EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: asdqwerty on June 05, 2016, 05:53:17 pm
-
Fluke 89 IV:
Last time it was successfully used, it was used to check 3~ capacitors and voltage.
Some days later it was to be used to measure some resistance.
But it shows 0.00 ohm, it beeps when switching to beep mode. It should read OL, before start.
Switching to voltage gves 0.00.
Althoug at times, it might rush around displaying any number, both it resistance and voltage mode. But for the most 0.00 and it does not matter what the test leads do. Open, connected to something or shorted.
Fuses are checked, batteries changed. Still the same.
So it seems like something in the input stage might have a break / gap.
Tested with a different ohm-meter, by setting one test lead in the input-socket, and followed the pcb. I got a reading in front or after all the components it seems. Nowhere there was OL, it varied from few ohm to a lot of ohms. Don't know if its a lead... :-//
Anyone have any idea of which component to investigate more or try to change??
There are pictures of the 89 PCB on google and this site.
-
Take the MOVs out and see what you get.
-
I'm having a hard time following what works and doesn't. As such, let's get a baseline.
Q1. If you measure a 1.5V AA cell, do you get 1.5V or 0.0V?
Q2. If you short the probes with meter set to DCV, do you get 0.0V?
Q3. If you short the probes with meter set to resistance, do you get roughly 0.2 ohms?
Q4. If you remove the leads with meter set to resistance, do you get 0L?
Q5. If you measure a 1k ohm resistor, do you get 1k ohm?
-
Q1: 0.0V
Q2: ~0.0V -> no change with probes shorted or not
Q3: 0.00 - same with probes open or removed
Q4: No, it says 0.00
Q5: No, it says 0.00
Resistance reads 0.0 , while VAC and VDC may have the last digits run up or down.
Holding "hold" while power up shows the lcd-dispay test.
I think the meter is working, but there is a gap in the signal path..
Also tried the mA and A, with the current in a ledflashlight.
No readings.
So it seems all 4 inputs are out of order, or the one thats common for all...
-
Well, if we use this picture from
https://www.eevblog.com/forum/repair/fluke-89-iv-repair-after-input-overload/ (https://www.eevblog.com/forum/repair/fluke-89-iv-repair-after-input-overload/)
as a reference, check the following components all "in-circuit".
Q6. The 1k fusible resistor (circled in green) should read 1k ohm using another working multimeter.
Q7. The three green resistors (circled in blue) marked 100k ohm should read 100k.
Q8. The yellow component (circled in yellow) with a cut through it is a spark gap. It should read 0L or open with another working meter.
Q9. The two red components (circled in red) are MOVs. They should read 0L or open with another working meter.
Q10. The brown/grey component (circled in brown) between the fusible resistor and 100k ohm resistor is a PTC. It should read around 1k ohm.
Report your readings for Q6-Q10.
-
@asdqwerty, please post clear focused pictures of YOUR board so we can see if there are differences. Please provide pictures of both sides of the pcb and one close picture of the COM jack solder area (i.e. 3 pictures in total).
-
Q6: Reads 3.5 Kohm - it is marked 3.5Kohm also
Q7: They read 100K
Q8: Says OL
Q9: The one closes to the yellow with a gap read 11.6Mohm and 11.9Mohm when changing polarity. The other one is OL.
Q10: Reads 1.15Kohm
All measured with a Fluke 189.
Its a slightly different PCB it seems.
BTW thanks for taking your time. :)
-
Q9: The one closes to the yellow with a gap read 11.6Mohm and 11.9Mohm when changing polarity. The other one is OL.
Looks like Monkeh has the right idea. The MOV reading 11.6M ohm might be bad. Remove it has Monkeh suggests and retest Q1-5. Also measure the removed MOV out of circuit just in case something in-circuit is affecting the resistance measurement. If the MOV still reads 11.6M ohm, it is definitely bad.
-
BTW, the contacts on the rotary switch could use a good cleaning with IPA (unless it is light photography trick?). Another reason to see pictures of YOUR pcb and not one already posted here or on the Internet.
-
One more picture
-
Another one
-
Also tried the mA and A, with the current in a ledflashlight.
No readings.
Hmm, I missed your edited statement above. Let's see if removing the suspect MOV first fixes V/ohm. Then you can retest the current measurements. One problem at a time.
-
I removed the actual MOV. It gives a readout OL when tested.
Placing the test-probes on the soldering points of the MOV gives 11,4Mohm, so thats where it came from.
I then put back the MOV. No change.
When putting the multimeter on VAC, it says OL for a short while, then displaying results in the range of 450-500 V, when test-probes are not connected.
VDC = 0.00
Ohm = 0.00
-
11Mohm sounds familiar.
Put it in VDC and manually change the range, see if you get sensible readings.
-
Regarding the dirt on the your 89IV rotary switch as per your pictures, see
http://mrmodemhead.com/blog/fluke-83-iii-rotary-switch/ (http://mrmodemhead.com/blog/fluke-83-iii-rotary-switch/)
Give them a clean.
-
Placing the test-probes on the soldering points of the MOV gives 11,4Mohm, so thats where it came from.
Those two soldering points should definitely read open circuit.
-
Placing the test-probes on the soldering points of the MOV gives 11,4Mohm, so thats where it came from.
Those two soldering points should definitely read open circuit.
If the meter is switched on at the time, you should read across the input impedance - 11.4Mohm is not far off the mark. Assuming it's a single MOV design, not the series arrangement some Fluke models use. I'm not familiar with the specific circuit here, but that figure is.. rather close to what I expect to see across the input.
-
Maybe worth mentioning... the multimeter was exposed to rain late last summer. It was forgotten outdoor, and quite wet. But after drying, it has worked flawlessly since. At least 8-9 months..
--
So I tested more, cleaned the rotary-switch circuit, and put on a tiny solderingtip on the switch itself, in case the pressure isn't sufficient to make a steady contact. But no difference.
It doesn't behave 100%the same every time:
Either VDC started working, or I tested too few times.
But now, VDC is working 9 of 10 times. My bench-powersupply gives 2-15VDC and it reads as expected. (auto and manually)
mVDC however is steady 0.00. - It shoud read like 2900mV @2.9v - The 189 does that.
Ohm-setting 99.99% of the time shows, steady 0.00 Ohm - the beeper mode beeps all the time.
VAC is totally wrong. Often peaks at OL and the bar indicates >1000V
Or it is around -0- , and when I try the wall-outlet it says 40-50 or 65-75 or doesn't react.
Or it displays approx 450VAC with no probes, and runs up and down.
Amps gives nothing. 0,0A 0,0mA -Lightbulb/led in the flashlight is lit.
The measuring done with the MOV out of the pcb was done with no power attached.
-
Maybe worth mentioning... the multimeter was exposed to rain late last summer.
Can you remove the LCD from the pcb and post clear focused pictures of the front and back pcb? After I see pictures, I have more test/measurement suggestions.
-
Can you also replicate my 4 measurements using just 1 set of probes shown in the pictures? Assume my 187 = non working 89IV and my 77 III is the "good" meter.
-
Here are more pcb pictures - part 1
-
part 2
-
Measurements
-
Additional comment:
Measurement picture 2:
If I on the faulty 89IV press "range" and cycle through the manual ranges, it affects the readout on both meters.
Like:
519K - 1.13Vdc
0,12M - 0.38Vdc
2.44K - 4.83Vdc
37.5K - 2.83Vdc
Auto - as picture
It also had a small impact of the current measurement, leading to decreasig current to almost zero.
-
I need a break to digest all this. I'm participating on your 89IV, defective BM-235 and now Velleman 9912 all at the same time. It is all starting to blur together.
I'm also gone for a few hours later this afternoon.
-
Going by your pictures, the input block seems to be very dirty. You might have some contamination within or even under the input block especially since you mentioned it got wet.
I would remove the input block and give it a good cleaning along with the PCB under the block. Or leave the input block connected and submerge it in some IPA and let it soak for a little while. Make sure you let it dry before retesting.
-
Going by your pictures, the input block seems to be very dirty. You might have some contamination within or even under the input block especially since you mentioned it got wet.
Good suggestion. I totally forgot to suggest that. See
http://mrmodemhead.com/blog/gallery/fluke-189-clean-up/ (http://mrmodemhead.com/blog/gallery/fluke-189-clean-up/)
-
The input-block is the part where the water entered the 89, and also the part where it was most wet.. and the only part not totally disassembled..
I'll give it a look..
-
Note the use of paper towels soaked with IPA and then cotton buds to dry.
http://mrmodemhead.com/blog/wp-content/gallery/f189_00/F189_021.JPG (http://mrmodemhead.com/blog/wp-content/gallery/f189_00/F189_021.JPG)
BTW, I'm now reminded that modemhead fixed a 187 that had wonky input jacks. It was a mechanical lug/broken solder problem. Similar to his blog post below
http://mrmodemhead.com/blog/fluke-80-series-jack-assembly-rebuild/ (http://mrmodemhead.com/blog/fluke-80-series-jack-assembly-rebuild/)
I'm not saying your 89IV has wonky input jacks because your VDC seems to work 90% of the time.
-
Free_electron fixed this Fluke 189 wrt to the input jack block.
https://www.eevblog.com/forum/repair/fluke-189-teardown-repair/ (https://www.eevblog.com/forum/repair/fluke-189-teardown-repair/)
-
I can easily get a good reading with another ohm-meter, from within the inputblock to the rear side of the pcb, or to the MOVs or resistors. So I guess it is something else.. Worth a try though..
Thanks for all the inputs anyway :)
-
Maybe worth mentioning... the multimeter was exposed to rain late last summer. It was forgotten outdoor, and quite wet. But after drying, it has worked flawlessly since. At least 8-9 months..
I'm just curious how you dried it? Air dry or did you scrub the entire pcb with 91% IPA? How did you get the water out of the big ICs?
-
I remember it was most water around the input-block, so I guess I was shaking the pcb and then using a hot-air blower to dry that part, and some paper-towels.
No IPA, only air. Can't remember I spent time drying other parts. Maybe it was positioned in such a way most water was in the bottom...
It was left outdoor on a friday, I got it back following monday, can't remember when the rain set in, or what time that other guy found it. But it sure was wet for many hours.
But since it powered-up and started working after some drying, I have just forgotten about the drying process. Only remember I spent some time with the input-block. And I'm quite sure I removed the LCD, and if it was water there I must have used paper and hot air..
:-//
-
Any pictures or results from taking apart and/or cleaning the input block?
-
Additional comment:
Measurement picture 2:
If I on the faulty 89IV press "range" and cycle through the manual ranges, it affects the readout on both meters.
Like:
519K - 1.13Vdc
0,12M - 0.38Vdc
2.44K - 4.83Vdc
37.5K - 2.83Vdc
Auto - as picture
It also had a small impact of the current measurement, leading to decreasig current to almost zero.
Well, at least the readings aren't 0 ohms. The readings, obviously are way off, perhaps due to contamination on the pcb or input block?
-
The 89 IV has a lifetime warranty. Why are you messing about with it?
-
The 89 IV has a lifetime warranty. Why are you messing about with it?
I see a lot of people make this common assumption about Fluke's lifetime warranty. Fluke's "lifetime" is a marketing term. Details in screenshot below.
-
I called Fluke, asking for advice. They said this unt is more than 15 years old. It better to buy a new than sending in for repair. Nobody said anything about warranty.
--
I have not disassembled the inputblock. It seems to be a compact-piece, and has 8 soldering point.
However, when I take another ohm-meter and put one test probe in the inputblock, and the other probe on the soldering-point on the rear, or a component along the input-stage, I get very low ohm reading, so I concluded there's no problem along that signal path...(?!)
-
A lot of those side by side readings are obviously wrong suggesting there might be leakage current somewhere. If it were my meter, I would have already desoldered the input block and cleaned it. I buy abused/used meters and have seen what pcb contamination, however small and not visible to the eye, can do in terms of funny readings. There are a number of cases here where we have helped users fix their multimeters with funny readings just by a good IPA bath.
You have nothing to lose but 15 minutes of your time because this Fluke is not under warranty.
-
That's right. It doesn't work as is, so it can't be any worse.
I'm gonna desolder the input block and give the pcb a good IPA bath.
-
I'm gonna desolder the input block and give the pcb a good IPA bath.
I'm not saying this is going to work, but if we eliminate this as a possibility, then we can move on to other things. If the readings also change that gives us a clue too.
BTW, I still can't shake the feeling those two solder MOV points should be open circuit when the 89IV is powered off. However, I have been fighting a cold for almost 5 weeks so I may not be thinking clearly or I just don't understand how it works in the 89IV. I have no formal electronics education. :-//
-
My 87IV had more or less the same problem: no reading at all.
The black binding post of input block was defective (broken and interrupted)
As this input block is very expensive and I don't use current inputs, I mounted the input block of the wrong way (rotated of 180°) and it work again. (see picture)
-
BTW, I still can't shake the feeling those two solder MOV points should be open circuit when the 89IV is powered off. However, I have been fighting a cold for almost 5 weeks so I may not be thinking clearly or I just don't understand how it works in the 89IV. I have no formal electronics education. :-//
No power attached is not necessarily the same thing as being switched off.