EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: norEL on February 05, 2017, 08:38:40 pm
-
This is my 7year old Agilent U1253B DMM. The meter has worked fine until a couple of months ago. Took the meter apart and found the fault. Contacted the dealer who sold the meter new. He agreed to take the meter in and send it to one of keysights labs. I just got responded with a repair cost that was way too expensive for this old meter. Got the meter back with no explanation on what had gone wrong and why this had happend to my meter...
Video:
https://youtu.be/K573yx-OFOQ
-
In the video where the rotary switch burn marks are shown, its not very clear, the image is blurry. Could you redo the video and focus on the burned area?
A common reason for those burn marks is a high voltage spike or surge. The protective elements prevent a more serious blow up, but its can leave damage on the PCB, including other components beyond those switch traces.
A concern I always had for Keysight/Agilent designs is they rely on gas discharge tubes as a main element in surge protection. Like Joe Smith, I think they are not fast enough to protect the internal circuitry, but it will provide protection for the user.
In an old video, Joe showed an experimental surge with the arc happening in a Keysight meter. The video shows the arc on other sections of the board too. For the switch see 8:50 and 9:40
https://www.youtube.com/watch?v=lE5CKaLzvqU&t=757s (https://www.youtube.com/watch?v=lE5CKaLzvqU&t=757s)
-
Yes, the image of the damage got blurry. |O Thank you for notifying me! :-+
Something I forgot to mention are the deep scratches the rotary switch have made in the PCB.
Instead of making a new video, here are some images of the PCB.
To me it looks like it has got warm when I was measuring over longer periods. But if so more would be black around maybe.. If this damages are from surge voltages I would have noticed it I think? And the meter have only been used to meassure 300vac/vdc and below..
I also do not trust the Keysight protection as I did before.. This is a higher end class meter..
-
Thanks for photos. It does look like surge damage.
(https://www.eevblog.com/forum/testgear/keysight-agilent-u1253b-killer-switch/?action=dlattach;attach=290437;image)
-
Then it must have happened when I was measuring/logging over some period of time without my presence at the moment.. But still, if it was surge it looks like it had to be some power in it and then I really think the protection on this meter is too bad!
-
The repair looks like it would be a swap of the circuit board, which is why it would cost so much. I am not too surprised they would not cover it under warranty.
In the video saturation links, there really is not enough energy to do much damage. You having the meter on actual AC mains gets the benefit of a much better test than what I am doing on the bench. I'm not a fan of their input protection. After seeing Dave's review on some new Keysight meter he had received, and hearing him say how it had GDT's so no worries, I wanted to see for myself. Not that GDTs are bad, just they may not be fast enough without some additional protection to save the meter. The HIOKI I ran also used GDTs and that meter did very well in the testing.
After I damaged the Keysight, I found the CPU was drawing excessive currents and there was no signs of life (clock dead). I had no plans to try and repair it so I ran the half cycle test on it with it set to the AC volts mode. You can see the results here:
https://youtu.be/ju9JcAzgOeo?t=110 (https://youtu.be/ju9JcAzgOeo?t=110)
Again, this is a fairly small transient but similar to your real world results.
You may want to read the following document put out by Fluke that talks about line transients. http://media.fluke.com/documents/1263690_6003_ENG_J_W.PDF (http://media.fluke.com/documents/1263690_6003_ENG_J_W.PDF)
-
Some troubleshooting tips, is there any damage upstream of the switch? If so, its more likely to have been a kV transient [ my error, surge is the wrong term to use.] If custom chips were damaged, the meter is a loss.
Its not likely prolonged logging over time damaged it because the input impedance of DMM is so high; to cause that much burn a low impedance path has to have happened to provide the total energy, it could also be debris inside the meter, or one of wiper arms of the switch broke off and fell between the traces.
Can you take a photo of the back of the switch, where the wiper arms are?
For what its worth, its not often to see a surge transient fault [ or a broken wiper arm] with photos and details, so its worth looking into.
Then it must have happened when I was measuring/logging over some period of time without my presence at the moment.. But still, if it was surge it looks like it had to be some power in it and then I really think the protection on this meter is too bad!
-
Yes, the repair was a swap circuit board and the cost was more then a new meter at my dealer so...
I have not bought me a new one yet because I haven't decided which one to buy.. Think I will buy a other brand then Keysight.
Have several meters from Keysight that is almost the same age that have worked and are working just fine, but because of this incident my mind have decided that Keysight have bad input protection and my life is worth something better then this.
There was no loose parts in the meter and the meter functions on all ranges but with error readings on some of them.
Here is a picture of the switch contact plate. You see signs of a burst of something..
Thank you for commenting this, I appreciate your help!
-
Thanks for the photos. You are welcome. It looks like the wipers on the rotary switch are untouched and intact. In a blow up you can see the blast burn mark is on a different section of the rotary switch so the event happened separate from the switch contacts, on the copper strips alone on the PCB.
(https://www.eevblog.com/forum/testgear/keysight-agilent-u1253b-killer-switch/?action=dlattach;attach=290572;image)
Because of the transient protection in a CAT III setting, a typical transient on the line is often not exciting, you may feel or hearing nothing, or at most a hear a small crack or pop, and the DMM appears to go berserk right at the moment. If its damaged, the meter will not work right after the event, if not, it will be return to reading normal; Joe Smith's videos show how the meters will appear to react when hit.
-
I don't have one of these meters but it should be easy enough to see what mode the meter was in when it happened.
-
You may be able to repair this meter so that it works again? Remove the carbon residue first and see if the meter makes proper measurements. Some hints.
http://mrmodemhead.com/blog/ideal-carbon-footprint/ (http://mrmodemhead.com/blog/ideal-carbon-footprint/)
https://www.eevblog.com/forum/repair/fluke-787-repair-any-idea/ (https://www.eevblog.com/forum/repair/fluke-787-repair-any-idea/)
-
You have a point there " #saturation ", the blast marks are not on the contact's on the switch.
I sat down with the meter to check if I was able to set the switch between two ranges, and that was fairly easy! There is like
a dead spot between the ranges when turning the switch..
So could that be the case, that I had accidentally got the switch in a between position and that made the fault??
Maybe I will try to fix the meter " #retiredcaps " thank you for the tips and links!
-
Hi, its not likely, because then the switch would have some burn mark too. But If all is well but the trace on the PCB, it may be repairable:
http://www.engineeringlab.com/conductor-procedures.html (http://www.engineeringlab.com/conductor-procedures.html)
Best of luck.
You have a point there " #saturation ", the blast marks are not on the contact's on the switch.
I sat down with the meter to check if I was able to set the switch between two ranges, and that was fairly easy! There is like
a dead spot between the ranges when turning the switch..
So could that be the case, that I had accidentally got the switch in a between position and that made the fault??
Maybe I will try to fix the meter " #retiredcaps " thank you for the tips and links!
-
Thank you for the link " #saturation " ! If I got time to do a repair I will post the result here.