+

P.S. I think the big WW power resistor is needed to help the PTC take the hit.

Big resistor is probably carbon type, specially designed for this purpose... That is one area where carbon resistors outperform other types...
 

Online floobydust

  • Super Contributor
  • ***
  • Posts: 6926
  • Country: ca
Re: Handheld meter electrical robustness testing.
« Reply #1880 on: August 28, 2017, 12:05:00 am »
"big resistor" seems wirewound: 3k5 WW 5% 5W 20ppm on the 87, 867, 27 etc. in series with PTC.
I think I see 1k ohm probably MOX (green body) on Asian DMM's, Brymen.

Carbon comp power-resistors are expensive and end-of-life now, drift is terrible.
Porcelain-coated nichrome is pretty good for surge power handling, but turn-turn voltage breakdown might be the limiting factor.

I never did find much short-term (<10msec) impulse data/overload specs for resistors. Did an automotive load-dump design running a resistor at >100X overload and worked with Vishay engineers on it. It was hell because the SMT WW resistor's thermal properties, wire melting point were modeled.

A 6kV impulse, couple 1.5kV (clamping) MOV's, 1.5kohm PTC and 3.5kohm 'big resistor", that's only 0.6A instantaneous or 1.3kW for it, and 0.54kW for the PTC. Hmmm.
If my numbers are reasonable, 250X overload but really brief but really big...

 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 6451
  • Country: hr
Re: Handheld meter electrical robustness testing.
« Reply #1881 on: August 28, 2017, 07:50:06 am »

Carbon comp power-resistors are expensive and end-of-life now, drift is terrible.


They are non inductive, and spreading current and voltage field uniformly across... Very good, very robust.
Drift and tolerance is non important with the way they are connected into circuit..

I agree, ceramic composition solid body ones are stealing the show with even better characteristics...

There are thick film and wirewound resistors with enhanced surge ratings, but they have lower max overvoltage, good energy ratings though....
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1882 on: August 29, 2017, 01:13:03 am »
How many here or anywhere would take a full day to mod a $20 meter, just to test it to failure, for NO financial compensation? You are far too modest, and way UNDERPAID! :)

The problem with taking money or even meters is it could be viewed as a conflict of interest.   Dave presented a rare opportunity which was a little hard to pass up and I am sure there were people out there thinking I was in Dave's pocket, all the while Dave cringing as I continued to push his meter.  :-DD 

I try and repeat the same set of tests, collect the data and let you see it for free.  Fairly cut and dry. 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1883 on: August 29, 2017, 02:02:49 am »
It's for science  :-+

P.S. I think the big WW power resistor is needed to help the PTC take the hit.


Shown are the Fluke 115 which was tested up to 12KV 50us FWHH pulse with no damage), the HIOKI DT4252 (which started to arc around a plastic spacer at 10KV but with some added plastic, made it to 14KV 50us FWHH with no further damage), and while the Gossen M248B has lots of problem the one thing it has going for it is that's it very robust, surviving 12KV 50us FWHH).   I've shown you the Fluke 107.  Again, I've never had the 101 apart.  Everyone has seen the Brymen BM235.  These are all CAT III 600V and up rated and certified for both EMC and safety.  This is true for the BM235 as well.  BIG PTC, series resistor.  It's not back magic and it's not what you find on cheap meters.

Meters like the UT61E need a little help.   


Online floobydust

  • Super Contributor
  • ***
  • Posts: 6926
  • Country: ca
Re: Handheld meter electrical robustness testing.
« Reply #1884 on: August 29, 2017, 04:46:35 am »
The Gossen's gas tubes are fast enough?

The Hioki PCB DT4252 layout- a 1,000V fuse with tiny spacing between the trace and clip.
I'd expect an arc there after the fuse clears. Sigh. Engineer's screwup telling PCB CAD guy that both sides of a fuse are at the same potential. Not after it blows...
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1885 on: August 29, 2017, 11:12:39 am »
The Gossen's gas tubes are fast enough?
Three meters I have ran used GDTs.  Both this Gossen and HIOKI survived.  There was also a Keysight meter that failed at 5KV.  Because the Keysight failed with the small generator, it was a candidate for the half cycle generator as well.  Not a great test but the added energy gives you some idea how other areas in the meter may fair once we have a breakdown.   

The Hioki PCB DT4252 layout- a 1,000V fuse with tiny spacing between the trace and clip.
I'd expect an arc there after the fuse clears. Sigh. Engineer's screwup telling PCB CAD guy that both sides of a fuse are at the same potential. Not after it blows...
There is no need to guess about the HIKOI.  If you look on page 31 of this thread, we spent some time going over it.   

EN 61010-1:2001
Quote
Fuse holders with fuses intended to be replaceable by an OPERATOR shall not permit access to parts which are HAZARDOUS LIVE during fuse replacement.

EN 61010-2-033:2012
Quote
101.3.2 Protection by a certified overcurrent protection device
If the protection device is a fuse, it is replaced with an open-circuited fuse. ....
A voltage of two times the highest RATED voltage for any TERMINAL is applied to the TERMINALS of the overcurrent-protected measuring circuit for 1 min. The source of the test voltage shall be capable of delivering 500 VA. During and after the test, no damage to the equipment shall occur.

Online Fungus

  • Super Contributor
  • ***
  • Posts: 16561
  • Country: 00
Re: Handheld meter electrical robustness testing.
« Reply #1886 on: August 29, 2017, 11:24:21 am »
There is no need to guess about the HIKOI.  If you look on page 31 of this thread, we spent some time going over it.   

Link!
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1887 on: August 29, 2017, 11:46:11 am »
There is no need to guess about the HIKOI.  If you look on page 31 of this thread, we spent some time going over it.   

Link!
Screwdriver drop test!!! :popcorn:

Online Fungus

  • Super Contributor
  • ***
  • Posts: 16561
  • Country: 00
Re: Handheld meter electrical robustness testing.
« Reply #1888 on: August 29, 2017, 12:41:27 pm »
I though you were going to add links to specific meters on the first page as they came up.  :popcorn:
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1889 on: August 29, 2017, 11:14:01 pm »
I thought you were going to do a drop test on some meters.  :popcorn:

Offline CustomEngineerer

  • Frequent Contributor
  • **
  • Posts: 463
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1890 on: August 29, 2017, 11:21:30 pm »
There is no need to guess about the HIKOI.  If you look on page 31 of this thread, we spent some time going over it.   

Link!
Screwdriver drop test!!! :popcorn:

That made me laugh a lot more than it should have.
 

Online floobydust

  • Super Contributor
  • ***
  • Posts: 6926
  • Country: ca
Re: Handheld meter electrical robustness testing.
« Reply #1891 on: August 29, 2017, 11:38:42 pm »
UL 61010 certifiers told me spacings must be met around the entire fuse; (not just the end-bells or clips). Why :-//  the mid-section is ceramic.
Their answer:
Because you never know exactly where inside a fuse the link has melted i.e middle, left, right- regulatory consider the entire body of the fuse energized to hazardous live, and carbonized (=conductive).

So I had to do placement and PCB layout that leaves a large island around the entire fuse. If parts are too close, you see heatshrink covers used.

It's too bad a 5x20mm fuse could not be developed with a (DMM) 1kV high-interrupt rating. These DMM fuses are pretty huge.
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1892 on: August 30, 2017, 11:11:28 am »
UL 61010 certifiers told me spacings must be met around the entire fuse; (not just the end-bells or clips). Why :-//  the mid-section is ceramic.
Their answer:
Because you never know exactly where inside a fuse the link has melted i.e middle, left, right- regulatory consider the entire body of the fuse energized to hazardous live, and carbonized (=conductive).

I assume you are referring to a hand-held meter. 
EN 61010-2-033:2012
Quote
Additionally, spacings surrounding the overcurrent protection device in the equipment and following the protection device in the measuring circuit shall be sufficiently large to prevent arcing after the protection device opens.
This is under 101.3.2 and requires the same test as above.  This assumes you are using a certified fuse. 

So I had to do placement and PCB layout that leaves a large island around the entire fuse. If parts are too close, you see heatshrink covers used.

It's too bad a 5x20mm fuse could not be developed with a (DMM) 1kV high-interrupt rating. These DMM fuses are pretty huge.
I've shown some meters that were really bad in the fuse area.  As long as you are self certified, I guess you can get away with it. 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1893 on: September 06, 2017, 01:51:05 am »
Trying out some StackPole ASRM series parts.  These are a metal film, flame retardant, pulse, blah blah safety resistors.  Also some tiny little EPCOS 500V PTCs.  The PTCs by themselves would survive a direct hit from the low voltage generator.  Well, 20 hits anyway.  With the high voltage generator turned up, they came apart with a single hit.  The resistors will also arc across with a single hit with this setup.  Combining the two, it survived 20 hits at 15KV before I turned it off.  No sign of any problems.  Way outside the ratings of both parts but may not be too bad a combo to try. 

Picture showing the StackPole part after the single hit. 
 
The following users thanked this post: 2N3055

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1894 on: September 06, 2017, 05:40:03 pm »
Time to try some OX series parts.   Surge ratings are much better.  May only need the one part.   

Online floobydust

  • Super Contributor
  • ***
  • Posts: 6926
  • Country: ca
Re: Handheld meter electrical robustness testing.
« Reply #1895 on: September 06, 2017, 10:49:44 pm »
joeqsmith, it's good research you doing- anything to find a compact PTC/resistor solution for DMM front-end protection.

The only decent technical paper I found:
Pulse Handling Capabilities of Vishay Dale Wirewound Resistors

"...The cross-over point is the time where significant energy starts to be dissipated not only in the [wirewound resistor's] wire itself but is now being dissipated into the core, leads, and encapsulation material. This is the point where the pulse is no longer considered a short pulse, but is now considered a long pulse."

So we are under the cross-over point and my hunch is the resistors are breakdown voltage-limited. OX/OY look pretty tough, but 3/4" is big and their 14kV/20kV rating with tiny 1,000pF discharge seems tame. I think 6kV for Cat. III is reasonable?

I considered Vishay WSC/WSN series (SMT) but no info about 50usec impulse at high-voltages.
 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1896 on: September 07, 2017, 12:42:02 am »
I wonder if we say 6KV peak, 1K resistor and a 1K PTC, does it divide for the pulse?  Many of the PTCs I look at will give thermal time constants in the tens of milliseconds with an amp applied.  At 3A and up, we are outside any of the parts I have found.   The plots are showing the voltage across both the PTC and series resistor and just across the PTC.   PTC is 1.3K R is 1K so not quite 50%.   At M1/M2, pulse is around 1KV peak.  Seems PTC does not react.  At C1/C2 1.8KV, we can see the PTC appears to be starting to open.   In the second set, the pulse is over 4KV and now we can really see the PTC starting to change.  A physically larger part, more thermal mass, may not budge at all.  This may be better as the PTC seems to be the weak link.       

The energy rating for the stackpole parts compared with the Ohmite OX is much worse.  If I could find the room, I would use the OY with the large PTC.   I also have some old Ohmite little devil and little demon parts that would work but those parts are no longer available.   

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1897 on: September 08, 2017, 03:04:11 am »
A brand new Kasuntest ZT102 arrived today.  I started to run the normal functional test and discovered this meter will read Hz and duty cycle in the volts mode just like the AN8008 I had ran.  All three meters have the same revision PCB, came in the same box, included the same probes.   Now the really strange part, the SN of the latest meter is actually older than the last one I bought!  I assume this is setting in the PROM.  Maybe someone loaded the wrong image into the programmer that day but all the other features work correctly.   Maybe they recycle through serial numbers and this is actually a newer meter.   Also the springs used for the selector switch are silver like the first one I bought but all three use unique crystals.  Lots of variance in the process.

I have a few more things to test before I will do anything to the meter.  Maybe this weekend I will have some time to work on it. 

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1898 on: September 09, 2017, 01:30:44 am »
AN8008 CALIBRATION BASICS
I've figured out how to do calibration although I don't have the finer details worked out.  Here's what I know...

1. Short out J1 on the PCB (it's at the top right corner as you look at the PCB from the back of the meter)
2. Get your calibration reference ready and connected - in my case a 300 mA DC feed through the mA/A terminal and Common
3. Turn rotary control from off to the resistance position and CAL will appear on the display
3. Now WAIT until you hear a beep and then move the rotary control to the mA/A position - if you turn the rotary control away from resistance too quickly it doesn't show the values you've selected but moving the rotary switch around corrects that
4. Press [Set/Hold] (orange) button repeatedly until you see DC mA and a value will be displayed
5. Now you should see a value close to what you're providing the meter with, around 300 mA DC in my case
6. Press the [Range] (blue) button (quick press) to range down (but it only does it in 0.1 A increments)
7. Press and hold the [Range] (blue) button (long press) to range up (but it only does it in 0.1 A increments)
8. Press [Set/Hold] (orange) button to move off that setting (I think this is when the cal change just made is saved)
9. Move the rotary switch to off
10. Clear the link on J1 and power back on and test

As far as I can tell, the trick is to set the input to an exact value like 300 mA so you can set that value on the display during cal because you can't adjust the display to 303 mA so, for my slight discrepancy, I saw 298 mA displayed, I ranged down - I saw 200 mA, and then ,with a long press of [Range], ranged up and then I saw 300 mA displayed, pressed the orange button again and I was done.  Other parameters can be set by pressing the [Set/Hold] (orange) button repeatedly but you'll only see values corresponding to the rotary position selected so you'd leave it in the resistance position to cycle through the measurements associated with that position including resistance.

What I'm not clear about is when it actually stores a new calibration value, I guess that, once you use the [Range] (blue) button, it changes the calibration for that setting, and I think it saves it when - having got the display to show the value you want, you press the orange button once more. 

At first I did my above procedure providing 300 mA but ranged down to display 100 mA  and couldn't change that value because I hadn't figured out that a long press ranges up by that point.  So when I then switched off and removed the link, it had calibrated the meter to display 100 mA when 300 mA was supplied which it did.  I had to re-calibrate after I figured out that the long press increases the displayed value.

After I posted a video where I had modified the KZ102 (AN8002) to the capacitance readings were off about 100pF.  Someone had wrote me about modifying the contents of the PROM to realign it.   Seems like a lot of work.  Does anyone know if the above procedure applies to the AN8002 (and others) as well?  For capacitance and current, what are the standard values that are required?

I played around with the Z102's alignment.  Range is increment.  Hold is dec.  Mode select, selects the mode to align.  I made up some caps from my RLC meter as a reference.  There is a limit on what capacitors I can use and there is nothing it will accept in the 500pf range.  No null so seems I have to live with the capacitance or there is another trick.   

Offline joeqsmithTopic starter

  • Super Contributor
  • ***
  • Posts: 11631
  • Country: us
Re: Handheld meter electrical robustness testing.
« Reply #1899 on: September 09, 2017, 04:13:55 pm »
They always look so nice in the box when they first arrive.   Not much going on today so time to drag out the Dremel, iron and dope.  Maybe this time I will go in a little finer steps so we can get a better idea where it fails at. 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf