-
Fluke 117 "fatal" error message
Posted by
najrao
on 10 Sep, 2017 08:50
-
I was given an almost unused 117. It reports "F116/Err" after power up, and reverts to "bAtt". A new battery, indeed a bench power supply up to 10V, did not make a difference. The Fluke manual simply calls for return to have it 'serviced'. And the local service shop authorized by Fluke want an arm and a leg.
To start with, the instrument was in beautiful condition, even the screen protector intact. There was the slightest of tarnish at two switch contact pads, and at another location adjacent to IC marked LTFLK2. I have cleaned these out meticulously; and dried out the board at 50-60C for some half-hour. The display shows all characters at power up as should. All the discretes test out ok for basic checks on-site. The Vcc and Vref are 3.06 and 2.48V, and believed correct.
I appeal to members to help in any way possible.
Janardan, Bangalore.
-
-
A new battery, indeed a bench power supply up to 10V, did not make a difference.
What is the current draw when hooked up to a 9V power supply?
There was the slightest of tarnish at two switch contact pads, and at another location adjacent to IC marked LTFLK2.
Let's see a clear focused picture of the two areas that you are referencing above. Also, post pictures of both sides of the pcb with the rotary switch removed.
Can you also post a short video of the power up sequence so we can see the error messages as it boots?
-
#2 Reply
Posted by
Toasty
on 10 Sep, 2017 19:37
-
Dave (EEVblog) did a review of this on YouTube in 2010:
https://youtu.be/3k9nVg03Wi4I would grab the datasheets for
U3 &
U5 and see if they are voltage regulators.
Also, the error could be thrown because the voltages are not appearing at the microcontroller in the correct sequence-
Per the datasheet for the
TI M430F448 (100 pin package) pages 20-22:
http://www.ti.com/lit/ds/symlink/msp430f448.pdfAVcc pins 100(+) & 98(-)
must not appear before: DVcc
1 pins 1(+) & 99(-) or DVcc
2 pins 60(+) & 61(-)
The pictures that
retiredcaps requested would be of great help.
T
-
#3 Reply
Posted by
najrao
on 11 Sep, 2017 07:55
-
Thank you both.
I am attaching two pics as asked. The video did not compress enough. Direct gmail may be possible; will send to retiredcaps if you permit.
U3 is the 5V reg, and outputs OK. U5 is a high input impedance amp for Volt-alert detection; it turns on the ground connection of the beeper IC through a transistor. Not likely to defeat correct booting.
I had seen the M430F448 datasheet., and suspected the brownout function. Close observation of pcb tracks will reveal that pins 1 and 100 are shorted together, as are 98 and 99. Though not directly visible, pin 60 (DVss) is connected to 98+99, as is 61 to 1+100 (continuity check). The datasheet recommends feeding both from the same source.
The green Kevlar wire in the pic is my doing, in an attempt to shunt a poor plated through connection at the bottom end. Feeds Vcc to the beeper IC , innocuous.
LTFLK2 does not decode, so U1 remains a mystery.
The current draw is 780uA on start up, and drops to 125 once the display reverts to bAtt.
Janardan.
-
#4 Reply
Posted by
najrao
on 11 Sep, 2017 11:51
-
Correction: Pin 60 connects to 1+100, not 61. Pin 61 to 98+99. Sorry for error.
Datasheet does say analog power must not apply before digital. Conflicts somewhat with the datasheet's own recommendation of using the same supply for both. Still, the two can power up at most at the same exact instant, and not one before the other! The design should have considered a margin, and added circuitry to delay the onset of analog.
Both ana and digi have their own bypass caps, close to pins.
I cannot SIGHT the trackwork connecting the ana and digi, and have only checked by continuity; hope a shorted component does not lurk between, hidden somewhere not obvious. Pls can you check continuity on another 117? Also, don't know what sources their feed, does U1 come in here?
I have a large collection of multimeters, and restored several just for the heck of it. This 117 is nagging me to distraction.
-
#5 Reply
Posted by
kasone
on 11 Sep, 2017 11:55
-
My Fluke 117 displays F117 upon power up and a friends Fluke 115 displays F115. I would suggest maybe a corruption of the firmware loaded into the processor? I have attached a Fluke calibration manual for this series of meters that may be of some help.
-
#6 Reply
Posted by
najrao
on 13 Sep, 2017 05:18
-
Thanks, Jason. I already have the English version. Unfortunately it offers no help.
Because I can't upload the boot video, I shall describe what happens:
1. Most times, bAtt shows directly upon power up.
2. Holding pressed the HOLD button and turning on shows, first all characters for under 2s., and then "1.02". Releasing the HOLD button reverts to bAtt.
3.Holding Min/Max similarly results in "bEEp" first, and releasing gives " - - - -, and bar graph" for an instant before going bAtt. This is not consistent, going bAtt directly some times.
4. Holding RANGE, the YELLOW button, and the BACKLIGHT produce the same results as 3., except that bEEp is replaced by "LCAP", "PoFF", and "LoFF" respectively.
Note that the "F-116/Err" did not report in this message; today, it did show but once. Earlier, I have seen it alternating F-116 and Err ad infinitum. The F-116 itself is sometimes "F-11-" instead.
Any video would at best show part of this description.
May be this post leads to something?
-
#7 Reply
Posted by
Toasty
on 13 Sep, 2017 05:48
-
Can you get into calibration mode?
Can you do a password reset?
Try applying a slight pressure on different parts of the board when powering up. Try a slight twisting action also. I'm wondering if there are bad vias...? You have a bunch of micro ones there to check for continuity from one side to the other. At least this doesn't appear to be a multilayer board...
T
-
#8 Reply
Posted by
najrao
on 13 Sep, 2017 13:28
-
No, can't get to first base for calibration mode or password reset.
Will try pressure at sweet spot, and/or deforming the board and get back tomorrow.
Thanks for yr interest.
-
#9 Reply
Posted by
najrao
on 14 Sep, 2017 03:54
-
No sweet spot, no response to pressure or twist.
The board has at least THREE layers, checking out the myriad via's is well nigh impossible.
Need more ideas.
-
#10 Reply
Posted by
Toasty
on 14 Sep, 2017 07:52
-
Well then, perhaps the M430F448 got its code scrambled a bit or there's a fault in the chip that even a re-flash won't fix.
I'd get a hold of Fluke support directly, not through a service station.
>>checking out the myriad via's is well nigh impossible.<< And I'll bet there's a bunch under the µP you can't get to.
T
-
#11 Reply
Posted by
najrao
on 14 Sep, 2017 08:45
-
I have discovered that shorting S7 on the board BEFORE power up SOMETIMES and only sometimes gets me to CAL mode, only to put up "CAL"/"Err" alternately for ever and anon. Password entry is not possible. Again SOMETIMES, shorting S7 AFTER power up starts a never ending beep beep beep, which stops if the S7 short is removed.
Thanks for your interest.
-
#12 Reply
Posted by
najrao
on 15 Sep, 2017 08:29
-
Now it is stuck on CAL/Err at all times! No bAtt or F11-. Very funny.
-
#13 Reply
Posted by
najrao
on 11 May, 2021 15:43
-
Refers back to Sep 2017, pls forgive me.
My F117 remains bricked, I fished it out to see again.
New findings:
1. Input terminal shows 18 to 20M* in all V positions, not 10 or 11. Auto-V Lo-Z shows 3.2K. Protection elements test OK.
2. Ohms, Continuity, and Diode modes show zero V at input terminal. Any of my other DMM's show a source voltage, e.g., F112 puts out 2.5V for the diode position, and 650mV for Ohms (ranging megohms).
There is but one dc reg ic on the board (TPS77050), and it puts out 5V. I do not know how the resistance modes work, nor how the 3.15V for the MSP is derived.
Surely, my wise friends should make something of this!
*>80M , i.e., OL when checking later; not consistent