Fluke 289 defective or not?

[mqsaharan]

Hi.
My fluke 289 shows negative figures in Conductance function. Does this meter show negative values for very high resistance or very low (near zero) conductance? Is this normal behaviour?
Also, I am experiencing some random figures in mVDC function with shorted input probes that keeps on drifting slowly. It used to show zero and one day it suddenly decided to change its behaviour. Is it normal for this meter?
On the other hand, mVAC and VAC functions used to show some counts (random figures) with shorted probes now they show zero. Is it OK?

Thanks in advance.
Qasim.

[Electro Detective]

Check for low battery, internal battery or supercapacitor leakage on the circuit board, dirty input sockets, faulty leads etc first... in no particular order

before considering to get the sledge out 

[mqsaharan]

It is hardly ever used or never used kind of situation. I have installed new batteries and changed probes, etc. There is no super-capacitor leakage as it was purchased new in 2016 and it has spent 95% of its time in its bag.
It is with Fluke service centre at the moment. And the guy who is in charge of my meter is saying that it is all right. And he has given me the following explanation for the negative reading in Conductance function.


"A negative conductance value is normal as well. The meter measures resistance and conductance by passing a small amount of current through the device under test (DUT). The meter then records the value of this current (along with the voltage needed to generate it) over a short period of time and calculates the derivative of both values. The meter then divides the derivative of the voltage values by the derivative of the current values, and this results in the resistance value that the unit displays. This is more specifically known as a “differential resistance” measurement, and the conductance (“differential conductance”) is just the reciprocal of this resistance value. If one of the derivative calculations results in a negative value, this will cause the resistance or conductance value to be negative. This is common in very-high or near-infinite resistance measurements, open-circuit measurements, no input situations, or other situations where the conductance of the DUT is close to 0. For example, this unit is reading -0.03 nS with the conductance function selected and no input whatsoever. This is normal behavior and does not affect the unit’s accuracy over the specified measurement ranges."


Now, he says that it is common for a meter to generate a negative value in very high or infinite resistance measurements but I haven't seen one. So, I thought that more experienced people might have seen it. That is why I posted it here.

Two things that are bugging me are:
1. It has started showing random value and keeps on drifting slowly with input probes shorted in mVDC function. It never did that before.
2. It is showing negative readings in Conductance function. None of my old Flukes 8020A and 8060A are showing negative values near or at zero nS. And I don't seem to correctly remember if 289 was like that before.

So, if you have a Fluke 289, please tell me if your meter shows negative values in Conductance function for, say, a reverse bias silicon rectifier diode.

Qasim.

[Electro Detective]

I just tested the 289 alongside 87V and 189 in nS/conductance mode, all three behave in a similar fashion,

the trusty 8060A to compare with is awol at the moment => 


If I jiggle the leads of the 289, I get random plus and negative values, and OL when leads shorted

I only had a diode and white LED to play with on nS, and got the expected OL readings one way, and 70 to 150 nS reversed,
same deal with the other meters

FWIW A 'Reset' of the meter and total memory wipe may be an option to consider, and take the batteries out and let it sit for a while, or do the process vice-versa 

I noticed on mine the inputs had fine dark soot (free resistance? lol) which I cleaned out with some WD40 lightly applied to a small head cotton bud/Q tip
with the meter facing upside down so no chance of fluid sneaking in or residing during the clean and lube. Sockets are always shiny now and easy to view 

Mine has the leaking capacitor issue, but cleaned up and the dried cap behaving itself.. till I do something about it 

Lastly, the supplied standard Fluke leads (and my other Fluke leads) do not sit well in the 289 meter, and I'm not the only user/owner to notice

The fit is way too tight and who knows what kind of actual metal to metal contact is going on in there,
with the plastic sheath rubbing up against the sides or whatever is going on, and it can be a hassle yanking them out without KO-ing youself   
I use those same pesky leads with my other Flukes and no problem, as they are basically all the same anyway   

I have a nice pair of flexible APPA branded leads with same Fluke specs, and they plug and unplug beautifully     

Not sure if my limited experience getting my 289 up to speed may help you iron out yours, but good luck   

and hopefully some other 289 user members here may take this further

[evava]

Two things that are bugging me are:
1. It has started showing random value and keeps on drifting slowly with input probes shorted in mVDC function. It never did that before.
2. It is showing negative readings in Conductance function. None of my old Flukes 8020A and 8060A are showing negative values near or at zero nS. And I don't seem to correctly remember if 289 was like that before.

So, if you have a Fluke 289, please tell me if your meter shows negative values in Conductance function for, say, a reverse bias silicon rectifier diode.

Qasim.

Here https://www.eevblog.com/forum/beginners/is-my-new-fluke-289-multimeter-faulty/
I recorded fluctuations on my F289 in the mV range with input posts shorted (over 24 hours) and I expressed disappointment there.
Others in that thread confirmed that behaviour is normal at F289, so...
After that disappointment I bought also UNI-T UT181a, which behaves fine and stable in mV range.

However, no negative (-) reading at nS range at my F289.

[mqsaharan]

Electro Detective
Thank you for checking your meters along with F289 and also for your helpful hints and reminders.
As I wrote before that my meter is with the Fluke service centre at the moment and hopefully socket cleaning should have been taken care of but I'll check it thoroughly. I played with it including resetting it before forwarding it to the service centre but it didn't have any effect.
If I remember correctly, one of the probes I got also had a tight fit but I jammed it in with force and it has been loosen up to the normal insertion force since then. But I am not suggesting you to try it. I am just stating my experience.
 
evava
Thanks for the link. My multimeter behaves exactly like yours in mVDC range. And additionally it is showing negative values in Conductance function. When I asked, the Fluke service centre replied in detail which is in my previous post. Summary is, they are saying it is normal. But I don't like it.
If I remember correctly, it wasn't negative when the meter was new. Its just one day (and that was about 9 months after the date of purchase) my meter suddenly decided to change its behaviour and started showing the problems which I stated in my original post. And the important thing is before changing its mood, it told me by flashing some message which I couldn't read and then getting totally stuck/frozen (there were no leads attached at the time) and didn't accept any input through any buttons including on/off button. In less than a minute it automatically reset itself. On next turn on, It again flashed some message that I couldn't read because it disappeared before it was completely on the screen like the first time. But at least it didn't get stuck ever again.
Now the problem is I cannot tell Fluke service centre how to reproduce its behaviour of getting frozen and prove them that something is wrong with the meter. And they are constantly saying that the meter is all right and totally within specs. And its "changed" behaviour is also "normal". Then this could be a software bug.
Oh I forgot one thing. You'll be surprised. When I got my meter, I opened it up after a few days and the first thing that caught my eye was something attached to the plastic shield due to static cling. If my memory is serving me right, It was some dirt and some thin small flat fibers of some plastic type material having kind of same texture as plastic shields on the PCB. It was perhaps from the packing material. It made me curious and I disassembled the rest of the meter and found the same stuff stuck, again due to static cling, to the plastic shield on PCB behind the display as well. Anyway, I brushed it off and assembled the meter and it kept working fine.

All I can say at the end is that Fluke is not that Old Fluke any more.

Qasim.

[Electro Detective]

If I remember correctly, one of the probes I got also had a tight fit but I jammed it in with force and it has been loosen up to the normal insertion force since then. But I am not suggesting you to try it. I am just stating my experience.


 
All I can say at then end is that Fluke is not that old Fluke any more.

I got fed up with jamming both my probes in with jack hammer force and avoiding Mike Tyson jaw taps pulling them out    

The APPA leads are just awesome, fit and connection perfect and flexibility heaven   


And yes, unfortunately the old Fluke is all but gone    no matter what paid and unpaid apologists will say to the contrary   

[evava]

Its just one day (and that was about 9 months after the date of purchase) my meter suddenly decided to change its behaviour and started showing the problems which I stated in my original post. And the important thing is before changing its mood, it told me by flashing some message which I couldn't read and then getting totally stuck/frozen (there were no leads attached at the time) and didn't accept any input through any buttons including on/off button. In less than a minute it automatically reset itself. On next turn on, It again flashed some message that I couldn't read because it disappeared before it was completely on the screen like the first time. But at least it didn't get stuck ever again.

Interesting findings - I wonder if other people that bought meter new noticed that.
Unfortunately I can not confirm it cause I bought used meter, I am experiencing this behaviour from the beginning of my use.
I do not like unstability of uV DC reading (but it still meets the spec probably), otherwise it is useful and worthwhile meter and I do not want to part with it.

[evava]

I have tried today again and the same result as blueskull - (no wires in input posts) and almost all the time F289 shows -0.01nS, sometimes shortly jumps to 0.00nS.
I attach photo where F87V and F289 are struggling vainly to find zero on nS range (but using REL key helps certainly), while UT181a having no problem with zeroing 

[Electro Detective]

Surely all that input protection on the Flukes would contribute to some Conductance nS   -+sway-+   

Whereas the Uni-T and other BANG4buck meters being almost wide open for a zap nuke, zeros out

Not sure if it's relevant, but the 289 input impedance drops down to 8 Megohm when fed with 1000v DC from an insulation tester,
whereas with 500VDC and below the meter reads the expected 10 Megohm

Apparently this is due to the input protection stuff discussed at another post/s.

The Fluke 87V on the other hand holds at 10 Megohm at 1000VDC and below,
so it must have a different input protection arrangement, as there is less stuff in it to zap like in a 289

I am assuming 'if' the input protection was toned down or bypassed on some high end Flukes they would kick some serious precision/accuracy butt and settle/zero out faster, 
but you can kiss their CAT ratings goodbye, and most likely their insides too on a mega oopsie or tragic carpet surf     


Perhaps it's time to invite in some better clued multimeter heavyweights on this one: Dave Jones, Joe Smith, Fluke 289 designers etc 

[Electro Detective]

Page 17 in the Fluke 289/287 Calibration Manual:

"Note
The DC mV measurements in Table 4  are subject to offsets caused by thermals due to temperature gradients between connecting materials.
After connecting the calibrator to the Meter terminals, allow time (could be several minutes) for the thermals to dissipate.   
When readings are stable, perform a REL and record the Meter reading."

----------------

Fluke owners manual Page 34:

"There is normally a residual conductance reading with the test leads open.

To ensure accurate readings, press the softkey labeled Menu.

Move the menu selector to the menu item labeled REL and press the softkey labeled REL
to subtract the residual value with the test leads open."

-------------

Specs:

Conductance Range 50 nS (Add 20 counts above 33 nS in 50 nS range)

Resolution 0.01 nS

Accuracy 1%   +10   

-------------


EDIT:

Tested the 289 vs 8060A in nS Conductivity mode using the same Brymen gold color tipped leads

The 8060A settles on 0.00 within a few seconds upon switch on, and or after fiddling with the leads or shorting them etc
Always been a solid performer IIRC


The 289 may need a bit longer to get close to 0.00, and varies between   - 0.01 to 0.04 

I noticed if you short and open the tips repeatedly the meter appears to settle in faster


FWIW If you really need    [0.00 nS]    and refuse to settle for anything less       
Menu up the meter for Relative /REL mode,
and switch it in and out till you hit on 0.00

It's elusive but do-able after a few tries... just don't move about too much, breathe heavy, push Menu buttons too hard, or rattle the leads

or curse too loud in case the neighbors freak  and make another phone call again reporting LAB RAGE

[evava]

Surely all that input protection on the Flukes would contribute to some Conductance nS   -+sway-+   

I am assuming 'if' the input protection was toned down or bypassed on some high end Flukes they would kick some serious precision/accuracy butt and settle/zero out faster, 

My assumption too - fully agree with you.
Whatever, Flukes are most trusted meters in my shop.

P.S. Zeroing is sometimes fake, like at AN8008 (mV and mOhm ranges have unpleasant and annoying gap around zero), IMHO much better is zeroing by hand and preserved linearity.

[Electro Detective]

That's the catch with cheap Landfill bound meters and BANG4buck$ bells n whistles Fluke competitors,
they give great readings in some modes and questionable digits in others

No thanks,    if I'm sticking a meter in somewhere to test a suspect component or tracing a circuit path/fault,
the last thing I need is to pray the 'test' meter is doing it's job,
and not another suspect or intermittent item I need to be testing, unpaid R+D safety modding and repairing too   

hey, I'm cheap     but not that cheap (I think...)

 

[hgjdwx]

DT4282's DC60mV range and nS range  have a good performance in this respect

[mqsaharan]

I do not like unstability of uV DC reading (but it still meets the spec probably), otherwise it is useful and worthwhile meter and I do not want to part with it.

I also don't like instability of a multimeter's reading with no input. It is distracting and sometimes misleading. But I am also not well-versed in these things. There might be design limitations or budget limitations or may be the designers just get bored at the end and call their design final.

Mine shows -0.01 on open circuit, and with some fluctuation in air, -0.10~-0.01.

I have tried today again and the same result as blueskull - (no wires in input posts) and almost all the time F289 shows -0.01nS, sometimes shortly jumps to 0.00nS.
I attach photo where F87V and F289 are struggling vainly to find zero on nS range (but using REL key helps certainly), while UT181a having no problem with zeroing 

It is clearing up my mind. I have been using only 8020A and 8060A before 289 showed up. They do fluctuate with leads plugged in. With no leads, open inputs, both are solid at 0.0nS.

Surely all that input protection on the Flukes would contribute to some Conductance nS   -+sway-+   

Whereas the Uni-T and other BANG4buck meters being almost wide open for a zap nuke, zeros out

What I have seen in pictures, some of them do have input protection. Their effectiveness is a whole other question, like in the case of UT181 with Joe's grill starter test.

Not sure if it's relevant, but the 289 input impedance drops down to 8 Megohm when fed with 1000v DC from an insulation tester,
whereas with 500VDC and below the meter reads the expected 10 Megohm

Thanks for some more details about 289. Such information is usually hard to find. With auto-ranging meters, their behaviour is lot less documented in user manuals than the old manual ranging ones.

FWIW If you really need    [0.00 nS]    and refuse to settle for anything less       
Menu up the meter for Relative /REL mode,
and switch it in and out till you hit on 0.00

My meter was reading negative values during measurement.
It still bothers me if it does not say 0.00nS with open inputs but I guess I can make my peace with it.
When it was new, it used to say 0.00nS with open inputs. I guess it got tired of staying in shape all the time and wants to relax from now on.

My U1461A is packed with bugs, and its faulty input mux occasionally gives me 0V reading on hot input. If I trust it on mains measurement, I will no doubt kill myself sooner or later.
However, it is the only meter with 1V resolution, 1.1kV power supply built in with nA resolution current read back, and the comparable model I use at work, Keithley 2410, is $7600.
I paid $730 for my U1461A, and for nearly 7 grands of price difference, I have nothing to complain about.

Good to know more about U1461A. it was on my list. At least I'll be more mentally prepared for its quirks.
I think you should report its bugs to Keysight. They may get resolved in the next firmware, if they decided to work on it.

My F289 came with 1.7V null reading on LoZ mode, and it climbs slowly to 2.1V, even with input short.
I didn't find any threads discussing this issue here, but from a Chinese forum, the conclusion is that the true RMS chip has a bug, and how severe it is is pure silicon lottery.
This stupid bug was with my meter for years, and it slipped through one calibration. Until later the meter was damaged by battery juice and was sent to repair, a board swap got rid of the faulty tRMS chip.

Mine used to read somewhere around 0.0 volts in LoZ mode. I don't seem to exactly remember its status.

And how may people here bitch on the self corroding ultracap? It's a stupid design decision that should not be made by Fluke.

It seems more like an overlook by having an understanding that the meter will be back after one year for calibration. Companies usually do that, most individuals don't. And the test gear is usually specd towards companies' needs and not towards individuals'.
I am just trying to take a guess about their thoughts on the matter.

Furthermore, an industrial data logging DMM that doesn't come with PC software? Seriously? All, I mean all, UniT and Keysight, have free software. Maybe you need to purchase a dongle, but that's it.
Fluke not only sell the dongle, it actually sells the licensed software, so grabbing a $20 used dongle on eBay won't work due to the fucking license model.

You should take a look at Gossen. I was very surprised to see their prices vs their specs.

Surely Fluke doesn't have a good competitor for now, but if UniT rolls an UT181 with pimped up input protection, or KS rolls an U1272 without firmware bugs, there will not be too many reasons to buy an F289 anymore.

That is true. Fluke lack serious competition. Others copy them instead of being creative.

DT4282's DC60mV range and nS range  have a good performance in this respect

Whenever such information is needed, it is impossible to find.

[mos6502]

I have tried today again and the same result as blueskull - (no wires in input posts) and almost all the time F289 shows -0.01nS, sometimes shortly jumps to 0.00nS.
I attach photo where F87V and F289 are struggling vainly to find zero on nS range (but using REL key helps certainly), while UT181a having no problem with zeroing 

As you said, that's just Chinese Fake Zeroing. The Uni-Turd 139C does that as well. Also the tachometers one many modern cars work like that (engine RPMs are fluctuating at idle when cold, but the software just displays a rock steady RPM value to give the impression of "quality".)

[KMSL]

The operation of your Fluke 289 seems "normal". Mine of 2012 does the same.

[Electro Detective]

...It seems like you have some idolism with Fluke, and let me tell you that's not good.


...Surely Fluke doesn't have a good competitor for now, but if UniT rolls an UT181 with pimped up input protection,

or KS rolls an U1272 without firmware bugs, there will not be too many reasons to buy an F289 anymore.

Before I feebly defend any incinuation of Fluke idolism (lol)    

I'd like to state there won't be any Uni-Turd or buggy halfassed eye candy Agisight or Keylent in my humble inventory any time soon       or Windows 10 PC or High Sierra Mac

TBH it sux that member blueskull got shafted on two big dollar meters    but up to him to kick up a big stink  and get them to repair or replace said items,
or a compromise 'out of warranty' price, 
or he can offer the refusers  a filmed Youtube video of the issues and or possible dangers...
and you never can tell what content will pop up from Youtube commentors.. who in their frustration may post  -overpriced dud product-  videos too       

 
That said, may I state unbiased that anything outside of Fluke 289 ~normal~ behaviour may be an indication of time for a calibration, possibly from component drift perhaps,
who knows where the parts are sourced from much less batch quality checked nowadays   

Get a calibration preferably done at a professional and reputable facility that know what they are doing and have strong service ties with Fluke in case you need a board swap to replace the Friday afternoon assembly line 'Revision'
and don't charge what a new meter would cost 

Avoid the pretty Cal sticker    slapping cash grab turkeys, who prey on the easy clueless corporats


FWIW I bought my 289 used and abused, and after opening it up and doing all the TLC stuff described earlier above, it works like a new boxed one,
unbelievably still running with the no name cheapo batteries it came with   

That's another point too, if having issues check the batteries individually under load or try another set,
and make sure they are secure in their slots with no play in the space between the meter's back
A soft piece of cut foam or similar used as a packer will keep them where they need to be, especially if the meter gets knocked or bumped
This applies to any meter with AA or AAA batteries for that matter,
the last thing you need is the suckers slightly popping out and sort of drifting between 4.5 and 6 volts with a 4 x AA set
especially in resistance, nS, diode and capacitor modes