-
Yes, another repair thread about benchtop precision DMM. This time it's not so precise, as usual stuff I mangle with, but more typical and commonly used Fluke 8846 6.5-digiter.
This is possible, thanks to John for his donation in the name of science (and making another article).
I almost feel like it's Fluke invasion in the lab recently. It went from zero Fluke DMMs (87V is not accounted for) to three in no time.
Rear side already show some damage to us, by broken tab of the power entry block. No biggie, just 26$ part.
However when I try to power on thing, it just locks up in booting logo and nothing happens.
Let's see who's inside? Good old LM399H, Fluke resistor networks, ALTERA Cyclone 1 FPGA's and bunch of other goodness.
Transformer is rusty special, that's for sure. Nothing obviously broken. According to date codes on chips, it's made early 2006.
So I mangled with it for few hours, replaced all electrolytic caps (usual rule when doing any older than 10yr equipment repairs), applied freeze spray on Cyclone FPGAs and stuff around in testing for cold joints, and poof, thing suddenly came to life..
Too easy repair, I thought! I turned it off, on, off, on, seem to be working fine.
Okay, let's have it meet the Hulk (that's nickname for my zombie calibrator).
Connected, turned on for 10 minute warmup, hooked cable, and...
-3.8ppm on 10V, -3ppm on 1V, no worries.
-6.6ppm on 100V, spot on at 1000V, woohoo.
Resistance checks, +10.4 ppm on 10kohm 4W, -47 ppm on 10 ohm. Not too shabby.
ACV and DCI also working okay.
Almost spot on, lol, with zero warming up time. Too good to be true, right?
LAN port however does not work, cannot connect to it either in static IP mode, or DHCP. DHCP does not get any IP address from network. Perhaps LAN Phy is killed?
Whatever, connected GPIB, that worked. So I setup simple python script to datalog 10VDC and went to bed (it was 5:30 am already).
And sure enough, 10 minutes after I closed my eyes, poor meter died.
It hangs on same power on initializing screen.
BTW, Firmware on it is really old 1.0x, does anybody have newer 2.xx firmwares for 8846A? Fluke removed it from website for no good reason 
Any ideas what to pursue next?
-
Ask Fluke first for any FW updates, and on a previous thread someone had some boot up issues, check eevblog threads (maybe you have seen it ?)
Fluke must had a good reason to remove it loll to keep some curious eyes away loll or complicated updates procedures, i don't know, you loose nothing to ask.
Retry a more region precise freez spray to replicate your findings, as you say it can be any "cold joint soldering" and maybe an Flir camera to check thermal problems ? -
... the Hulk (that's nickname for my zombie calibrator).
You don't know I searched few days ago in google the "Hulk calibrator" nothing, "Fluke Hulk" nothing... and I gave up. -
....applied freeze spray on Cyclone FPGAs and stuff around in testing for cold joints, and poof, thing suddenly came to life.
So just to clarify, you didn't reflow or rework the solder joints? Just used the freeze spray? If you just used the freeze spray to test for cold solder joints and it started working again, then suddenly stopped once left on for a while I would look at reworking the solder joints, as it sounds like the freeze spray was enough to make the joint contract and make contact, but when on for a while it expands and breaks the connection. -
Correct, didn't reflow anything. I freezed FPGA (as my suspects for BGA ball contacts) - nothing happened. Then freezed one of the IR diodes, and it got to work those 30 minutes.
I have opto's and FPGA already in upcoming digikey order. -
I searched for
Fluke8846A-Firmware2.10.882.24.zip (which I think should also work for the 8845A)
and got a hit at
https://doc.xdevs.com/doc/Fluke/8846A/fw/
but I get a forbidden error. -
-
LOL
........
But on the "Forbidden" ...
I also ran into this error when looking to see what keithley 177 schematic quality was, (linked to it in another thread recently).
Couldn't see the xdevs one so I uploaded a decent copy to that 177 thread.
-
There is easier way?
Repository tree migrated, so old URLs borked, sorry.
I have parts from Digikey ordered for meter repair, will see if my intuition is right end of the week.
IR shot reveals not much wrong
-
What the part heating up at 78.43 degree, i find it very hot for an small part ?
-
If I remember correctly it is the LM399.
-
Yep, that is ovenized LM399. It's internal constant temperature by design is about 95 °C.
-
I have a couple of questions about 8846A. I hope you dont find it irrelevant to this thread though
-most serial numbers for this unit start with 9xx like 923 962 etc...but there is a unit I am negotiating for and the serial number starts with 118
and I also have seen SN starting with 108. How do the firmwares match to these serial numbers?
-can this Noritake display work in this DMM http://noritake-vfd.com/gu256x64d-7000bx.aspx
-finally, does this meter has the measurement auto HOLD feature that holds the measurement and shows the last few of them like the newer Keysight models? I cannot see a hold key on the front panel
EDIT: attached is the datsheet for that Noritake VFD
sorry if this seems out of scope for this thread but since TiN is here and most other DMM experts will visit this thread I am hoping someone could help
-
http://noritake-vfd.com/gu256x64d-7000bx.aspx
Ohhhhhhhhhhhh nice display would be nice on a HP 34401 like the eevblog thread ...
https://www.eevblog.com/forum/repair/hp-34401a-dmm-with-leaking-segments/ -
even mouser has it.
although it does not have the BX suffix. I dont know what is the difference.
however, my question remains: does this fit 8846A?
https://www.mouser.ca/ProductDetail/Noritake/GU256X64D-7000?qs=sGAEpiMZZMvkC18yXH9iImO64zrkImZBrdAMGtROkPM%3d -
The display is a graphics type display with driver board -so it at least the same class of display like in the 8846. So there is still a chance is might work, but I don't know.
It would definitely not fit the 34401 or the Keithley 200x meters, as these use 7 seg based displays. -
@TiN
Where did your manual repository go . I really liked it. -
Its all there, just update links due server config changes.
-
I was still hoping someone could answer my earlier question:
does this meter (8846A) has the auto HOLD feature that holds the measurement and shows the last few of them like the newer Keysight models? I cannot see a hold key on the front panel -
I was still hoping someone could answer my earlier question:
does this meter (8846A) has the auto HOLD feature that holds the measurement and shows the last few of them like the newer Keysight models? I cannot see a hold key on the front panel
As i understand they implemented "Histogram Function" which is more advanced, but not really hold function as we used to. -
It has a Memory key (next to F5). If memory is turned on, the meter can store several thousand readings internally or optionally externally if it is an 8846A. (The 8846A has a USB port that can be used to store readings.) If you wanted to see the last few readings, it has a Last softkey that could be used then you could scroll backward. It also has an Analyze key (which is separate from the Memory function) that can be used to turn on STATS. Then things like average and standard deviation and high/low can be displayed and it can show a graph of the readings. Histogram is a graph of the standard deviations.
-
It has a Memory key (next to F5). If memory is turned on, the meter can store several thousand readings internally or optionally externally if it is an 8846A. (The 8846A has a USB port that can be used to store readings.) If you wanted to see the last few readings, it has a Last softkey that could be used then you could scroll backward. It also has an Analyze key (which is separate from the Memory function) that can be used to turn on STATS. Then things like average and standard deviation and high/low can be displayed and it can show a graph of the readings. Histogram is a graph of the standard deviations.
yes, I know but that would also store all the readings between probings. With reading hold (or probe hold as keysight call it) it stops
capturing when you take your probe away and reads a new measurement when you touch the probe and only if the new value
is different than the last one by more than 1%. With simple memory it keeps reading and storing the noise when you are not probing.
I have looked all over the manuals and I am 100% convinced now that neither 8846A nor DMM6500 have this simple useful feature.
All keysight multimeters and most older Fluke have it and even rigol and siglent have this. It's a simple software thing and i dont know why they decided to do away with it
-
Just curious what it took to fix unit ? If you needed FPGA.
-
It's still dead.
-
Bummer
Just asking cause I had chance to buy a couple of non-functioning units or a couple of functioning units.
After your answer I decided on functioning units.
I have to correct that... I bought a couple of units with errors which keep unit from functioning!