Keithley 2001 Meter went for Cal working, Arrived Dead

[smgvbest]

Hi
I put this here because I did repair this meter and the issue was the memory needed to be reseated.
I ran it for 2 weeks staight plus on/off test and never a failure.

So, I just sent my Keithley 2001 in for a cal to Top Dog Test and it left here fully working. Passed all Self tests and actually measured DC/AC/Resistance all as good as my Keysight 34465, Rigol 3068 and HP 34401. I wanted this to be my reference meter hence the cal.   
 
I just got notified it arrive dead, won't read, fails all self tests.    they of course wanted to know if i wanted it repaired which i said no, ship it back

Has any one ran into a problem like this send a meter out for cal.
I packed it well,  actually in a BK Precision box with foam that fit it perfectly so I'm not sure what would have happened shipping wise.
they didn't mention shipping damage

This is not saying anything about this company I just am wondering if anyone has had issues like this

[coromonadalix]

The box must have been dropped flat or badly shaken in transportation, even if the box is not deformed.

If you knew how sometimes they work, or pack the van trailers,  i've seen some carriers being very "non careful" with packages ...

In the past has a spare time i have worked at two carriers,  boy i've seen bad things happen, even an tv box with fork lift holes on the side

[Electro Detective]

If they didn't mention shipping damage... and that's not hard to spot even if externally the package looks ok,
the Cal place may either be trying to drum up some extra work ('jobs for the boys' )
and or someone dropped or zapped it on the bench.

Either way, get it back, open er up and see what the go is, before sending it out for Round 11   

[smgvbest]

I did ask them about shipping damage and they said there was none.

I am afraid of them trying to drum up repair work.
The Orginal Error was a Comm Error message that was it.   that was the memory chip needing re-seating was all.  it worked great since.

I also asked if they are sure it was my meter,  I gave them my SN and asked to verify.

But yeh,  I already said to send it back.

when I get it back and if its been zapped.  that could only happen on their bench.   it went from reading a 2v cal voltage to the box so working  there is no way it was zapped in transit that I know of.

[bitseeker]

Curious to find out what happened as well. I've seen their name around a fair bit, but don't have any experience with them to confirm or deny anything.

It's possible that something just came loose inside. Hopefully, that's all it is.

[TiN]

Comms error means brains can't talk to legs.

Sometimes, esp on old meters the cable between boards get dodgy, resulting in your comms errors.
Other times DC/AC/DC converter (ERG potted brick or little PCBA on digital board) that powers VFD glass crap out, making CPU to go sick and resulting same comm errors, that happen to units I serviced (more than one) .
Or ADC board went loose, that will result comm error. Or the TLP opto on analog board went poo, also comm error.
Or the analog board caps leaked their goo, and ruined everything. 

2001 is meter with lots of fun to have...

[bitseeker]

TiN knows. He's Dr. DMM. Brings 'em back from the dead, buried, and decaying, even.

[Zucca]

smgvbest only when you will open the box you will know.
The rest is just speculation.

take pictures of everything, package close, open, package open 10% or do an unboxing video even better.

Keep us posted.

[SoundTech-LG]

Purchased an Agilent/Keysight counter from them 53230A  and it was not cheap...  also a replacement VFD for HP E3631A Power Supply. It will be good to know what you find. Good for them to know they are under scrutiny here.

[smgvbest]

Ok, here's the update.

I called up and spoke to Van today.   On the Display it has a Overload showing,  it would respond over GPIB but otherwise dead.   he believed it had taken a hit on the front end either voltage or current and blew the front end.

I asked how a meter that left working could take such a hit in the shipping box and said he could not answer that question.
He did tell me it was package very well and there was no sign of any shipping Damage.

This lead me to a new question.   Is it possible that it took a bad bump and something got knocked loose on the inside?
He did not think this was possible because of how it was packaged.

I ask if he could check and he asked if I was giving permission to proceed and open the meter up.   I said yes to take a look.

A hour or so later I got a email with the attached photo

Well, short story (yes, now I do the short story)
it did take a bad bump in shipping and the ADC board got knocked loose

He replaced it and it now powers up fine and they are letting stay powered to to begin the cal procedure on it. 
He's going to secure it for return shipping.

I asked for pre and post call data for it as I want to see how I did on it's comprehensive cal
I believe the total cost is 125 for cal+post data + 25 for pre data so 150 total.

If anyone is interested when I get the Keithley back I'll share the pre/post data here

[bitseeker]

Thanks for the update, Sandra. By "replaced it" do you mean he reseated the original board or actually replaced it with another one?

I've seen lots of things in PCs and servers come loose during shipping. It can be quite an ordeal for equipment.

[smgvbest]

Yes, reseated it,  much better choice of words

[Specmaster]

I have a HP3478A that needed the memory retaining battery replaced, during which it lost the cal data. It did work OK but it kept flashing "CAL" as it could not be relied upon to be 100% accurate. I sent it away to RS calibration services on what was supposed to be a 5 day door to door service for recalibration. After many phone calls and emails after the 5 days had elapsed and no sign of the meter, I was told that it would not calibrate, it kept crashing a wiping out any calibration they had managed to do.

I gave them the links to the service manual and also to the software data that they needed for their Fluke calibration tool.

In the end I had a long telephone chat with the engineer doing the calibration and he told me the meter was beyond economic repair but if I liked to select another from their catalogue, they would give a free calibration with it? I though t all new meters were meant to be calibrated by the manufacturers anyway. I declined and they returned my 3478A FOC.

I sat down, armed with the service manual, a AD584-M, A MS7221 calibrator and some precision resistors and set about calibrating it myself and within a couple of hours it was was done and it passed the self test on switch on. This was about 3 months ago and it still works flawlessly, since then I have gotten a Fluke 8840A and 3478A agrees with the reading on the 8840A and within the last few weeks I purchased a new Brymen BM867 which is a 5.5 digit meter and it also agrees with the 3478A and the Fluke 8840A.

So my question is why could they not get it to calibrate? I manged it with ease, in fact it was so easy I wondered why didn't I do it myself in the first place?

[texaspyro]

ERG potted brick or little PCBA on digital board

I have found that ERG is very good about keeping ancient converters in stock (or at least orderable).  I had some Tek 1503B/C TDRs with bad EL display inverters and some VFD inverters used in another product (can't remember which) and they had them in stock or built up a batch.   There was a reasonable minimum order.

[bitseeker]

Yes, reseated it,  much better choice of words

Ah, yes. Glad that's all it was. Phew!

[bitseeker]

So my question is why could they not get it to calibrate? I manged it with ease, in fact it was so easy I wondered why didn't I do it myself in the first place?

I dunno. Maybe because it wasn't auto-everything simply by connecting it to their calibrator and hitting a button.

[TiN]

Comprehensive cal is NOT enough for serviced K2001. Make sure lab will do manufacturing low level calibration, otherwise your AC and ACI/DCI functions will not be adjusted/calibrated to spec.
Equipment requirements for low-level cal are same as user level cal ("comprehensive" in Keithley words) plus the signal generator to outout 2VAC 1Hz (57XX MFC can't output below 10Hz).

[Gyro]

...

A hour or so later I got a email with the attached photo

Well, short story (yes, now I do the short story)
it did take a bad bump in shipping and the ADC board got knocked loose

He replaced it and it now powers up fine and they are letting stay powered to to begin the cal procedure on it. 
He's going to secure it for return shipping.

...

I'd take that as a total design fail by Keithley! 

There's no way that, in a properly designed piece of equipment, a board should be able to pop off like that! Just looking at it, the only mechanical support and retention is at the short edges of a long PCB, and then just flexible (board to board?) plastic clips. Board deflection must be huge under mechanical shock, most of it being absorbed on the connector pins and solder joints (and the smd parts). I'm surprised they don't get a lot of doa.

I'd have been crucified if I'd implemented something like that in a piece of consumer or industrial gear.

[bd139]

Have to agree. Philips were good at making stuff that disassembled itself in transit. Sounds like they hired some Philips engineers.

[mcinque]

Happy to hear that. I wonder if we should start to apply shockproof stickers on our packages to avoid this kind of issues...

[bd139]

A few years back i was working with someone who was making tracking devices with GPS and accelerometers for exactly that purpose. It’d track when something was buggered. They were just shipping them in the boxes and working out which courier was the most shit.

Hermes won.

[Gyro]

Those shockwatch stickers rely on you being a company and being in a position to claim full value off the shipper (or for something unavoidably shock sensitive that shouldn't be put into service - like old removable disk platters).

As an end user, I'd trade them for the manufacturer not skimping on shock and vibration testing in the first place.


P.S. Hermes won?     Ok, I've never had a problem with Hermes but you'd hope it would be someone high cost like Fedex!

[bd139]

Sorry I should say lost. Hermes won the prize of most likely to fuck up your stuff. I think UPS won he least likely.

[smgvbest]

Comprehensive cal is NOT enough for serviced K2001. Make sure lab will do manufacturing low level calibration, otherwise your AC and ACI/DCI functions will not be adjusted/calibrated to spec.
Equipment requirements for low-level cal are same as user level cal ("comprehensive" in Keithley words) plus the signal generator to outout 2VAC 1Hz (57XX MFC can't output below 10Hz).

Thanks, I will double check with them that it is a low level cal they are doing.
I did the comprehensive cal on it myself with a pair of precision resistors and 2 AD584 references in series for 20V and divider resistors on the 2.5v to get it to 2v.   actually did quite well to my surprise.
first cal I did on a meter.

[TiN]

User cal do not calibrate/adjust ACV/ACI/DCI at all