Agilent U1272A crappy soldering

[wraper]

I taked apart my new Agilent U1272A multimeter and saw that a places where parts were not populated looks weird. Inspected it with a microscope and wow, all soldering is a pure crap. On some places it looks like that the parts will fall off if I touch them. Seems that a soldering paste they used was well expired. I can't believe this , shame on you Agilent. All images are here: https://plus.google.com/photos/117212633093763827202/albums/5870856500768671777

[DavidDLC]

Man, I will return it immediately. You mentioned it is brand new.

David.

[wraper]

I bought this as Agilent bench PSU + free U1272A promotion about a week ago at EU Farnell.

[DavidDLC]

It does not matter, you should receive a quality product.

Contact them and tell them about the problem.

David.

[KJDS]

That's either very poor solder paste or something is amiss with the thermal gradients in the oven.

It's certainly not acceptable.

[2nOrderEDO]

I was very happy because that PSU + Free Agilent multimeter offer. We should wait until someone else buy that kit and teardown the multimeter. Maybe they have a bad batch and they are offering this promotion because they want to get rid of their stock.   

[sullyRD]

Wrapper is correct "shame on you Agilent."...Bad Quality control, a meter like that will not last long i would return it.

[casper.bang]

Shit.. just ordered the same combo. However, Dave did a teardown which did not reveal anything remotely this nasty.

[Corporate666]

That is absolute crap! 

I have had boards come out like this and it was when we tried to use solder paste that was way, way beyond the expiration date and that had been left out and had most of the flux evaporate. 

Solder paste is actually very accommodating - you have to *really* abuse it to get results like that.  That is shockingly bad... should never have been let out the door like that.  Hope you can return it for one that was made right!

[2nOrderEDO]

Shit.. just ordered the same combo. However, Dave did a teardown which did not reveal anything remotely this nasty.

Teardown your multimeter and keep us updated!

[uoficowboy]

I took apart one of our U1272As that we got through the promotion and the soldering looked just fine on it. I think you got a bad batch.

[Barny]

My U1272A should arrive tomorrow.

I'll take a look and share the results.

[wraper]

If anyone takes apart his multimeter to check this please use some kind of magnifying. It didn't look bad at first glance when I looked with naked eyes.

[Mike Warren]

That's some of the crappiest silk screening I've seen for a long time, too.

I, too suspect that this is why they are being given away, but is it really worth Agilent's good name to do something like this?

[Lightages]

So with the bugs in the software, the weird recovery from over voltage on the ohms range, and now this do we need to rethink Agilent as something to recommend? I hope we get to see inside more of these from others soon.

[BravoV]

You should pm Mike Kawasaki, he works for Agilent and also a forum member here -> https://www.eevblog.com/forum/index.php?action=profile;u=12309 or email him directly mike_kawasaki@agilent.com pointing to this thread to hear his reply here.

This is definitely a bad press for Agilent, they should rectify this asap, not good especially in this highly competitive market for handheld DMM.

[Corporate666]

So with the bugs in the software, the weird recovery from over voltage on the ohms range, and now this do we need to rethink Agilent as something to recommend? I hope we get to see inside more of these from others soon.

I actually hope that someone from Agilent gets in touch with this forum, or at least with Dave.  I notice the OP is in Latvia.  I am not sure if Agilent uses a central facility for manufacturing or regional centers, but there is no way such soldering work can be in line with company standards.  So the big questions are what happened and what are they going to do to make sure it doesn't happen again?  I wonder if it was just a lazy employee at Agilent... or perhaps it's a Chinese supplier trying to save a few bucks by using old expired solder paste? 

[BravoV]

I wonder if it was just a lazy employee at Agilent... or perhaps it's a Chinese supplier trying to save a few bucks by using old expired solder paste?

With the price range they charged, if this happened to me, I don't care, its their problem, I'd expect they will provide a full replacement and free of charge. 


To Dave : If you're reading this, you should make some noise at your tweet there about this.

[EEVblog]

Shit.. just ordered the same combo. However, Dave did a teardown which did not reveal anything remotely this nasty.

Not at all. Mine was acceptable.
http://www.flickr.com/photos/eevblog/sets/72157626761928610/

[reagle]

That's some of the crappiest silk screening I've seen for a long time, too.

I, too suspect that this is why they are being given away, but is it really worth Agilent's good name to do something like this?

Drill registration is a crappy as well

[poorchava]

Looks to me like the PCBs might have expired. I have seen similar effect on some occasions when the vacuum sealed container with PCBs was damaged. Even ENIG goes bad after some time. HASL or OSP even more so.

Maybe they have realized that they manufactured a bad batch and decided to give them away thinking that people won't file that many claims ("well, it was free, what did u expect").

[casper.bang]

Teardown your multimeter and keep us updated!

I definitely will. The combo is in back-order though, back in stock may 1'st so I reckon it will be a few weeks before I receive it.

[wraper]

Hope you will get different batch then.

[blackdog]

Hi,

I opend my Agilent u1272a and the soldering is CRAP!


Soldermask is OK.


This meter game with a Agilent powersupply.

I have never seen this bad quality in a HP ore a Agilent product :-(

Kind regarts,
Bram

[casper.bang]

Assuming Agilent is aware of this and it's at the heart of their free campaign (meaning not liable for guarantee since it was never technically purchased), what is the prospect of repairing it yourself? I assume a reflow ain't going to change much?   

[Barny]

Even you got the multimeter for free, you have warrenty at the multimeter.
The multimeter ist practical a materially discount and part of the contract.

[BravoV]

Even you got the multimeter for free, you have warrenty at the multimeter.
The multimeter ist practical a materially discount and part of the contract.

Let alone warranty, how about the words like "top tier" or "product consistency" ?

Do these jargons still exist in their dictionary ? 

[eliocor]

In mine 2 pieces of U1272A the soldering is in perfect state!

Also please DO NOT confuse RoHS (Pb Free) compliant soldering (which seems rather dully) with the SnPb one which is shiny.

[digsys]

I opend my Agilent u1272a and the soldering is CRAP!
I have never seen this bad quality in a HP ore a Agilent product :-(

Something is fishy with yours !! The ICs have nice leaded solder, while the rest of the components use sand mixed with spit.
Weird. It can be a 2 part process .. maybe ICs soldered by hand? Questions.

[wraper]

blackdog, where and when did you get yours?

[Barny]

I've checked my U1272A

The soldering is not shiny but looks ok.
Only one chip looks a little bit suspicious.
But it is not that bad like the one from wraper.

The Board is rev 005.
On the box was a little yellow sticker.
One sticker at the bottom of the box says: Cmpl Date:20130409
That means my U1272A is a little bit more then two weeks old.

[Edit]
In my U1272A is no tephlon band like in Daves U1272A.
There are no other seal too.
The case is from 10.10 and the battery cover is from 01.13 rev A
The stand is rev A too.
[/Edit]

Sorry, my camera is not that good.

[mike_kawasaki]

I taked apart my new Agilent U1272A multimeter and saw that places where parts were not populated looks weird. Inspected it with microscope and wow, all soldering is pure crap. On some places it looks like that parts will fall off if I touch them. Seems that soldering paste they used was well expired. I can't believe this , shame on you Agilent. All images are here:

Wraper,

Request:  Can you share the serial number? And it would be also wonderful to get your contact information in case an Agilent support engineer needs to talk to you directly.

I work at Agilent and take great pride in our quality.  This looks unacceptable based on your pictures.  I'll commit to forwarding information immediately to our Manufacturing Quality team and have them investigate based on the serial number.

At a minimum - return to Farnell and ask for a replacement unit!  If you have ANY issues let me know immediately.  I can insure that Farnell does a return and Agilent approves return of the unit (so that there is no cost to Farnell).

As a side note:  The "free" multimeter is only a marketing promotion to leverage our strong product supply presence.  It has NOTHING to do with the multimeter quality.  All the multimeters are the standard units.

My apologies for your experience!

Feel free to write me at on the EEVblog with personal note if don't want to publicly share any information.

Mike Kawasaki
T&M Marketing Manager
mike_kawasaki@agilent.com

[wraper]

Hi, Mike

Serial number: MY53010059

Alexander

[mike_kawasaki]

Hi, Mike

Serial number: MY53010059

repair.incolab@inbox.lv
Alexander

Alexander, thanks!  I'm sending to the Division Team right now... they are in Malaysia so already started the weekend, but know for a fact that the Marketing Managers works on the weekend (I used to live in Singapore and we used to talk on Saturday mornings while doing conference calls with the US Friday afternoons -- don't miss that!!).

Hope Latvia is beginning to receive some warmer Springtime weather.  My friends in Amsterdam mentioned there's been some nice weather during last few weeks.
Mike

[wraper]

Thanks, Mike!
Weather here is rather interesting. We had winter until mid April and then bang! something more like summer in one day. And now we have floods. But today weather is more like in autumn.

[mike_kawasaki]

Thanks, Mike!
Weather here is rather interesting. We had winter until mid April and then bang! something more like summer in one day. And now we have floods. But today weather is more like in autumn.

Alexander, glad you got one day of summer.  We are so spoiled with weather here in California.  It has been a dry winter and the weather has been absolutely gorgeous, actually too warm as plants bloomed few weeks early.  Last weekend we hit 32C.  On a personal level (as a golfer and runner) am LOVING the dry and warm weather.  Now if someone could just lower the crazy house prices in Silicon Valley... friend just bought a 900 sq ft home in Palo Alto, had to offer $1.3M (asking was $1.0M) and all-cash to win the deal! 

I'll be working part of this weekend so commit to get an update whenever I hear from the Division.

Mike

[T4P]

Hey mike, when i was putting back my review set U1273A together without applying excessive force one of the screw heads snapped. Seriously 
You think Agilent might take it back?

[paul30003]

I tore down my cheap uni-t and the soldering was much better than this

[ddavidebor]

guys, a lot of this meter have this problem.

i think agilent has REALLY SERIOUS quality control issue on his portable meter, and this is not acceptable in a company that had make some of the best pcb board in his history.

[mike_kawasaki]

Hey mike, when i was putting back my review set U1273A together without applying excessive force one of the screw heads snapped. Seriously 
You think Agilent might take it back?

T4P, when you state it is a "review set" does this mean you received it for free??  If you give me the serial number I can check on the warranty status, or you can enter into the Agilent.com site.  In the past, worked with another customer who received a free demo unit (think it was for Element14 testing) and there was a service issue; we verified that the unit comes with full warranty and we replaced for no cost.

Noticed you're from Singapore.  I spent 4.5 wonderful years in Singapore, just returned to US in Dec 2011.  We'd move back if Agilent would pay for the move again...  I'm especially missing my S$3 Indian lunches, ice kachang and jumping in the pool almost every day of the year.

Mike

[ExtremeXS]

Hi Mike,

I have a serial number very close, also recently purchased from Farnell:

Serial number: MY53010054

I've only had a very quick peek and not unclipped the case fully, but looks like it may also have a solder paste issue as per the pictures.

Cheers,
Mark

[Barny]

My Serial starts with MY531100 which implies that it is build ca. 100000 Parts later.
Is this possible?

[EEVblog]

Mike Kawasaki
T&M Marketing Manager
mike_kawasaki@agilent.com

Thanks Mike. Good to see Agilent responding proactively with the community (yet again)
I wish more companies did this.

[EEVblog]

I have a serial number very close, also recently purchased from Farnell:
Serial number: MY53010054
I've only had a very quick peek and not unclipped the case fully, but looks like it may also have a solder paste issue as per the pictures.

Please post some photos if you can.
It is most likely a batch issue.

I just had an idea...
What if SMD assembly lines also had automated cameras that snapped and archived a photo of every board/panel as it came off the line?
This sort of issue could be pinpointed pretty quickly.

[quarros]

I just had an idea...
What if SMD assembly lines also had automated cameras that snapped and archived a photo of every board/panel as it came off the line?
This sort of issue could be pinpointed pretty quickly.

You would run out of space before you know it, or had many low resolution photos that not necessary pinpoint all of the possible issues...

[EEVblog]

You would run out of space before you know it, or had many low resolution photos that not necessary pinpoint all of the possible issues...

A run of 10,000 panels at say 4MB each image is only 40GB. So you could get 500K panels (say 5 million boards) onto a single 2TB hard drive.
Completely doable.

[ExtremeXS]

Yeah mine (MY53010054) does look the same as Alexanders, I will try and get a photo or two as requested.

Cheers,
Mark

[quarros]

You would run out of space before you know it, or had many low resolution photos that not necessary pinpoint all of the possible issues...

A run of 10,000 panels at say 4MB each image is only 40GB. So you could get 500K panels (say 5 million boards) onto a single 2TB hard drive.
Completely doable.

Hmmmmm It seems I haven't thought it trough.
You're right because 4 meg is enough for most small boards to reach acceptable resolution
for fault finding, and bigger boards for bigger instruments on the other hand wont be manufactured
that many so the individual pictures for those can be bigger...

Sorry I was a little hasty with my statement.

[SeanB]

Even keeping a single actual panel as well is doable, as you just keep the last one out, and then slip it in an envelope with the details ( panel without the value added things like switches, displays and such are not too expensive) and file it. It then allows you to check later if needed, and you just can either make up as a unit later if desperate or just put it through a shredder when you eol the product range and no longer do any support or cal of it. On a batch of 10 000 panels it will be cheaper than the hard drive and server space.

[T4P]

Hey mike, when i was putting back my review set U1273A together without applying excessive force one of the screw heads snapped. Seriously 
You think Agilent might take it back?

T4P, when you state it is a "review set" does this mean you received it for free??  If you give me the serial number I can check on the warranty status, or you can enter into the Agilent.com site.  In the past, worked with another customer who received a free demo unit (think it was for Element14 testing) and there was a service issue; we verified that the unit comes with full warranty and we replaced for no cost.

Noticed you're from Singapore.  I spent 4.5 wonderful years in Singapore, just returned to US in Dec 2011.  We'd move back if Agilent would pay for the move again...  I'm especially missing my S$3 Indian lunches, ice kachang and jumping in the pool almost every day of the year.

Mike

Yeah, i recieved it for a review. MY52380014 is the SN i think. Busy on OCN these days
I read the agilent TOS and it said "user damage" ... i don't really think it was my fault though ... the head just snapped
Dec 2011, hmm. Just about 5months in when i stepped into electronics 
$3 lunches? No more man  Now my lunches even as a student starts from 3$ ...

Even keeping a single actual panel as well is doable, as you just keep the last one out, and then slip it in an envelope with the details ( panel without the value added things like switches, displays and such are not too expensive) and file it. It then allows you to check later if needed, and you just can either make up as a unit later if desperate or just put it through a shredder when you eol the product range and no longer do any support or cal of it. On a batch of 10 000 panels it will be cheaper than the hard drive and server space.

I don't know man. Surely as Agilent you will definitely have some spare servers lying around with storage space in them. This is 2013 ... c'mon ...
Even i use a Xeon as my daily rig ... Scratch it ... he's got a 3930k at half discount, lol.

[peter.mitchell]

$3 lunches? No more man  Now my lunches even as a student starts from 3$ ...

Cmon, chicken rice is perfectly fine for lunch

[Galaxyrise]

Hey mike, when i was putting back my review set U1273A together without applying excessive force one of the screw heads snapped. Seriously 
You think Agilent might take it back?

Sorry if I'm the only one confused, but is this a unit you need to return after the review and you're worried that you'll be charged for the broken screw?

[blackdog]

Hi, mike_kawasaki

This is the serial of my U1272A
I is purchased from Farnell, the Netherlands.



Kind regarts,
Blackdog

[T4P]

Hey mike, when i was putting back my review set U1273A together without applying excessive force one of the screw heads snapped. Seriously 
You think Agilent might take it back?

Sorry if I'm the only one confused, but is this a unit you need to return after the review and you're worried that you'll be charged for the broken screw?

I get to keep the unit.

[logictom]

I have a serial number very close, also recently purchased from Farnell:
Serial number: MY53010054
I've only had a very quick peek and not unclipped the case fully, but looks like it may also have a solder paste issue as per the pictures.

Please post some photos if you can.
It is most likely a batch issue.

I just had an idea...
What if SMD assembly lines also had automated cameras that snapped and archived a photo of every board/panel as it came off the line?
This sort of issue could be pinpointed pretty quickly.

I thought this was common practice? The assembly company we use around the corner from us does this with every board so they can do exactly what you suggested.
I've not been to any other assembly houses so perhaps they are better equipped than most?
You've got me curious now, I have a friend who works in a different assembly house, I'll have to have a chat with him to see what they do there.

[kx5]

Here is mine (Rev 4 board).
Is this acceptable?

[wraper]

Yeah, not the best they could but I think it's at least good enough to not cause any problems. Mine is so bad, that I think parts will fall off if I drop it.

[digsys]

Many of the IC pads are truly woeful ! Almost as if there was no flux component and/or under-temp. I agree it's better, but NOT a
whole lot better. It's possible that this is how the batches came out of production, and the ICs in the last pix were hand soldered
(leaded) to fix up what we see here. Sooo many questions need answering.

[SeanB]

Looks like they were just short on the temperature settings and the dwell times there, it has melted, but has not reflowed properly or fully. Probably an early batch after changing to a new lead free solder paste, or a new line still being fully profiled. The components passed the pull force tests in production, but just do not look good.

[wraper]

I would say that it's just crap solder paste not thermal profile. Bigger parts with larger thermal capacity looks better than smaller ones. So I think It's not solder paste melting problem, just solder paste isn't wetting pads well.

[DonRon]

Hi Folks,

got mine two weeks ago from Conrad in Germany.
Opened it after I have read this thread - looks not as good as barny's but better as kx5's (sorry - I have no camera).
The unit was sent within the power supply carton, orginal box of multimeter was not included -  I think they put it out of the original box , so it fits inside the power supply box.
Date of cal certificate is 10-march-2013.
Serial number MY530600xx.

Greet's
Ronald

[ExtremeXS]

Sorry not very clear photos, could not get the camera to focos any closer....

But the area that is unpopulated looks like the model a couple or digits away from my serial number.

[wraper]

Best way to take good photos of PCBs I found so far is microscope + Nokia N95  . Good camera doesn't work because of big lens. But tiny phone camera with enough megapixels is the best.

[K6TR]

Mike Kawasaki
T&M Marketing Manager
mike_kawasaki@agilent.com

Thanks Mike. Good to see Agilent responding proactively with the community (yet again)
I wish more companies did this.

Dave I've enjoyed your Youtube page for well over two years. The topic in this thread has prompted me to register and comment.

While I appreciate the prompt attention Mike has given to this matter the poor soldering we see in these photos is very troubling ! The reason why it is troubling is this is not the first such instance. If you will remember you found a poorly soldered turret when you reviewed the U1253A. This would indicate that Agilent has a systemic problem maintaining quality control out of the plant that manufactures it's multimeters in Malaysia. You also will remember that your U-1253A kicked up a furror. My personal take on that matter was your review contained a level of thoroughness that demonstrates why they are so popular. In that review I thought some of your criticisms while certainly warranted were excessively harsh. This is just my personal perspective. I think the level of harshness seemed incongruent with the thumbs-up rating you gave to the 1253A the way most people perceived it.

What is before us in this thread is a very serious matter. One that deserves rigorous scrutiny and brutal candor ! Anyone can make a mistake but repeated negligence should raise some serious alarms. Agilent a company world renowned for producing some of the finest Test Equipment ever produced has seriously dropped the ball. They charge Top Dollar for their Equipment and their Multimeters are no exception. Their chief competitor in this market, Fluke, assembles their meters in the US paying US Labor wage rates. Wages which are substantially higher than those in Malaysia. The consumer does not gain anything out of this arrangement as the Agilent Meters are priced identical to the Fluke Meters. The least Agilent could do would be to apply a small percentage of the excess profits they gain from producing in Malaysia to see that the quality of what they produce out of that plant matches the quality of the meters produced in the US by Fluke.

[mike_kawasaki]

While I appreciate the prompt attention Mike has given to this matter the poor soldering we see in these photos is very troubling !

K6TR, I agree it is troubling.  Must admit I'm not a Manufacturing Expert but am working on a Manufacturing Quality Engineer to be named as the owner of this escalation.

Agilent a company world renowned for producing some of the finest Test Equipment ever produced has seriously dropped the ball. They charge Top Dollar for their Equipment and their Multimeters are no exception. Their chief competitor in this market, Fluke, assembles their meters in the US paying US Labor wage rates. Wages which are substantially higher than those in Malaysia. The consumer does not gain anything out of this arrangement as the Agilent Meters are priced identical to the Fluke Meters. The least Agilent could do would be to apply a small percentage of the excess profits they gain from producing in Malaysia to see that the quality of what they produce out of that plant matches the quality of the meters produced in the US by Fluke.

K6TR,
Don't think this is a company secret, but Agilent builds almost all their T&M equipment in Penang, Malaysia.  The manufacturing team is worldclass (IMHO) and we've done thorough benchmarking.  The Agilent T&M quality is #1 in the world when measured against all competitors and their repair records.  We've even documented that the Malaysia team's quality is BETTER than when we manufactured in the US.  (You might imagine this was an internal question the team asked and we've verified.)

However, that doesn't help solve the issue raised in this thread about the U1272A. 

As stated above, the Agilent team is committed to constantly improve our quality as we see it as a competitive differentiator.  This issue is being taken very seriously and I am talking to the Division Management later tonight to get an update; with the first action requested to assign a quality engineer to analyze the issue.

I take great pride in our quality; but we all have issues and what differentiates companies is how quickly they respond and solve problems.  This will be another test for Agilent!

Mike

[eKretz]

I must agree that it seems to be indicative of a recurrent issue if this has happened before with a different product from a different line, but it seems that Agilent is taking the matter quite seriously, as they should. I'm guessing that they will pass the test and improve their QC at the same time. Obviously there is a problem with finished parts inspection as well as process control.

[KJDS]

What has surprised me is that the issues would have been evident to any of the employees assembling the boards into the housings. If those employees areb't empowered to stop the line, call over the line management and get the issues addressed then the facility isn't world class.

I've seen many world class facilities and they've never called themselves world class, merely stated their limitations and shown how everyone is contributing to improvements.

I've also seen a lot of facilities that describe themselves as world class and none of them have been close.

[mike_kawasaki]

What has surprised me is that the issues would have been evident to any of the employees assembling the boards into the housings. If those employees areb't empowered to stop the line, call over the line management and get the issues addressed then the facility isn't world class.

KJDS, touche!  That's a great viewpoint.  We should find these problems earlier in the manufacting process!

(and sorry for mis-typing your user ID last message -- too much multi-tasking).

Mike

[eKretz]

Indeed, someone in management at that facility should be getting a serious workover for this. Especially if this  had been brought to management attention by a line worker. I don't see how it couldn't have been now that I think about it. Most likely is the person who was in charge made the decision to let it pass, hoping it would fly under the radar, and as it seems it always must, that approach backfired.

[mike_kawasaki]

Quick Update... The Division Marketing Manager has confirmed that the Support Team has placed this as a high priority.  They are investigating the issue and expect an update within the next week.

Thanks again for the EEVBlog community for raising the issue. 

Mike

[Corporate666]

What has surprised me is that the issues would have been evident to any of the employees assembling the boards into the housings. If those employees areb't empowered to stop the line, call over the line management and get the issues addressed then the facility isn't world class.

I've seen many world class facilities and they've never called themselves world class, merely stated their limitations and shown how everyone is contributing to improvements.

I've also seen a lot of facilities that describe themselves as world class and none of them have been close.

I agree with you and I think a lot of people are waiting for an answer on this - I *hope* the answer will be an actual explanation and not something like "we've taken steps to resolve this but we can't comment further". 

I doubt there is a company in the world that doesn't experience problems - even those renowned for quality.  Apple had their Wifi antenna issue, Intel has released bad batches of processors, etc.  The main thing is how the company handles an issue when it's been raised... looks like Agilent is doing all the right things so far, and it's great that the hounds have not descended on them

Looking forward to Mike's updates.

[amyk]

What has surprised me is that the issues would have been evident to any of the employees assembling the boards into the housings. If those employees areb't empowered to stop the line, call over the line management and get the issues addressed then the facility isn't world class.

Or maybe the employees couldn't care less/don't know, given that their job isn't to inspect PCBs but to assemble them as fast as possible?

[KJDS]

What has surprised me is that the issues would have been evident to any of the employees assembling the boards into the housings. If those employees areb't empowered to stop the line, call over the line management and get the issues addressed then the facility isn't world class.

Or maybe the employees couldn't care less/don't know, given that their job isn't to inspect PCBs but to assemble them as fast as possible?

In which case, teh facility isn't world class, but a sweatshop.

[mikes]

In which case, teh facility isn't world class, but a sweatshop.

That's only your opinion.

[KJDS]

In which case, teh facility isn't world class, but a sweatshop.

That's only your opinion.

No, it's a fact based on the definition of a world class manufacturing facility.

[EEVblog]

I doubt there is a company in the world that doesn't experience problems - even those renowned for quality.  Apple had their Wifi antenna issue, Intel has released bad batches of processors, etc.  The main thing is how the company handles an issue when it's been raised... looks like Agilent is doing all the right things so far, and it's great that the hounds have not descended on them
Looking forward to Mike's updates.

Agilent have done the right thing before:

[Bored@Work]

What has surprised me is that the issues would have been evident to any of the employees assembling the boards into the housings. If those employees areb't empowered to stop the line, call over the line management and get the issues addressed then the facility isn't world class.

Or maybe the employees couldn't care less/don't know, given that their job isn't to inspect PCBs but to assemble them as fast as possible?

Or the employees have been conditioned to never exercise the option to stop the line. I have seen that once. On paper everything looked fine. Quality awareness programs, empowered employees, and whatnot. But what really happened was that any new hire was told in no uncertain terms by fellow workers to never stop the line. If necessary that message was delivered by dragging the new employee in a dark corner of the parking lot and roughing him up a bit. Bonus payments depended on the line not stopping.

[eKretz]

Yeah, that's kind of what I was thinking when I wrote my last response. I have been on the receiving end of similar B.S. before, it does happen. This problem should have been prevented from the top of management down. Don't forget the facility manager almost certainly gets bigger bonuses than the line employees.

[wraper]

My multimeter started to go bad today. Restarted several times when I switched between ranges. Then display gone blank on mA.A range. Once I pressed on switch It gone back.  When released, it died again, tried it several times, same result. Then I switched it off and now it works normally so far. Not sure if it something with switch or crap soldering already cracked but I'm sure that it's not because I removed back cover. I didn't even touch PCB, just took photos and screwed cover back. When writing this, tried to repeat problem and it died again on mA.A range but just for a moment.

[mikes]

In which case, teh facility isn't world class, but a sweatshop.

That's only your opinion.

No, it's a fact based on the definition of a world class manufacturing facility.

LOL. Where did you find this "definition?" ITYM "KJDS class manufacturing facility." Nowhere does ISO9000/9001 require employees to be skilled at each other's jobs.

[kx5]

Agilent have done the right thing before

They are supposed to do something. I (we) paid a lot of money for a high quality instrument and their support.

[Salas]

Rant mode [ON] The hp duo Bill & Dave were always stressing that their key success was not technical innovation but their groundbreaking management concepts. Staff got inspired, even stated they felt like the bosses were working for them than the other way around. Silicon Valley spurred out of hp. Best growth managers ever proved to be company owning EEs and programmers. Alas, brilliant heritage isn't always cutting it in an increasingly barbaric accountant cult 21st century world. Someone close down the economic and business faculties for a generation please. They churn out too much entropy. Rant mode [OFF].

[Bored@Work]

LOL. Where did you find this "definition?" ITYM "KJDS class manufacturing facility." Nowhere does ISO9000/9001 require employees to be skilled at each other's jobs.

Anyone arguing with ISO9000 when it comes to world class quality automatically loses the argument. ISO9000 is an incredible pile of shit.

[wraper]

Yeah folks, it is becoming even worse. Looks like this multimeter was born to die.

[c4757p]

How sad this thread is. It's starting to look like their definition of "quality control" has become "does it fit in the box?". How in hell was this not caught??

[Lightages]

Anyone arguing with ISO9000 when it comes to world class quality automatically loses the argument. ISO9000 is an incredible pile of shit.

I agree there!

I hope Agilent shows the right stuff here and removes the doubt on their quality. I have been recommending Agilent multmeters to some but am now not until we see what is going on. The list of meters than can be recommended is getting smaller and smaller. Right now I only feel safe recommending Brymens and their rebrands, Fluke, and the Amprobe XR series for high safety requirements. For low safety requirements there are others. Agilent is in limbo in my books right now for either use.

[K6TR]

I doubt there is a company in the world that doesn't experience problems - even those renowned for quality.  Apple had their Wifi antenna issue, Intel has released bad batches of processors, etc.  The main thing is how the company handles an issue when it's been raised... looks like Agilent is doing all the right things so far, and it's great that the hounds have not descended on them
Looking forward to Mike's updates.

Agilent have done the right thing before:

Dave it will be interesting to see how they respond this time. So far Mike Kawasaki's prompt response with alacrity bodes well.

I do have a hunch that this problem may test Agilent in ways we have not seen before.

I say this because previous problems have been of an engineering/development nature where the issue was lack of thoroughness. This problem appears to be of a manufacturing nature. Rarely do these types of problems stem from a lackadaisical or uncaring assembly line employee. But from a Mid Level manager who exerts undo pressure on the line to complete a production run in order to satisfy a sagging shipment schedule.

Yes Agilent have responded proactively in the past but if you will remember Dave in one of your reviews of an Agilent multimeter you found a poorly soldered turret. This latest incident would seem to indicate a trend or systemic problem.

Again this is only conjecture on my part and I keep an open mind while waiting an Official Explanation


Bob

[EEVblog]

Or the employees have been conditioned to never exercise the option to stop the line. I have seen that once.

Yep, line stoppages are a huge deal. You don't want to be the low level worker that cries wolf.
Sounds stupid of course, until you've worked in a production factory and see how they operate and are managed.
The buck stops with Agilent though. Even more so given that this is likely Agilents own factory in Malaysia.

[EEVblog]

Yes Agilent have responded proactively in the past but if you will remember Dave in one of your reviews of an Agilent multimeter you found a poorly soldered turret.

That was not really a poorly soldered turrent. It just looked poorly soldered. Quite a common look for those style of turrents with lead free solder.

[eKretz]

Or the employees have been conditioned to never exercise the option to stop the line. I have seen that once.

Yep, line stoppages are a huge deal. You don't want to be the low level worker that cries wolf.
Sounds stupid of course, until you've worked in a production factory and see how they operate and are managed.
The buck stops with Agilent though. Even more so given that this is likely Agilents own factory in Malaysia.

Even so, the blame should never be placed on the employee. The way the process should go is: Line Employee notices problem, reports to Line Supervisor. Line Supervisor investigates, determines whether it is necessary to report to QC or if the problem can be immediately rectified and whether it needs to go any further than that. If it needs to go further than that, then QC is notified to check and pull defective units. The higher-ups are the ones that should make the decision to stop the line in such a case. Line-level employees should only be stopping the line in an emergency situation, I.E. for safety concerns or if the units are missing parts and non-functional.

[c4757p]

Even so, the blame should never be placed on the employee.

Any manager who even thinks about discouraging an employee from reporting something like this should be fired so fast he'll wonder where the assembly line went.

[ExtremeXS]

Hey Wrapper,

Jeez your unit's really not happy.  I can get mine in V/mV modes to fluctuate values by applying a little pressure on either side  the to 'bar' of the rotary dial too, that does not seem to be something that should happen.

Can you let me know what you decide to do with yours and how it goes?  I have been in touch with Farnell but got a reply that didn't quite clear up my questions just yet (I need to go back from slightly clearer detail).

What I want to avoid is that whole send it in and the same unit just gets returned as 'no fault found' had that kind of issue before on other devices where the reported issue just has not been checked.

Cheers,
Mark

[wraper]

Hi ExtremeXS,
I will not return it now, will wait some time. I have returned my bench multimeter for warranty repair a week ago. And that doesn't appear to be fast. If i will return this, I would be left only with some crap multimeters which I don't trust. Moreover there is very big possibility that Farnell will replace it with same one from that crappy batch. Actually I'm starting to think that I have some kind of curse on me (of course I don't trust such things  ). When I buy new equipment somehow all crap goes to me. And what you think, this multimeter works normally again.

[Bored@Work]

Or the employees have been conditioned to never exercise the option to stop the line. I have seen that once.

Yep, line stoppages are a huge deal. You don't want to be the low level worker that cries wolf.
Sounds stupid of course, until you've worked in a production factory and see how they operate and are managed.
The buck stops with Agilent though. Even more so given that this is likely Agilents own factory in Malaysia.

Even so, the blame should never be placed on the employee. The way the process should go is: Line Employee notices problem, reports to Line Supervisor. Line Supervisor investigates, determines whether it is necessary to report to QC or if the problem can be immediately rectified and whether it needs to go any further than that. If it needs to go further than that, then QC is notified to check and pull defective units. The higher-ups are the ones that should make the decision to stop the line in such a case. Line-level employees should only be stopping the line in an emergency situation, I.E. for safety concerns or if the units are missing parts and non-functional.

If I remember my quality lessons correctly, it was Toyota who came up with the original. In their rise from a small local manufacturer to a large international player they invented and implemented the Toyota Production System (TPS). Part of TPS was (and likely still is), that every worker could stop the line immediately, should he or she detect a problem. They installed something like a pull string in parallel with the line. Similar to those strings used in the old days in buses to signal the driver that he should stop. You pull, the line stops.

No multi-step process involving the line supervisor, QC or some higher-ups before the line is stopped. The decision to stop the line was moved to the lowest level. We are talking about a country strongly believing in hierarchies, so implementing that must have been a shocking thing.

The good thing is that if the line is first stopped, there is less incentive for higher-ups to look the other way to avoid having to stop the line. It is already stopped. The bad thing is it takes bigger balls for a worker to flag a problem.

[retiredcaps]

I will not return it now, will wait some time.

I used to work for a vendor and have been through this crisis management before. I'm not affliated with Agilent in any way, but if I were an empowered Agilent employee reading this, I would

a) ship you a new thoroughly inspected and tested replacement multimeter at no cost to you
b) ask for the faulty multimeter to be sent (with prepaid box and tracking number) to the engineer responsible/tasked to investigate this
c) maybe throw in some extras like extra probes, extra fuses, blue tooth, etc
d) post regular updates back on this forum

[eKretz]

Even so, the blame should never be placed on the employee. The way the process should go is: Line Employee notices problem, reports to Line Supervisor. Line Supervisor investigates, determines whether it is necessary to report to QC or if the problem can be immediately rectified and whether it needs to go any further than that. If it needs to go further than that, then QC is notified to check and pull defective units. The higher-ups are the ones that should make the decision to stop the line in such a case. Line-level employees should only be stopping the line in an emergency situation, I.E. for safety concerns or if the units are missing parts and non-functional.

If I remember my quality lessons correctly, it was Toyota who came up with the original. In their rise from a small local manufacturer to a large international player they invented and implemented the Toyota Production System (TPS). Part of TPS was (and likely still is), that every worker could stop the line immediately, should he or she detect a problem. They installed something like a pull string in parallel with the line. Similar to those strings used in the old days in buses to signal the driver that he should stop. You pull, the line stops.

No multi-step process involving the line supervisor, QC or some higher-ups before the line is stopped. The decision to stop the line was moved to the lowest level. We are talking about a country strongly believing in hierarchies, so implementing that must have been a shocking thing.

The good thing is that if the line is first stopped, there is less incentive for higher-ups to look the other way to avoid having to stop the line. It is already stopped. The bad thing is it takes bigger balls for a worker to flag a problem.

I am pretty sure you do remember correctly. My point was more that an issue such as the one under discussion may have been beyond the line workers' skill and/or knowledge level to determine whether there was a serious problem, so in such a case, the correct thing to do would have been to bring it to someone's attention who could make that determination, and then take the necessary action. However, you have a valid point there as well about the incentive.

[casper.bang]

Just received my U1272A, and I am pleased to see that the board (rev 5, like Barny's ealier in this thread) looks fine, certainly nothing like the OP's pictures.

Very nice meter I must say. Only issue I find is that the V/mV range is really susceptible to interference when I i.e. put my fingers near the front panel buttons, but I guess this is normal and perhaps a property of high input impedance.

[lowimpedance]

Had a look inside my U1272a and the soldering seems ok. PCB is rev.4..
How ever when I tried to ut in back together I found the two tabs at the top would no longer engage which would allow the two case halves to be separated quite easliy.
No amount of pressure would help to re-engage these tabs. 
Why the hell did they use these crappy small tabs instead of another two srcews !!.
I would suspect these would be a weak point if the meter was dropped on this end, if they were to break the meter would be essentially unsafe as the case could then be able to be separated without much effort.
 The tilt bail latch is another very crappy part with little or no positive retention feel. (the first unit I purchased this bail did not even latch in but flopped freely!).

Some pictures of the pcb (sorry about quality), and the body latches for the top part of the case.

[mike_kawasaki]

Hello EEVBlog Forum,

Thanks again for raising this issue with Agilent.  I have an update and action plan.

1.   Confirmed problem with Wraper’s U1272A.  You’re probably saying ‘duh’, but I wanted the Division’s Production Team to review the pictures and confirm this does not meet our manufacturing standards.
2.   Confirmed this U1272A PC board should not have passed our inspection standards.  The Production Team is reviewing their process and insuring all production workers are retrained on finding this problem.
3.   Confirmed that the Division’s Product Team reviewed every board currently within their production environment – production units sampled in the factory do not have this issue.
4.   Confirmed that we have NOT seen an abnormal number of failures in this serial number range

Wraper, because of your issue want to guarantee a replacement unit.

1.   Wraper, please send me an email (through EEVblog or my work email) with your contact information – full name, complete shipping address and phone number?
2.   The Division will send you a new U1272A immediately upon receiving your shipping information.
3.   The Division will also include a U1171A IR-USB cable, as a thank you for raising the issue.
4.   The European Customer Center will contact you and help return the unit to Agilent (no cost to you).

The division will have the unit shipped back to Penang so the Production Team can analyze the board and propose any new production process improvements.

As always, Agilent will stand behind their products.  I hope nobody else has an issue with our handheld DMMs, but if you experience a problem we will replace any faulty unit during the warranty period. 

As a reminder, all Agilent products sold after 1 March 2013 come standard with a 3 year warranty.

Thanks again to everyone for their test equipment passion and loyalty to Agilent equipment.

Mike

p.s.  Special thanks to Carol, the Division Support Manager, who took quick and decisive action to investigate everything in the last few days.

[wraper]

How ever when I tried to ut in back together I found the two tabs at the top would no longer engage which would allow the two case halves to be separated quite easliy.
No amount of pressure would help to re-engage these tabs. 

Did you tried to close the top side with hooks first in the way that hooks catches and only then closing bottom side? Hooks don't seem to be damaged on photos.

Thanks to Mike for hiss effort. I wrote message to him.

[lowimpedance]

Oh yes, tried all permutations and  the problem was the slot could not just quite make it over the tab when pushed together to clip in.
Had to gently heat the plastic with a small Weller heat pencil and expand the slots downwards a little and now they both clip soundly in.
Clearly there moulding tolerences are a bit slack. In my meters case it must have been a very tenuous the amount of engagment. (personally I think they should have used screws instead).
My first U1272a I purchased the tilt bail would not clip into the body, ie missed the VERY small dimple which it was supposed to engage on when closed!. I returned it for a replacement and it was only marginally better. Did the heat pencil trick and improved the retention to an acceptable amount.
 These are clearly not very good aspects to the case design. On other models there seems to be much more attention to the case design. (I have both a U1242b and 1252a).

@wraper , I guess your meter case clipped back together no problems? hows the tilt bail retention, bit wishy washy or firm when clipped in?.

edit: added some photo's, (for anyone considering this meter), of the tilt bail locking clip. Also compared to the U1242b which has clips on both sides and has a much more positive lock in than the 1272a.

I should also indicate that I am happy enough with the general quality of the 1272a and I can live with the minor points concerning the case design.

[wraper]

My case is completely solid, no problems with that. Tilt bail is OK, however I didn't use it much so don't know how rugged it is. It's almost new meter.

[ExtremeXS]

Hi Mike,

Nice to hear up the update fro Wrapper.  What is the process for others who appear to have the same issue?

Thanks,
Mark

[mike_kawasaki]

Hi Mike,

Nice to hear up the update fro Wrapper.  What is the process for others who appear to have the same issue?

Thanks,
Mark

ExtremeXS,

Hope all is well with you.  Happy to answer your question.

Bottom Line is that we haven't seen any units on the EEVBlog Forum that are significantly 'unacceptable', except Wraper's unit.  This is why we're taking the special process to insure the unit gets into the Product Manager's hands.  Wraper's unit will go through a special process - but expect all other repair/service issues to go through our normal process.

The other issues are based on a case-by-case basis.  The units have a warranty and any failures will be honored by Agilent.  We have a service process and the team will verify if there's a failure.  If a failure is confirmed, then a new unit will be exchanged for your failed unit.

You can simply click on Agilent.com's 'contact an expert' link from the U1272A product page, and will give your local team's phone number and email.  (and interestingly, I sit next to our Worldwide Service & Support Manager so happy to assist if any troubles in contacting the right people).

As noted in a previous note, want to document the warranty -- it has changed to a 3 year warranty after 1 March 2013.  This is worldwide; and for all sales channels/distributors.

Mike

[wraper]

Actually ExtremeXS multimeter board is similar to mine. He posted a picture in post #61 but it isn't with such magnification as mine. But I can clearly see the same soldering defects on bigger pads, especially where connector is not populated. I also see suspicious soldering on SMT parts but unfortunately picture isn't quiet clear. And his serial number is different only by last digit. His multimeter is also from Farnell.
ExtremeXS please make better pictures if you can. I also wrote about microscope and mobile phone camera. If you can access a microscope somewhere, then you can make quiet good photos. Or you can try some magnification lens and small camera combinations, maybe it works.

[ExtremeXS]

He Wrapper and Mike,

Yeah as Wrapper says mine looks the same just not very clear photos, I will try again I did use a mobile to take these as my camera would not focus at all - I do not have a microscope to try to get a better magnification (I can try and get some brighter lighting maybe).

There are some components (again not clear in the photos) that are not sitting straight as the reflow did not work and thus pull them into position.

I do understand Mike that you are saying should it fail it can be resolved at anytime in the next 3 years, but if there is a manufacturing defect on a batch then the affected devices are all likely to have the same eventual failure I would suspect.

Cheers,
Mark

[whwong]

I am the product marketing egineer from Agilent managing handheld products, we acknowledge the concern from all gentlemen here. If any of you have the poor soldering issue on PCB, please contact us directly and we shall get Agilent local service center to look into your case. We are certainily want to collect back all reported units and we will do further investigation to root cause the issue.

[retiredcaps]

And his serial number is different only by last digit.

The above alone should be enough to get ExtremeXS's multimeter back to the Agilent (same team that will investigate).  It is reasaonble to assume that the same poor soldering job would exist in the very next unit made.

How often do you come across sequential serial numbers from two different users?

[ddavidebor]

Maiby dave can put a advise somewhere here in the forum?

[tomi]

Bottom Line is that we haven't seen any units on the EEVBlog Forum that are significantly 'unacceptable', except Wraper's unit.  This is why we're taking the special process to insure the unit gets into the Product Manager's hands.  Wraper's unit will go through a special process - but expect all other repair/service issues to go through our normal process.

Hi everybody,

I'm another not-so-happy owner of an U1272A (serial MY52520012) which seems to come from the same batch as the 'only significantly unacceptable' one from wraper. I've sent photos and serial to mike_kawasaki soon after this thread started in case they needed more info to investigate the issue and I didn't want to make more fuss about it.

The photos of its internals are here:

https://dl.dropboxusercontent.com/u/82714359/bad_1272A_1.jpg

https://dl.dropboxusercontent.com/u/82714359/bad_1272A_2.jpg

Looks crappy, but it worked. Then I tried to gently pick a diode with SMD tweezers and it lifted straight off the board (there's an empty D10 footprint in the first photo). Just held it, no force was used.

Come on, Agilent. This is supposed to be a precise measurement instrument. It's riddled with cold solder joints and I'm supposed to trust it to be within 0.05% for a full year after calibration, since the certificate says so. And in the mean time the components will rattle off the board after I put it firmly on the table. And now I'm supposed to sort this out with the local distributor, who probably don't have a clue and will presumably conclude that it's my fault, since it's not allowed to poke inside it anyways...

Sorry for the sarcasm, but I'm really not in a good mood today.

Tomi.

[mikes]

Get real. Agilent has a warranty process in place. Use it. It's much more efficient for everyone, and processes ensure that things get properly documented.

Because the first person to report the issue got a bit of special handholding is no reason for anyone else to expect the same. You have a statement from an Agilent employee that bad soldering at the level of Wraper's is a defect. If your's is similar, take a picture, write a note explaining the problem, put it in the box with your meter, and go through the warranty process.

And if you broke your meter by opening it, then poking and prodding, consider yourself lucky if they do honor the warranty. Because it was in fact you who broke it.

[wraper]

tomi, I checked your multimeter warranty status on Agilent website. And I think it may be from different batch because warranty start date is 2 weeks earlier. Mine and ExtremeXS warranty start date is the same. Where you bought yours?

[EEVblog]

I have a u1273ax (OLED version) in the lab that I haven't opened yet. Perhaps worth a look also...

[tomi]

Hi wraper,

bought mine from Farnell UK in the last week of March in a bundle with a bench PSU. Same as you, just a little earlier.

mikes,
the reason for my grumpiness wasn't that I broke the multimeter. Yes, technically it was me who picked the diode off the board, but I find it highly unusual that this could be done at room temperature. And yes, both the diode and the PCB are just fine, just separated. If Agilent decide to refuse the warranty claim, so be it. I'll live.

The issue I have is not that I expect some special handholding, but I was expecting a little more active involvement from them (RMA?, recall?). Instead, they seem to have decided to wait for warranty claims (does this mean with the local distrubutor or shipping it back to UK?) and see, how many come back.

But I broke mine anyways...

[mikes]

It's way early for them to do any sort of recall. There's simply no reason for them to go through some big, expensive recall before they know the facts fully. Agilent makes good stuff, at a premium price. I'd rather they not simply do a blind recall which is likely to cover a lot of perfectly good items, and then have to raise their prices further to pay for that unnecessary action.

They've been extremely candid here - they've admitted there's a problem, and they're working to resolve it. Let them do their work. This may be limited to a particular shift, a particular PCB panel, a particular soldering line. Give them time to determine what they need to about the cause - what additional info they need, figure out how widespread the problem might be, etc., and create an appropriate response. Everything I've seen here indicates they're taking this seriously and moving forward at a perfectly reasonable pace. If they feel they need to actually see more samples, I'm sure they'll get in contact with people who've said they have the problem.

Until then, there's a warranty available should anyone know that theirs has bad soldering, or if they should have an actual functional problem, bad soldering or not.

[MacAttak]

Glad I saw this - I was considering getting an Agilent as my third DMM, but I think I'll wait this out first. I'm on vacation for a few weeks anyways, so it's not like I need it right away.

It is encouraging to see the vendor taking it seriously and making firm steps to make things right. That's exactly why I don't mind paying a few extra dollars for gear sometimes, if I feel like I can trust the vendor to back up their product with good service.

[digsys]

... Yes, technically it was me who picked the diode off the board, but I find it highly unusual that this could be done at room temperature. And yes, both the diode and the PCB are just fine, just separated. If Agilent decide to refuse the warranty claim, so be it. I'll live.

I've been in this boat a few times. I have never been knocked back when it was an obvious design / manufacture fault. In OZ, at least,
you are able to open / repair stuff and STILL RETAIN warranty, as long AS you don't CAUSE the fault. Warranty is per base rights per
country ! NOT the 1 year / limited crap they tell you. AND I have won MANY claims for friends on failures after the "listed" warranty period!
There's NOTHING to say that the part/s wouldn't have dropped off at some point anyway. A jolt here, a knock there.
If I were you, put the loose part back in the meter, stick a note referring this Forum thread, and don't back down (IF it is rejected).
I get fed up with companies pushing crap (NOT saying this is the case here), and then trying to hide behind their bullshit terms.

But I broke mine anyways... 

Not really. You don't have to defend yourself, you did the right thing.

[K6TR]

I have a u1273ax (OLED version) in the lab that I haven't opened yet. Perhaps worth a look also...

Dave an episode like the one illustrated in this thread drives home the point to everyone to take your adage seriously "Don't turn it on.....Take it apart !"

Once people have their sense of confidence in the product restored then things lighten up.

[Baliszoft]

Took my U1252B apart, looked fine.

[wraper]

Just couldn't resist to take a photo of a solder joint of the same diode which Tomi took off from his board. Well, there is no solder joint...

[SeanB]

That is some poor soldering there. Probably the only thing holding the components on is the remains of the flux used. Pads are very oxidised as well, and probably this is the underlying reason, the boards were left exposed to humid air for a period and corroded.

[wraper]

That is not oxidation, that's the only places where solder catches the pads. Places which are smooth are just clean from solder. I couldn't make a photo as good as I see it with my eyes but I think it's detailed enough to see what's going on.

[TerminalJack505]

Hopefully I don't come across as being snarky here but... 

Damn!  Those are some bad boards.  I've seen better boards in videos where junky consumer electronics are taken apart just for a laugh about how bad they're assembled.

Take apart a $20, no-name multimeter and I can pretty much guarantee you it will be better assembled than those boards.

[marshallh]

That last pic looks like ENIG black pad, unless you caught it at the wrong angle. Regardless the paste didnt reflow worth squat.

Major process fail.

[wraper]

ENIG looks ok, that pad isn't black. That looks like this because photo is made from big angle to show soldering from the side.

[zaoka]

They are ignoring bugs with U3401 and U3402 and are not providing a solution (they did not even notify their customers that a problem exist), I am sure they would do the same with U1272A.

 

[Corporate666]

Hi Mike,

Nice to hear up the update fro Wrapper.  What is the process for others who appear to have the same issue?

Thanks,
Mark

ExtremeXS,

Hope all is well with you.  Happy to answer your question.

Bottom Line is that we haven't seen any units on the EEVBlog Forum that are significantly 'unacceptable', except Wraper's unit.  This is why we're taking the special process to insure the unit gets into the Product Manager's hands.  Wraper's unit will go through a special process - but expect all other repair/service issues to go through our normal process.

The other issues are based on a case-by-case basis.  The units have a warranty and any failures will be honored by Agilent.  We have a service process and the team will verify if there's a failure.  If a failure is confirmed, then a new unit will be exchanged for your failed unit.

You can simply click on Agilent.com's 'contact an expert' link from the U1272A product page, and will give your local team's phone number and email.  (and interestingly, I sit next to our Worldwide Service & Support Manager so happy to assist if any troubles in contacting the right people).

As noted in a previous note, want to document the warranty -- it has changed to a 3 year warranty after 1 March 2013.  This is worldwide; and for all sales channels/distributors.

Mike

Can you please comment on the one with the diode that came off.  Per Agilent's standards, does that qualify as "significantly unacceptable"?

If only the one was significantly unacceptable, does that mean that the other pics posted meet Agilent QC standards?  Or do they not? 

[mike_kawasaki]

Can you please comment on the one with the diode that came off.  Per Agilent's standards, does that qualify as "significantly unacceptable"?

If only the one was significantly unacceptable, does that mean that the other pics posted meet Agilent QC standards?  Or do they not?

I can't comment on every case, but let me reiterate and restate the Division Support Engineer's statement in a previous post:  Agilent Will Stand Behind These Products And Replace Anything That Is Defective.

Many of you have been great fans and loyal buyers of HP/Agilent equipment.  We don't want to risk our relationship with any of you!  A failed units (including the diode that came off) will be exchanged for new units.  As noted earlier, it is required to work through the Support Center as they manage all the local return/exchange processes and can manage the regional or country logistics.

I've talked to the Division Support Manager live.  She will insure all units are replaced if you deem them defective.

The team has not found a systemic problem based on their manufacturing records; however, that is why they urgently want to return some units to the Factory (as you might imagine, not all returned units actually make it back to the factory because the logistics costs are not warranted).  We have policies around when to do a recall but think it is too early; since the division is only going from a few pictures.  Agilent commits to fully analyze all returned units and investigate any faulty manufacturer process and decide if a certain 'batch' was effected.

Now I'm going back to enjoying my weekend with the family  .  Hope everyone had a good weekend in their part of the world!

Mike

[ivan747]

I think we all appreciate you took time from your weekend to contribute to this thread. Personally I know it can be a problem sometimes.

[retiredcaps]

Hmm, 3 weeks since the last public update here on the forum.

When I dealt with crisis management, a "regular" update meant daily or weekly information.

Even a note saying "we are still looking into it" every monday would suffice.

[wraper]

Multimeter itself arrived 5 days ago but I said courier to leave it at my work. Last week I was at some fascinating place called Prokopyevsk (Siberia, Russia) . Was repairing some lab equipment there. Returned home 2 hours ago. Picture from there - 284V  in mains socket instead of 220V nominal in Russia. Some good stress test to equipment, hell yeah. I'll write about new multimeter shortly. I already took some photos. While soldering on PCB is good, not all things are so nice about it.

[wraper]

I added the pictures of new multimeter in comparison with bad one. As you can see there is HUGE difference in soldering quality. Both boards are REV005. Agilent also added bodge wire in parallel to Amps terminal probe sensing trace. Now about bad things. I noticed that one inductor has cracks on both sides above terminals. Also I noticed that buzzer sounds an order of magnitude quieter than on old one. Default buzzer frequency setting is 3491Hz, on that old multimeter is very loud, however on other frequencies it sounds much quieter and similar to new one. New one has almost the same loudness on all frequencies.

[wraper]

Then I started to figure out why new is not loud as old one. I hooked buzzer from old to new and bingo, it became as loud as old. I also noticed that in the old one buzzer is soldered nice and clear but in the new one soldering is dodgy. I hooked them to signal gen and discovered that the old one has most loudness at 3.50 kHz, so default frequency setting of 3491 Hz is not coincidence. However new one hasn't resonance on that frequency and is most loud on 3 kHz but not as loud as the old one at 3.50 kHz. I even measured capacitance of both and they are almost the same. Then I decided to make that dodgy solder joint normal, it just pissed off my eyes. And I found why that solder joint looked dodgy and much bigger than on the old one. There is a hole in metallization which was hidden beneath solder. Looks like wire was broken off with piece of metallization and worker decided to hide it beneath solder. That's describes why this buzzer has not resonance on right frequency. Buzzer is not big deal and I can live with it but what the hell Agilent, what's happening. Of course it's possible that metallization came off while I was soldering. But that would be too easy, I even didn't move that wire because it goes under the shield. Anyway that buzzer wasn't as good as another one.

[tomi]

I had my multimeter packed and about to be shipped back to Farnell UK (the local distributor said they couldn't bother since they didn't sell it), when Agilent contacted me and stated they would arrange the exchange directly.

They really did and the new meter arrived on Wednesday. The DHL courier was supposed to pick the old one for return, but said he didn't know anything about it, so it's still lying around. So I'm left with a complete set of functioning spare parts... 

Yes, of course I checked: the soldering in this one is ok. I was just about to write a nice post praising their responsiveness and support - after all it wouldn't be right to raise a stink here first and then keep mum when the problem was resolved. And then I put in the batteries and turned the thing on. No battery and APO (auto power off) indicators on the LCD. They could be brought to life by gently twisting the multimeter case and disappeared again when it was put back on the table.

The plastic display frame was screwed in loosely and had some play. And there was something that looked like flux residue on the gold-plated contacts that mate with the silicone zebra rubber. After cleaning the contacts and screwing back the frame everything works fine. I really didn't think this would merit another round trip via DHL to Malaysia.

So, Agilent kept their promise to replace the faulty unit, the new one works after being slightly repaired In contrast to wraper's, the buzzer on this one is ear-piercing loud. Occasional quality hiccup aside  I like the U1272A and have actually ordered another one at work in the mean time. I haven't received it yet, but sure hope the third time's a charm. Or am I nuts?

 

[Lightages]

Wow. I am not impressed. They had the chance to make this right and the two examples we have here are bad too?

Well I am glad I did not get an Agilent now. This is getting crazy. You can't count on any company anymore.

Well at least we still have Brymen......

[Hydrawerk]

  I hope that this was only a bad luck or something.  Most Agilent stuff should be OK. 

[wraper]

Shipped the old one to Agilent. They arranged DHL pick up.

[Corporate666]

I added the pictures of new multimeter in comparison with bad one. As you can see there is HUGE difference in soldering quality. Both boards are REV005. Agilent also added bodge wire in parallel to Amps terminal probe sensing trace. Now about bad things. I noticed that one inductor has cracks on both sides above terminals. Also I noticed that buzzer sounds an order of magnitude quieter than on old one. Default buzzer frequency setting is 3491Hz, on that old multimeter is very loud, however on other frequencies it sounds much quieter and similar to new one. New one has almost the same loudness on all frequencies.

Wow!

Will be very interested to hear Agilent's response to this.

Honestly, this is what QC is all about.  Every meter should be the same in form and in function.  A buzzer not working right should not make it out of QC.  Neither should the cracked inductor, but I can see that being harder to detect.

[mikes]

Ouch. Seems like Agilent is following HP downhill.

[mzacharias]

Up until now I haven't felt the need to open my U1272A since I had seen plenty of pics here and elsewhere - I thought it would be better to leave it "pristine" inside.
That certainly changed after seeing this thread.

I now plan to open my U1272A for inspection, and my U1253B as well.

Edit -

U1272A  s/n MY52120110
U1253B  s/n MY51250065

Both OK, just the usual RoHS dullness, a few connections maybe "look" funny but aren't.

Turrets on the 1253B are good.

[true]

After reading this I also checked my U1272A, s/n starting MY5231 - solder itself looks like crap but that's just the solder; the soldering is fine. Also get the voltage fluctuations if my hand is near the buttons. Meter works great, thankfully =)

[whwong]

Hello,

We have received the poor solder quality boards and have performed the failure analysis. It is below our solder quality expectations, which prompted us to pull out samples from boards manufactured within the same date code. We do not detect failure in those samples and we suspect the returned boards poor soldering  are most likely related to board contamination.

We have implemented more stringent board cleaning process as well as a more higher magnification end of line inspection.

We apologize for the inconvenience cause and appreciate the time you have taken to highlight this issues to help us further improve our quality.

Regards,
Agilent Technologies

[Lightages]

Well I for one am impressed. It is not usual for any company to state what the exact problem was/is for a process nor the possible fix, especially publicly.

Thank you for being open and honest.

[retiredcaps]

We have received the poor solder quality boards and have performed the failure analysis.

Your response on what steps going forward is exactly what I expected to be shared publicly.

Thank you.

[mike_kawasaki]

We have received the poor solder quality boards and have performed the failure analysis.

Your response on what steps going forward is exactly what I expected to be shared publicly.

Thank you.

Thanks again to the EEVBlog Group.  You highlighted an issue.  The beauty of this 'social media' is that we heard about the problem very quickly, while also understanding that it affected multiple units. 

Glad to hear the Division was able to pinpoint the issue and have updated their process.

Mike Kawasaki

[Hydrawerk]

Owon (with their noisy scopes) should learn from Agilent. 

[sparkybg]

U1272A, serial MY53050166: No problem with soldering and no cracks on the inductor:

[Psi]

Damn, if i was sorting out replacements to send to customers in this kind of public situation you can bet i would take the replacement apart and confirm it is 100% perfect before sending. I wouldn't be grabbing one off the shelf and putting it straight into a DHL box.

 Otherwise Murphy's law says you're going to pick the only other bad unit on the shelf .