Fluke 87-V stuck on BATT and never turning off

[AngraMelo]

Hey guys,
Ive been trying to troubleshoot my 87-V for some time with no success. I have 2, one working fine and the other displaying BATT sometimes and other times when the wheel is on the "off" position it never turns off, it goes to the uA mode.

I tried everything including cleaning, visual inspection and probing around while comparing to the good 87.

What Ive found is that the reading on the transistor Q4, Q8 and Q17 are lower compared to the good one.

I just realized that there is a schematic for it but I dont understand how those transistors that control the backlight could cause it to never turn off.

MOVs, other transistor and diodes all read normal. The exception to this is a tamtalum capacitor that reads a different resistance but it is not off by so much.

Could anyone help me out?

Here goes the schematic if you dont have it.

http://www.ko4bb.com/getsimple/index.php?id=download&file=Fluke/Fluke_87V_Full_Service_Manual_with_Schematic_and_Electrical_Parts_List.pdf

[shakalnokturn]

Looking at schematic I'd be checking the obvious: Making sure about suspect tantalum, cleaning switch contacts, checking Q13.

[goaty]

It seems the rotary switch is passing a voltage from a voltage divider to select modes. I had failed devices which use the same scheme.
Problem is if one resistor is bad, the whole thing behaves weird.
Check the resistor divider network on page 41 B8/7.
Check the -2.5V, the SWPWR and the TP20.
It should be 3.3V from SWPWR to -2.5V (0.8 + 2.5)

Maby this helps

[AngraMelo]

Thank you for help guys
Ill check all those ideas when I get home tonight.

Also, can anyone see why the buzzer and backlight circuits malfunctioning could cause the "off" position to short to the uA position?

[AngraMelo]

I need to check if the BKLT U2 (MCU) either goes high or low to activate the back light. given that the emitter of Q4 and Q8 are tied to -2.5V, it could activate the back light either way. Right?

[goaty]

BKLT needs to go high (0,8V) to switch backlight on,
HIBEAM needs to go to 0.8V if high brightness is on.
The Power is BTP, which is always Battery (9V) and always on.
So maby a short in Q4/7/17 cause the power to be backfeed into the CPU even when its of. Check for shorts on those transistors.

[shakalnokturn]

Do you have -2.5V between logic ground and battery minus?
Having less difference could explain your low measurements on the backlight switching.

[AngraMelo]

I will check all of those tonight.
I did not have time last night because I had a order to fulfill but now it is over and ill have time later

[AngraMelo]

Ok so I did the test and we made progress!


Transistor Q4, Q8 and Q17 of the pins that are supposed to have any voltage on them all read around 40uV but on the good meter measure around 400uV. I know it is a small voltage but the difference is an order of magnitude more.

Test point 22 that is supposed to be at 1.2V if the battery is running low, is always at 1.2V on the bad meter, even when on the off position.
The resistors that form the voltage divider on TP22 read fine. So do the capacitors c13 and c28.

Now, as the TP22 is only a voltage divider and it does not appear to be connected to anything else besides de MCU and the series P-mosfet, what I think is happening is the following: (please let me know if that is possible)
The mosfet is somehow damaged and the on resistance has increased thus dropping more voltage thus bringing TP22 to 1.2V and triggering the BATT alert.

I measured the drain to source voltage and when the meter is on I measure 7V.


Now, I changed Q14 and the meter now turns off. BUTT!
It displays BATT for 10 seconds, then some gibberish on the screen for a second then turns off.
When powered off the drain to source voltage goes to 2V.

I ordered a replacement for that mosfet and Im hopping this will be the end of it.

Can anyone verify if my idea of what went wrong is possible?

[shakalnokturn]

Reverse polarity (what shape is CR3 in?), ESD, an attempt to measure it's own battery... Or sometimes components just fail without help.

[goaty]

So it is probably Q13.
For safety maby also measure the 1Meg resistors R73, 74, 80 ?

Also measure Battery (BTP - BTM) current and voltage is ok.
Because if it is on, you have measured 7V across Q13, but battery is only 9V, this is strange as you would need around 6V on the voltage divider to get 1.2V on TP22.

[Ordinaryman1971]

Did you take the meter apart, did you take off the range switch, did you put it back correctly? To me it looks like your switch is not assembled the way it should be.
I mean... that would be my suspicion. Please compare the way it's put together with the other meter you have.

[AngraMelo]

Ive done that dozens of times. Swapped all the swap-able parts with a known good 87-V I have here. Cleaned multiple times. It is not mechanical in any shape or form. As I mentioned above I change the transistor that pulls the gate of the P-Mosfet low so the meter turns on. Now the meter truns off but does some weird stuff on the screen before turning off. Also, BATT still shows up sometimes. The drop on the Mosfet is varying so I ordered a replacement.

[AngraMelo]

Guys!!!!

It is fixed! It was the P-Mofest!

For the record:

Q13 and Q14 were both changed.

[goaty]

Congratulations ! 

A good meter saved.

[AngraMelo]

Congratulations ! 

A good meter saved.

Thanks buddy! I also checked the resistors you mentioned. Everything is fine!
Thank you and everybody else for the help!
You guys rock!