Fluke 8010A wacky readings - Diagnoses

[Jeff_Birt]

I picked up a Fluke 8010A a few weeks ago and besides cleaning it all I had the chance to do was power it up and short the test leads on the ohms setting. Tonight I sat down to see how it worked and if the calibration was ok.

The ohms function 'sort of' worked, by that I mean that I got readings that were off but in the right ball park. The volts and current measurements were very screwy. I would get silly numbers that mostly would just keep counting up, like something was charging/discharging/leaking.

An investigation of the board revealed that the input protection varistors VR1, 2, 3 looked a bit odd and the soldering looked odd. I tried to ohm them out thinking I should at least get consistent high readings between the three of them, I did not they were markedly different. These three varistors bias the base of a bipolar transistor but it checked out OK (with a simple diode test). I pulled each end of the varistor chain up and then saw the discolored are under them. Without the leaky varistors in the circuit all meters functions worked! I did not check the calibration yet.

The part number was in the 8010A manual and Mouser has then in stock for $0.86 each. I'll add them to my next order and then check the calibration.

Next up on the bench is an 8050A.

[retiredcaps]

Good job finding the bad varistors.  They should read open circuit with a resistance check.

The grey rectangular PTC, just north of the red varistors, looks like it has a chunk missing the right hand corner??

Without consulting the manual, I think it should read about 1.1k ohm?

[Jeff_Birt]

Good catch in the PTC. The manual lists it as 1K and it measured about 1K at room temperature. I did check all the resistor in the input section (which are all 1K oddly enough) and all seemed OK. I was not sure if the chip was mechanical damage or it was damaged along with the PTC. I'm looking at the same part on an 8050A and is not chipped and the 8050A manual shows it as the same 1K part. I do have an 8050A with a dead display I could rob it from if need be. I did a quick search on Mouser and did not find a 1K PTC that even came close to the original parts 2W rating.

[retiredcaps]

The manual lists it as 1K and it measured about 1K at room temperature.

It is likely good then.  I had a bad PTC in a Fluke 77 III.  It read 100k ohm when it should have been 1.1k ohm.

[Jeff_Birt]

Your concern about the PTC was quite prescient. I looked at the 8050A this morning and its PTC was reading 750ohms instead of 1K. I'll do a separate post on that unit as it had multiple problems. I stole the varistors out of the parts 8050A I have to fix this 8010A and now it should be ready for calibration.

[retiredcaps]

I looked at the 8050A this morning and its PTC was reading 750ohms instead of 1K.

I participate in a lot of "help me fix my meter" threads and measuring the input protection components, MOV, PTC, fusible resistor, power resistor, etc only takes 5 minutes or less to check.

On my used bought from ebay Fluke 77 III, I initially did not measure the PTC and it wasn't until 3 months after that I discovered the 77 III resistance readings was 2% off at 100 ohms and 10M ohms.  Only then did I discover the PTC read 100k ohm.  I replaced it and now the meter is within specifications at those ranges.

[retiredcaps]

I did a quick search on Mouser and did not find a 1K PTC that even came close to the original parts 2W rating.

In searching for a PTC on newer Flukes, the closest that I found was this YS4020.  However, no one seems to stock it.

https://www.cdiweb.com/datasheets/ge_thermometrics/ys4019.pdf

Recently, I mentioned the YS4020 here.

https://www.eevblog.com/forum/repair/brymen-bm869-meets-high-voltage-in-real-world/

[Jeff_Birt]

I do appreciate you help. The 8010A seems to work correctly in all modes except AC volts. It will not zero when the leads are shorted. When I measure 120V I get a reading in the right ball park but when the leads are removed the display slow goes from 123V down to 50ish volts. All ranges have similar results. I'm guessing this points to the RMS module which requires some more manual reading

[retiredcaps]

The 8010A seems to work correctly in all modes except AC volts.

I rarely need to measure ACV and I don't have a 8010A to tell you if that is the correct behaviour.

I know some of the gang switches can be problematic in these push style Fluke bench and handheld multimeters. Maybe give the ACV button and its switches good inspection?

Post a clear focused picture of the ACV/RMS section as well.

[Jeff_Birt]

I played with the switches and the AC/DC switch did not really snap back out crisply. I cleaned all the switches with alcohol (98% stuff) and then put a few drops of WD40 in each one. I know some folks detest WD40 but used sparingly in an application such as this it will work fine. I believe some equipment manufacturers actually recommend it as a switch lubricant.

Now all the switches work nicely and I have a decimal point! Looking at the schematics I can see that the decimal point is controlled directly by the switches so this makes sense. Measurements in all voltage ranges except the 20V range seem OK now.

The 20V range problem is an odd one. I used a jumper wire to short the volt/k input to common. If I measure at R2, the first resistor (1K) on the input line I see 0V. If I measure at TP4, this is the 1M resistor (R6) just before the A/D input I will see anything form a few mv to a few tens of mv (it varies with time), the LCD on the 8010A reflects the readings I am taking (this tells me the A/D, etc. is OK). On any other range I get 0V at TP4 which is what one would expect.

I pulled up one leg of R6 and confirmed that the stray voltage is coming from the input section and not the A/D. With the input shorted there should be no voltage source to the input voltage divider circuitry. Since the problem is only present on the 20V range (switch S10) my suspicion is that there is something in the switch itself, perhaps some small bit of metal, that is causing a partial short and the phantom mV. BTW, I also measure the phantom mV at pin 2 of U1 (the ceramic voltage divider) but not after RT1.

The switches have their individual contacts labeled on the schematic but I'm not certain at this point which end of the switch is A, I suppose that this could be deduced by tracing what each switch contact is wired to. I'll probably need to get some switch/contact cleaner in an aerosol can so I can really blast S10 from the rear opening to try and flush out what is in there.

Fixing this meter is a fun learning experience, that is for sure

[retiredcaps]

Fixing this meter is a fun learning experience, that is for sure

Yes, that is why I do it myself and help others.

Modemhead's blog has an entry about cleaning

http://mrmodemhead.com/blog/dirty-multimeter-clean-up/

Those gang switches can be tough to clean.  I recall one post about using contact cleaner and then leaving the device upside down?

[ModemHead]

You've probably already found it, but the manual has a procedure for the RMS converter DC offset adjustment.  On a 3-1/2 digit TRMS meter it would be normal to have a few residual counts in the ACV ranges, but 50ish is definitely a problem.  I hope there's nothing wrong with the converter.

It sounds like you've determined that stray current is leaking into the front end. Don't forget to clean, clean, clean the PCB itself with IPA.  It only takes a few nano-amps of leakage to cause trouble, so surface contaminants can easily do it.  And I see some heat damage under the varistors.  Carbon is conductive...

Everyone has their own ideas about cleaning those switches.  Spray, don't spray?  Whatever works, I guess.  I've never had any problem with the switches on old Flukes, but I do have an old HP 3466 with similar switches that were flaky.  As I recall, I sprayed them with Deoxit 5%, and then compressed air. Whenever spraying anything like this in meters and scopes, I try to cover the intended target with a paper towel to control over-spray.

For the switches, there is usually a detail diagram in the manual. I can't spot one in the 8010A manual, so here's one from the 8060A manual.

[Jeff_Birt]

Thanks, I could not find the switch detail in the manual but was able to work out what was what last night. The image you posted is much clearer than my scribbles on the print out of the component locations You are right the board is discolored under the varistors but I did ohm out across the pads when I pulled them out and even on the 20MegOhm scale had an open circuit.

In the course of determining the switch contact layout I found there was more than 2V at S11A which was very odd. I tested S11B and found the same voltage (actually about 2.72V, VDD). Hmmmm......I desoldered both the S11A and S11B pins and made sure the pins were not making contact with the pads. This got rid of most of the phantom voltage, there was a few mV floating around. There is a bottom side trace that goes between the rows of S11 and comes through a via right at the A&B contact area and it looks like it then branches out to both sideways in both directions. I think this is the VDD that goes up to the range switches for the decimal point; S11B is one of the decimal point switches which is where the 2.72V was coming from.

At this point it was almost midnight and I was pretty tired so I called it a night.

[Jeff_Birt]

I picked up some CRC Electronics Parts Cleaner from a local auto parts store this week. If you live in a smaller town like me, there are no local electronics stores anymore. Our local radio shack pretty much just sells cell phones now. Anyhow, I looked up the MSDS on the CRC product and a similar product by MG Chemicals and they are essentially the same thing. Just be sure that what you get is safe for plastics. There are harsher 'electric parts' cleaners which will damage clear plastics.

I first soldered the pins on S11 back up and then cleaned the switches on the 8010A with a liberal application of the electronic parts cleaner and operated each switch several times. To make the cleaning process easier the white low current lead was desoldered from the baord. Trying to snake the spring out of the front panel is a PITA. I was busy for a few days so the board just sat and I just got back to it this morning.

A quick test was done with the input shorted on the volt function, testing in all ranges by measuring TP4. The phantom voltage was gone. The meter was put back together and seems to function fine. I will take it to work and calibrate it and the 8050A.

The moral of the story is that if you are getting strange readings and/or the decimal point does not show up clean the switches thoroughly. I cleaned them twice with 98% alcohol and there was improvement but it still was not 100%. I suspect that the addition of the force of the blast from the electronic parts cleaner managed to dislodge whatever piece of crud that was in the switch causing problems.