Fluke 115 -- bad AC Amps function

[TrickyT]

I have a Fluke 115 DMM purchased in 2011 where the AC current function has stopped working. The fuse tests good, plus the DC current function, which also depends on the same 10A fuse, works just fine. On the AC Amps position if I toggle to the Hz measurement function, it reads 0 when it should be reading 60. All other measurement functions appear to be working fine, including AC volts and Hz measurement option for AC volts. The board shows no signs of damage, including the gold traces under the function switch and the spring contacts on the switch. No components on the board appear to be damaged. There's no evidence of cold or cracked solder joints. There's no sensitivity to vibration. 9v battery is new. In summary, it's only when the function switch is in the AC Amps position that there's a problem.

I have searched far and wide for a schematic but could not find one. The PC board has at least three layers, so it is difficult to reverse engineer the circuit. Fuke doesn't service this model, but does offer a "Repair by Replacement" program for US$190, which is 85% the price of a new one from Amazon.

Does anyone have a suggestion on how I can track down the problem? (A circuit schematics would be wonderful!) Thanks for reading this far.

[mqsaharan]

Hi TrickyT,
I don't have this meter but if ACV *and DCI functions are working fine then the only thing that comes to mind is some problem with the rotary switch. Disassemble the switch, clean it, examine it under a magnifying glass for any cracks in the traces. And try to determine from switch contact traces/vias and the continuity function of another meter if the contacts are getting shorted that are supposed to when the switch is in AC current function.

I hope it helps.

Edit: * and DCI

[TrickyT]

I spent more than an hour looking at the board under a microscope but could not find anything obviously wrong or even suspicious. It's virtually impossible to trace out where circuit traces terminate since they often connect to the middle layer of the board in more than one place and you can't follow what you can't see. I got my hands on the schematic for a Fluke 17B and a Fluke 27 and looked at them, but they were different enough from the 115 to be of much help. At this point I'm afraid I've run out of ideas. 

[TrickyT]

One last post before I move on to other things in my life. The input protection circuitry on the model 115 closely follows the well-proven approach used on many other Fluke meters. Dave has covered this circuitry in detail in at least two different EEVblog videos. I've checked for problems in this area of my meter and everything seems just fine. This is not surprising since DC amps seems fine in my meter but AC amps is dead and they use common protection circuitry.

The 110 series (114, 115, 117) appears to relegate all of the hard work for virtually every major meter function to the TI MSP430F448 processor, and if something has failed internally in it there's nothing I can do about it. There is a custom Fluke chip (an ASIC?) along with a bunch of discrete components (including 2 large capacitors) in one corner of the board that so far I haven't figured out what it does. In his 117 tear-down video Dave thought this circuitry was for the NCV function but I don't think that's correct since the 115 has no NCV functionality. But even if this circuitry was associated with RMS readings (similar to what the AD8436 does in the 121GW), that doesn't explain why AC volts works but AC amps does not.

I think my problem comes down to something internal to the MSP430F448 processor. Hence, it's not repairable. 

And so I'm moving on. I've stuck a label on the meter about AC Amps being dead. I've got some other meters I can use when I need to measure AC Amps: a 1980's vintage Fluke 75, a Uni-T UT210-E clamp-on, and my trusty 121GW. But this Fluke 115 got used a lot. I'm really sorry to see it die.

[mqsaharan]

All right. Please keep in mind that I don't have this meter. So, some of the things I am writing here are guess work.
Now let me try to explain what I wrote in my previous post.

The input protection circuitry does not play any role in current measurement. The only protection you have as a user while measuring current is the fuse *and a bridge rectifier or a few diodes downstream for the meter.

Both DC and AC current measurement use the same signal path and same input to the ASIC, U1. In AC current measurement, the signal also goes through AC converter or TRMS chip connected to the ASIC. The same AC converter or TRMS chip is used when measuring ACV.

Now if your meter is measuring ACV correctly, it means that the AC converter is all right. And if your meter is measuring DCA correctly, it means that the current signal input to the ASIC is also all right.

The only conclusion that I could draw from this is that when you move the rotary switch to ACA, the meter is not recognizing that. Means there is a possibility that the switch contacts are either bad or there is some breakage in the tracks. No matter how remote but the possibility is there. That is why I wrote the following sentence in my previous post

"And try to determine from switch contact traces/vias and the continuity function of another meter if the contacts are getting shorted that are supposed to when the switch is in AC current function."

In order to do that, move the rotary switch to the ACA function and carefully disassemble the switch and note the position of the switch contacts. There will be at least one set leading up to ASIC and another leading up to microprocessor. Then trace those contacts to the vias or solder pads. Assemble the meter again. Check with continuity function of another meter that those switch contacts are shorted when you move the rotary switch in ACA position.
And if there are any components between those specific tracks and ASIC and microprocessor, check those too.

May be I am completely wrong here and problem is somewhere else.

Eidt: *

[TrickyT]

mqsaharan, Sorry, I should have addressed some of your points explicitly in my previous reply. I did try to to trace circuit paths using a second multimeter. As far as I can tell the input to the processor chip from the current shunt resistor is direct and does not involve the function switch. (Note that on the model 115 there is only a single 10A current measurement jack -- no mA or uA.) When I move the function switch to the ACA position the display shows "Lead" for a second or two to remind the user to switch the test lead to the current jack. Then the display shows "A" and "AC" indicating that it's in ACA mode. If I move the function switch the the DCA position the display then reads "A" and "DC." Lastly, it appears that in the ACA and DCA switch positions only two sets of contacts on the function switch make contact with the gold traces. I believe one set is for meter power and the other to tell the processor which mode to use. So I really do think the function switch is performing just fine.

[mqsaharan]

Well, then I think you are right that the microprocessor could be faulty.
It is interesting that it shows/flashes for a second "Leads" when you rotate the switch to ACA function because it does not have such feature built in as per its manual.
Anyway, I am sorry that I couldn't be of much help.