New 17B+, Is This Normal?

[robocoyote]

Bought this meter on eBay from a US seller, it arrived Saturday.  It has some weird behaviors.  For background, I'm an electronic tech that works in the transportation industry as well as an extra class ham, I'm not new to electronics.  Previous to buying this I've had a series of cheap meters and this is my first nice meter.  Now on to the problems.  When I measure voltage it overshoots.  For example, when I measure a 9V battery the voltage shoots to 42V and then comes back down to 9V, the high voltage LED comes on also.  It also takes as long as 10 seconds before the reading settles down to the right value on resistance.  It fails the classic continuity test, only beeping about 50% of the quick contacts.  I've tried putting new batteries in but it makes no difference.  Is this normal for this meter?  It's much worse than my Mastech MS8264 in almost every way except build quality.

[Fungus]

Definitely not normal.

[MosherIV]

Nope. Sounds like it is damaged.

Did the seller describe it as working?
If so a)send it back and get a refund
B) get a partial refund and get help here to fix it

Not sure from your description what is wrong.

[joeqsmith]

I have a 17B+.  Nice meter.  No, it does not do what you describe.  Maybe someone sent you a used damaged one.   

[robocoyote]

Thanks everyone, I bought it new from someone who imports the Chinese model into the US.  I've contacted the seller and will see if he will exchange it. 

Joe: I watched many of your videos while researching a new DMM, nice work.

Rob

[crazyguy]

is it Fluke 17B+ or other clones with the model 17B+?

[robocoyote]

It's the Fluke 17B+.

I found an old review (2010) of the 17B model here:
http://robotics.ong.id.au/2010/09/07/fluke-17b-review-part-2/

Under "The Bad" he lists the exact problems I'm having:
Displays incorrect value just after auto-ranging. The reading overshoot for one sample reading.

Measurement are generally fast, however resistance autorange is slow. It takes about three and a half second to measure 100Ohm resistor.

Continuity tester is annoying. The meter goes beep, then random duration of silent for a fraction of a second, followed by continous beep.

I brought the meter to work today and compared it to the 87V's and the ExTech EX430 that we have.  I didn't expect it to be on par with the 87V but it is just terrible to use, worse than the ExTech.  I'm asking the seller for an RMA.  I don't expect a replacement to be any better so what are my options for a $100 - $125 meter?  I'm looking at the Amprobe AM570 and the EEVBlog Brymen.  I already have a Uni T 139C as my second meter, I want a high quality meter as my primary.

Rob

[Fungus]

It's the Fluke 17B+.

I found an old review (2010) of the 17B model here:
http://robotics.ong.id.au/2010/09/07/fluke-17b-review-part-2/

Under "The Bad" he lists the exact problems I'm having:
Displays incorrect value just after auto-ranging. The reading overshoot for one sample reading.

Maybe that reviewer had the same defect as you.

Somebody who actually owns a 17B+ (two posts up) says it doesn't, his video is here:

Measurement are generally fast, however resistance autorange is slow. It takes about three and a half second to measure 100Ohm resistor.

If it's overshooting internally then that might happen, yes.

[joeqsmith]

I swear, I did not sell him my slightly used one! 

[robocoyote]

Mine is much slower than yours Joe, even when you didn't have the test sped up.  Something is definitely up with this one and I'm returning it. 
Rob

[joeqsmith]

The settling time on R was real time. It's fairly fast.  I'm not sure if there is different revisions or not.  Mine has a sticker starting with A17BR091524...  Maybe 17B, Rev 9?   Last three digits on the PCB are .809   

If you wanted to try something, using the same 9V battery test, select the manual range.  Now try it.  If the meter does not over shoot, I wonder if you are just seeing an artifact like my CEM meters when they switch ranges being off by 10X. 

[robocoyote]

It doesn't do it when I manually set the range.  When in autoranging it doesn't always display the same overshoot, sometimes it starts at 42, sometimes 32, sometimes 25, and other times it doesn't overshoot at all.  It always takes about 6 seconds to find the resistance though.  The meter goes through this process when in resistance mode: shows a number that's too high, then shows a number that's close to right and then counts down to the right number.  For instance with a 100k resistor the display looks like this:
175
107
106
105
104
103
102
101
100
Like it gets close and then steps to the right value over 8 or 9 screen updates then stay steady, weird.
Rob

[joeqsmith]

Ok, I now think yours may not be defective at all.  You would have to just shoot a short video showing what you are seeing for me to really compare how mine behaves but they are starting to sound similar.  I think we just have a misunderstand of what is happening.

When I go from open to 100K mine will start low, then work up to about 99.6K within a few seconds but to fully settle out does take about six seconds total.  At 40M it's about 10 seconds to settle.

If I sit and watch mine while pulsing the input from 0V to 5,  I see it go through a decimal point error.  So rather than 2.2 it shows 22 sort of thing.  It does not always do it but I can get it to happen.  When it does, it recovers on the next sample.  Its a similar problem with the CEM except with the CEMs able to log the data it can cause some problems.  For me,  I would not call this an overshoot but a decoder problem with the decimal but I suspect we are talking about the same thing.     

Continuity I expect is normal.  Most people just touch the leads together for the reviews but I have been taking the time to actually collect a fair number of metrics on this feature for each meter which I include in the spreadsheet.   Not a fan of unrepeatable tests.   The 17B+ is slower than most.  So if what you consider a classic test is just touching the leads together, I can believe you are faster than the meter, even if you are 90 years old and have suffered a stroke!   

Like a I say, maybe just shoot a short clip or try and measure the continuity and compare with the data in my spreadsheet.  For what you refer to as oveshoot, see if in auto range if it is really an overshoot or is it a decimal point error like mine during the auto range for a single sample.   Resistance seems normal for the 100K. 

Let me know. 

[crazyguy]

It's the Fluke 17B+.

I found an old review (2010) of the 17B model here:
http://robotics.ong.id.au/2010/09/07/fluke-17b-review-part-2/

Fluke 17B is based on FS9721_LP3 chipset
http://www.ic-fortune.com/upload/Download/FS9721_LP3-DS-21_EN.pdf

Fluke 17B+ is based on FS98O24 chipset
http://www.ic-fortune.com/upload/Download/FS98O24-DS-02_EN.pdf

[Electro Detective]

Lose the 17x and get a good quality Made In China Fluke 114, 115, 116, or 117

One of the four models should have most of the features you need

They cost more, but are the real deal

and work as advertised ... 

or cast off the 'budget Fluke' mindset and get an EEVblog meter   


"Is Dave, Is Good..."   

[retiredcaps]

Fluke 17B is based on FS9721_LP3 chipset
http://www.ic-fortune.com/upload/Download/FS9721_LP3-DS-21_EN.pdf

Fluke 17B+ is based on FS98O24 chipset
http://www.ic-fortune.com/upload/Download/FS98O24-DS-02_EN.pdf

At one point, I knew off the top of my head, but I have now forgotten and not 100% sure without researching again, but many of these Fortune Semiconductor chip sets found in various meters all tend to overshoot the first reading and then settle to the correct one.

[robocoyote]

Ok, I now think yours may not be defective at all.  You would have to just shoot a short video showing what you are seeing for me to really compare how mine behaves but they are starting to sound similar.  I think we just have a misunderstand of what is happening.

When I go from open to 100K mine will start low, then work up to about 99.6K within a few seconds but to fully settle out does take about six seconds total.  At 40M it's about 10 seconds to settle.

If I sit and watch mine while pulsing the input from 0V to 5,  I see it go through a decimal point error.  So rather than 2.2 it shows 22 sort of thing.  It does not always do it but I can get it to happen.  When it does, it recovers on the next sample.  Its a similar problem with the CEM except with the CEMs able to log the data it can cause some problems.  For me,  I would not call this an overshoot but a decoder problem with the decimal but I suspect we are talking about the same thing.     

Continuity I expect is normal.  Most people just touch the leads together for the reviews but I have been taking the time to actually collect a fair number of metrics on this feature for each meter which I include in the spreadsheet.   Not a fan of unrepeatable tests.   The 17B+ is slower than most.  So if what you consider a classic test is just touching the leads together, I can believe you are faster than the meter, even if you are 90 years old and have suffered a stroke!   

Like a I say, maybe just shoot a short clip or try and measure the continuity and compare with the data in my spreadsheet.  For what you refer to as oveshoot, see if in auto range if it is really an overshoot or is it a decimal point error like mine during the auto range for a single sample.   Resistance seems normal for the 100K. 

Let me know.

Well, maybe it is working normally then.  Still, it's too annoying to be my everyday meter so I've boxed it up and the seller is going to refund my money.  I don't really want to unbox it to do a video.  I now have $110 Paypal credit, what should I buy?  I'm thinking the EEVBlog/Brymen but if you think there's something better in the $100 - $125 range, let me know.  Oh, and where can I find your spreadsheet?  Thanks for your help.

[joeqsmith]

Well, maybe it is working normally then.  Still, it's too annoying to be my everyday meter so I've boxed it up and the seller is going to refund my money.  I don't really want to unbox it to do a video.  I now have $110 Paypal credit, what should I buy?  I'm thinking the EEVBlog/Brymen but if you think there's something better in the $100 - $125 range, let me know.  Oh, and where can I find your spreadsheet?  Thanks for your help.

Hard to say.  You have a UT139C but "want a high quality meter as my primary".  But you went from 6000 counts to 4000 and an TRMS meter to averaging.  You even had NCV that you gave up with the Fluke.  You can see why I don't recommend meters.  I would really have no way of knowing what you need/want beyond the things you mention being a concern with the 17B+. 

If the 139C does what you need, just stick with it.  Save up your money until you find something you actually need or want.   I'll continue to use my old Mastech in the garage till it starts having problems. 

Overall, for electronics work I still like the BM869s as a general purpose meter.  I'm also liking the old 189 that my friend gave to me.  The more electrically robust brands (assuming that metric would have something to do with quality) are Brymen, HIOKI, Fluke and maybe Gossen.   The Fluke 87V seems to be a bit of a flyer but every other meter I looked at from them were very solid.   I was not impressed with Gossen from a service or product standpoint.  The meter I bought has way too many problems.  Still, from my tests it's right there with the best of them.  It's just not a well thought out meter.   You could have a look at Keysight.  The one I tested did not hold up near as well as I would have expected.  It was a new design, not like the antique 87Vs so really no excuse IMO for this brand to have done so poorly. 

If you just want something with a very fast non-latching continuity test, some of the meters like the UNI-T UT90A can handle over 10KHz.  Some people like the latching.  For me, I like the non latch for the feedback and speed for electronics.   In the garage or basic electrical work, I could care less.     

Anyway, hope that give you some idea on my view.  Download and read the manual before you buy.  If you have questions, ask first.  If there is something you want demonstrated, I would ask.  Chances are someone will have the meter you are looking at and have a camera.   

If you click the link in my signature and select about or discussions you will find the link to the spreadsheet.  It's also in the very first post of that long thread. 

[CatalinaWOW]

I am not familiar with this meter, but have observed similar behavior on another brand meter that WAS defective.  The problem was a near open (few hundred K) in a series resistor in the input circuit.  Capacitive coupling across this near open caused the initial overshoot.  The charge gradually bled off through the remaining conductance and the reading would stabilize at a correct or nearly correct value.  High impedance of downstream circuitry made the much higher resistance in the input network insignificant after initial charge transfer (but was the reason the tiny capacitance involved had such a large effect).

If you can't get the vendor to replace the unit you may be able to poke around inside for a similar problem in yours.