Extech MM570 (aka Brymen 859, non-s) with bad Diode Check

[ogdento]

When I set my Extech MM570 to Diode Check mode, which shares a rotary switch position with Capacitance, it shows .407 instead of OL with nothing connected to the com/V inputs.  When measuring a known-good rectifier I get .0073

All other functions of the meter check out fine... AC/DC volts/millivolts, Ohms, Capacitance, Frequency, Current - and the readings are bang-on when compared to my Fluke 87

Background Info:
The MM570 a version of the Brymen 859, and is similar to the Brymen 857 (different rms chip/fuses).  There's a Q16 NPN transistor just above the rotary switch that's in the path when you select the Diode/Cap range.  It seems like it shorts it's collector to ground when turned on, but I'm not sure if it's turned on for Diode or Cap mode. (I don't really want to pull it off until I'm more confident about its role).

Here's a pic of my 570 with the shield removed and some traces for Diode Check mode drawn in:



**I checked out some other Brymen pics from nanofrog, mjlorton, grapsus, and bdunham7... The Extech A, the Amprobe 140-A/160-A, and the Brymen S models do not seem to have Q16... or at least it has a different location/id.  I thought it might only be on boards that have Diode/Cap on the same rotary position (and are toggled with the select button)

Other tid-bits:
1. when set to diode mode, the meter shows .407 and draws 5.6mA from the battery with nothing connected to the inputs... the Diode check current on this unit happens to be .4mA, which seems coincidental at least?
2. when powered off and set to diode mode, the resistance from com to V (checked with another meter) is about 4.2M
3. the input impedance measured for DC volts is 10-11M depending on the range (seems right)
4. all the MOVs and PTCs check out, the big blue input cap checks out (i would expect other issues if any of these were bad though)
5. continuity through the range switch assembly seems to be ok

bdunham7 posted a datasheet for the FS970X that might be the basis of Brymen's chip... much of the reference design matches the mm570 but unfortunately the Diode check does not 

Anybody got any ideas?  or is anybody with a similar meter with Q16 willing to check the current draw and/or resistance from com to V when in Diode mode?

[edited:  updated photo showing path through C15]

[ogdento]

The other pic:

[mzacharias]

You've got something leaky in there. Like an MOV or the like associated with that function or possibly all functions associated with the AC, DC, resistance, etc input jack.

I suspect the input resistance of the meter itself on these functions will also be low for this reason.

You can check that with another multimeter to the Extech's inputs (while the Extech is in DC volts mode)

[ogdento]

Hey mzacharias thanks for the reply!

I checked the DC input impedance early on... for input impedance I get 11.2M on the lowest range, then 10.18M, 10.10M and 10.08M as I continue hitting the range button.  It's in line with my Fluke 87 and another Brymen 257.  It seems strange that all functions look perfect except diode. 

I also checked the input clamps and they look ok - there's a 0-ohm link for each pair of transistors that I'll temporarily remove to eliminate them as a potential problem.  in joesmith's BM786 part 4 video he blasted his clamps and broke the diode function (but also all the other functions using V input)

here are some pics of the volt and ohms readings (the resistors I used are 2.4k and 49.995k 0.01%)

[mzacharias]

I still think something's leaky in there.

I'd pull the battery on the Extech, open it up and check components such at MOV's, PTC's, etc with the other meter.

[ogdento]

Hey thanks mzacharias,

I did check the movs and ptcs... the mov's are open and both ptcs are 1.2k.  I pulled out one leg of each mov just to be sure. 

Also pulled c1 (big 0.01uf input cap) and c15 (27nf cap at top-right of precision resistor and they both measured fine), re-assembled the unit but diode check still fails.

I checked Q16 (npn) in-circuit and it didn't seem right so I'll pull it next and test it off the board.  I hate taking these tiny transistors off.

[ogdento]

I pulled Q16 and it checked out fine, but I think that was a red-herring...

While re-reading the FS970X-DS-44 doc that @bdunham7 posted (the chip is very close to the btc version) I finally noticed diode check also connects to HUNK/pin20 - or the hundred k input from the precision resistor.  I figured that input path might be the problem, so I tried to measure a 500k resistor and it couldn't read it!

100k and 1M work fine, so the problem seems isolated to the HUNK input.  There are 2 variable caps under the shield on the mm570, and there is a fixed and variable cap on each of the HUNK and ONEM inputs in the FS970x datasheet.  I'll check those out tomorrow

[ogdento]

Been a minute since I had any time to debug this guy, but I haven't found anything bad in the HUNK input path yet.

I did notice the upper black discs from the rotary switch mechanism is cracked (around 1 o'clock in the photo below)... I don't think this is my problem because the switch has continuity through it in diode mode, but I really didn't expect this part would break.

The mm570 is an old meter, but the rotary switch looks just like what Brymen are still using in their current models.