Electronics > Repair
Fluke 8600A - carbon resistors and tantalum caps, should they be changed?
(1/3) > >>
kblue:
Hello there!

I've got another Fluke 8600A with a bit NiCd (Battery Pack Option) corrosion damage. Nothing to worry about (i hope). The pack is already removed and while i'm working on this unit, i wonder ... should i change the carbon resistors? Most of them read 15% high (with gold ring it should be 5%) or leave them be? Same question for the "ticks" (blue and orange tantalum caps). Some of them read "okay'ish" some have high'ish ESR compared to "modern new caps".

The unit seems to "work" so far, but not quite right due to bad connections - already working on it.

Best wishes and thanks in advance!

/edit: Everything runs fine now ... tested some functions.
5 VDC (Ref) = 5,000
5 VAC (Ref) = 5,000
1 mADC = 1,0000
1 mAAC = 1,0003
100 Ohm precision resistor = 100,00 Ohm

Only odd thing is, fresh from start every "dot" between numbers (except the one used for the actual range) is blinking. So its like ,5,0,0,0, Over time, while the unit is running, the dots not used for the selected range slooooooooowly fade away but never gone completely. Gives me the feeling of "bad cap" due to "things fading while something is running". Dunno.
blue_lateral:
If it were me, since it is a piece of test gear, I would replace ANYTHING that is way out of spec like that. Are you sure? I don't think I have ever run into any single piece of equipment that had that many carbon resistors that far off. Since it's test gear, use something more stable.

I would be more worried about the battery contamination on the board and especially in the switches. A little deoxit in the switches completely disabled one of my flukes years ago thanks to the high Z. It was conductive enough to be a big deal and it was hell getting it back out. It took lots of isopropyl.

I suspect others will disagree, but it wouldn't take much to convince me to shotgun a bunch of tantalums in something that old. I suspect tantalums were poorly understood in years past. Engineers believed the datasheet and sometimes it did not go well. They fail short. If you aren't having that problem you can consider leaving them alone. If you replace, it is a good idea use higher working voltage.

It's probably going to need calibration after a rebuild that invasive.

TERRA Operative:
A 4.5 digit multimeter? I'd just replace the resistors with good quality metal films. It'll make the meter much more temperature stable for one.

There is a chance that the temperature coefficient of some resistors was used to counteract the temperature coefficient of some other parts, but for 4.5 digits, it's not that critical and carbon composite resistors vary wildly in their tempco characteristics.


As for the tantalums, if they aren't shorted, just leave them be, unless you really wanted to uprate their voltage ratings.


I say go for it with replacing the resistors then recalibrate the instrument. :)
Kleinstein:
The carbon resistors should be in the non critical parts, like setting the display brightness. The actual result should not depend much on carbon resistors.
So one could leave the resistors if they don't do harm if the drift further. At the supply regulation better replace them.

For the tantalums, there is a difference between metal case ones and drop types. The wet types (metal case) are usually very reliable and no need to change them, unless there is corrosion damage.
The plastic drop ones may fail short and if used with a powerfull (> 1-2A) supply they should be changed, as than a nasty failure mode with fire is possible.
For the replacement "normal" low ESR capacitors can often be used, unless the meter also needs to operate below freeezing.
kblue:
The carbon resistors are somewhere in the power line, the battery power board and stuff, true. So far, they don't do harm - just reading 10-15% or more high (22 ohms are 30 ohms and so on). Thanks for the reply, tips and tricks!

So far everything works fine, except for the decimal dots between numbers. With fresh start all of them are lit and while the unit is running, the "not used dots" fade sloooowly away but will never be off completely. The 8600 is just powered from my bench power supply and by accident, i noticed without the "battery power board" (missed to plug it back in) the decimal dots are as they should be while selecting ranges. But the instrument wont work correctly without it. Putting the board back in, the instrument works fine except for the decimal dots.

Also this "battery power board" while powered from my bench psu via battery terminals is making somewhat of a high pitch noise (exactly 9000 Hz and 18000 Hz together).

I could live with the "decimal dot problem" no worries, its just a bit odd. While fading, they do "pulse" with the actual measuring.

Best wishes!

/edit: The high pitch noise is comming from "T2" of the "battery power supply board". In the documents T2 is a "xfmr inverter".
/edit2: In the original schematics, i see U202 and U203 voltage regulators. My battery power board has a single chip solution with a RC4195N. Heating RC4195N up with some hot air, makes the wrongly active decimal dots fade away faster and cooling it down, makes them brighter again. I wonder, if i can replace this part with "usual regulators" for + / - 15 V
/edit3: Checked the +15 and -15 VDC from the RC4195N ... ripple @+15V is about 180 mVpp @-15V its about 500 mVpp. Goes down while heating up with a bit of hot air.
/edit4: Checked the RC4195N out of circuit and it seems fine (but this was just a simple test without load and stuff)
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod