Fluke 77 series ii (i believe) strange demise

[mikeydee77]

Could be the stuff of fairies, but . . .

I bought a Fluke 77 from ebay a few months back.  Despite being cheap it's been working as good as gold and is fairly accurate compared against my Fluke 87 so I thought I had been lucky in my purchase.

However when the battery indicator came on I opened it up prompted by also received a few weird readings to change the 9v battery.  But when I had replaced it, I found that it didn't power on at all.  Then I noticed that the fusible resistor R1 was fried and that there was a scorch mark underneath the pcb right by R1. 

So maybe I had damaged it just before without realising it but it seems strange that it shows no signs of life other than a little peep from the piezo when I connect the battery.

What I was hoping was that I had got something wrong when taking it apart and reassembling it for example alignment of the rotary switch or there is mention of a little spring which is missing from the meter altogether (or flew out when I dismantled I suppose) or something else and maybe it is working but just not the display.

I have checked the two test points TP1 has 8.33v and TP3 has -7.7v. These are supposed to be 3 and -1.5 so are clearly way out.

Any suggestions would be very welcome and appreciated.

[ModemHead]

Scorch marks and a fried R1 means there was a serious over-voltage event. If R1 was the only casualty, it would not prevent the meter from booting up and displaying random counts or OL.

The analog processor chip (U1) is responsible for regulating the analog ground (COM jack) to 3V below the battery positive terminal, aka Vdd, aka TP1.  So if TP1 isn't 3V wrt COM, then it isn't looking good for U1. Also, TP3 is one of the LCD drive signals and should be positive, so that doesn't look good either.

If the PCB is "scorched", then make sure you completely remove any carbonized material.  Carbon conducts quite well.

[lowimpedance]

 What were you measuring when the readings were no longer good ?. Seems odd if R1 was originally damaged and the meter was working good.
Post some pictures, maybe able to spot something mechanical (you should not need much disassembly to just change the battery!!).
Looks like the meter is probably toast though!.

[mikeydee77]

Hi and thanks for reading my post.  I have been away from the computer for a few days hence the absence.

So I was possibly measuring the voltage on a SMPS prior to the meter's demise approx 200v. I say possibly as I have a number of meters and I can't exactly be sure which one.  I just thought ooh better change the battery on that 77 as the battery indicator has been on for a while  .... and as I was putting it back together when I noticed the damage.  I suspect that it was there all along since ebay purchase but somehow R1 was still limping along.

Here are some pics. Notice in particular that:
Next to R1 the CC 1M resistor is a  little char grilled
The foil screen as a scorch mark just where R1 meets it
On the underside of the PCB R1 has pcb burn damage causing a dent in the board.  It looks enough to be conductive to my eyes but I would have thought the meter would just give bad readings rather than dying because of this.

Here are the pics all suggestions welcomed.

[retiredcaps]

What is this (circled in yellow)?

[ModemHead]

That looks like an original 77 (series I) board.

R1 is not original and has been replaced.  If you didn't replace it, then the damage must have been done before you got it.  The scorch mark doesn't look bad but should be cleaned up because it goes from R1 to a ground trace.

Since you mentioned test points I assume you have the service manual. Obviously the power supply is messed up, so check the battery connections to Vdd and Vss of U1 and make sure everything is ship-shape.

If you have a scope, you can check the clock.  If the clock isn't running, absolutely nothing works.  The watch crystal is just above the battery under that little metal shield.  Probe the via connected to the left-hand crystal lead, you should get a 3Vp-p signal at 32.76kHz.  A DMM with frequency function should also pick that up.  Use COM jack for ground.

[mikeydee77]

This is a piece of tinned wire bridging two sections of the circuit.  No idea if it is original or a mod or fix.

[retiredcaps]

No idea if it is original or a mod or fix.

Likely a mod by the previous owner.  I have never seen that in my two series 70 meters.

[mikeydee77]

To modemhead
Yes I replaced R1, the original was frazzled.

I made the assumption that this is a series ii meter without knowing how to identify the different models. I think the original eBay ad may have given me the idea.

I have read through the series ii service manual and made comparisons.  Ill make the suggested checks and post back.

Thanks for your contributions.  I am resigned to the meters demise but would be good to work out if repair is worth pursuing.

I contacted fluke uk and they charge about 30 quid to assess plus repair charge which I guess would wind up being 60 quid to replace a chip or two assuming that they have supplies.

Quite fond of this meter as it reminds me of the start to my working life.

Mike

[ModemHead]

Yes I replaced R1, the original was frazzled.

Drat, I was hoping it was a previous repair not related to this failure. Many times they survive events that blow out R1, but sometimes not.

I made the assumption that this is a series ii meter without knowing how to identify the different models. I think the original eBay ad may have given me the idea.

The fuse configuration on your 77 matches the series I meters.  The 10A input is unfused, the 300mA input has a 3A HRC and a 630mA glass fuse in series.  Later models added a 15A HRC fuse for the 10A input.

I have read through the series ii service manual and made comparisons.  Ill make the suggested checks and post back.

I tried to find the series I manual at Fluke but I'm either blind or they've removed it.  So here's my copy.  The very last page in this manual should match yours I think.
http://mrmodemhead.com/share/manuals/Fluke_70series_smeng0000.pdf

Quite fond of this meter as it reminds me of the start to my working life.

Well, as long as you have the time and inclination, no need to give up until you're sure that U1 is dead.  Even then, there's always the possiblity of finding parts donor.

[lowimpedance]

I tried to find the series I manual at Fluke but I'm either blind or they've removed it. 

No not blind, it appears to have vanished with the list starting at series 2. Flukes manual search is a PITA form what it used to be.
Plenty enough 77's around and a parts donor may be in better shape and the OP's current one which then becomes the spares unit!.

[mikeydee77]

Thanks for the manual.

OK an update since realising that my measurements were not with respect to the common (neg probe) terminal) 

As per 4-16 in the series 1 service manual, with meter on the  Vdc setting:
TP1 measures 3.05v
TP3 measures 2.5v (included although I didn't notice a mention of it in the manual)
The crystal is oscillating at 32768 hz and the sine wave amplitude I measure at 348mV on my meter and on the scope it looks like about 1v p-p which is a little high according to the manual
1v ref voltage looks ok
V Middle is 1.6v
Couple of other checks I made
Battery draw is 426 uA
Also getting a clock in to U2 on pin 23
Also just noticed that the display bezel is cracked
Thinking that the display might have been damaged during disassembly I also tried testing the continuity buzzer but it didn't peep.

Anything else I can check

[ModemHead]

OK, maybe U1 isn't dead yet.  He's getting bettah!  Voltage regulation and a clock is good, since U1 is responsible for both.

The beeper part doesn't sound good though.  See if the Ohms current source works. U1 is responsible for that too.  This one I have here will push 0.5mA from V+ to COM in Ohms mode, a little more in Diode/Continuity mode.

426uA battery draw sounds high.  Mine averages 252uA in DCV, 273uA in ACV.

TP3 seems high.  It should average at about Vmiddle since it swings from 0V to 3V at 50% duty cycle.  Check TP3 and some more of the U2 LCD drive lines with your scope.  You should see nice clean square waves at 64Hz.  If U2 is driving the LCD and nothing is showing up, starting checking the LCD, zebra strips, etc.

[retiredcaps]

426uA battery draw sounds high.  Mine averages 252uA in DCV, 273uA in ACV.

I agree.  I measured my Fluke 75 and it consistent with modemhead's reading above.

[mikeydee77]

The beeper part doesn't sound good though.  See if the Ohms current source works. U1 is responsible for that too.  This one I have here will push 0.5mA from V+ to COM in Ohms mode, a little more in Diode/Continuity mode

I connected a 1k resistor in ohms mode and measured the voltage across; it was zero - is this what you meant for me to try?

426uA battery draw sounds high.  Mine averages 252uA in DCV, 273uA in ACV.

OK - could be good to keep an eye on that.

TP3 seems high.  It should average at about Vmiddle since it swings from 0V to 3V at 50% duty cycle.

Yep it is 2.5, but with a little handling and flexing of the board it sometimes drops to 1.8v - interesting.  Haven't located the cause but could be a dry joint / plate through hole issue - will keep checking on this.  I haven't seen any sign of life when it drops.

Check TP3 and some more of the U2 LCD drive lines with your scope.  You should see nice clean square waves at 64Hz.  If U2 is driving the LCD and nothing is showing up, starting checking the LCD, zebra strips, etc.

ok I have some square waves on U2 Pin 15, 14 and TP3 etc.  Pretty sure my 30 yr old scope is not calibrated but they look like a lower frequency - I measure 14 Hz (70mS) on my old scope and my meter wont pick it up. Although the top half of the wave is clean the 0v has a load of noise on it. The ratio is not 50:50 and they jitter about a bit.

Feels like it could be just a bad joint somewhere given the changing nature.

[Excavatoree]

Feels like it could be just a bad joint somewhere given the changing nature.

Has anyone mentioned that some of the early 70 series models suffered from a soldering defect of U1?  Seems some didn't have enough solder, or maybe the installation was bad or something, but the result is poor contact.  I'm not the worlds greatest smd solderer, but I've fixed a few that had that problem.

I've seen some that were obviously re-worked.  Some have been done very well, others not so much.

This is difficult to test, sometimes I've looked at the pins under the magnifier and seen that they were "hovering."  Sometimes, I can press hard on the IC and have got meters to come to life.  (EDIT:  Of course, that's temporary - just to detect the problem.  The meter stops working when pressure is removed)

Just a thought.  Sorry if someone mentioned it before and I didn't see.

[retiredcaps]

Has anyone mentioned that some of the early 70 series models suffered from a soldering defect of U1?

Just a thought.  Sorry if someone mentioned it before and I didn't see.

Nope.  News to me.  Thanks for sharing.

[ModemHead]

Has anyone mentioned that some of the early 70 series models suffered from a soldering defect of U1? 

I didn't know it was an official defect either, but that explains why I've had to resolder U1 a few times.  In fact I think that's how I fixed this one, because I can see the evidence of my amateurish SMD soldering.

I connected a 1k resistor in ohms mode and measured the voltage across; it was zero - is this what you meant for me to try?

Actually I meant just connect another DMM configured to measure mA straight across the V+ and COM jacks of the 77 and measure how much current it could source.

The LCD drive lines should definitely be clean and 64Hz (3.1 div @ 5ms/div).

None of this sounds good and the hint of an intermittent fault makes it worse.  This is where I take out a popsicle stick and start poking and observe what happens.  Yea, poke it with a stick...