[FIXED] MetraHit Pro leaked Battery repair attempt

[Dachpappe]

Hello everyone,

I got (actually bought) a dead MetraHit Pro Multimeter. The unit was offered as defective - leaked battery.
Yesterday it arrived in the mail and I had a look inside. To be honest, it was quiet worse than expected. The engineers at Gossen had placed the battery holders in the X-Tra Series multimeters directly on the main board, without any barrier to the battery compartment.
End of the story, I now have a dead multimeter with quite a bit of corrosion inside. The interesting part (for me) is, that the path of corrosion is easily traceable but even parts of the device without corrosion are kind of oily. Has somebody seen this before?

PS: I'm sorry for the bad angles the before-cleaning pictures were taken at.
Some pictures of the board after the first cleaning attempt will follow.

Regards, Dachpappe

[Dachpappe]

Hello again,
here are some pictures after cleaning the board with a toothbrush, warm water and later rubbing alcohol for a while.
At first sight the traces do look intact, but some components are quite damaged.
The SO-8 part next to the terminal is a Fairchild FDS6894A logic level dual mosfet, the TSSOP20 directly to the MSP430 is a HCT573 latch. Both should be easily replaceable. But there are 3 identical chips I am not able to identify. They are labelled as "CT2 16Z". Can anybody help me out with the identification?

And as always, here are some photos for you:

[retiredcaps]

Can't help id the part, but in your last photo, you will need to clean some more.  The battery corrosion is still there on the pcb.

Louis Rossmann does a lot of liquid damage repair on his youtube channel and always uses a ton of flux with hot air to get rid of all that corrosion and to make all those solder pads clean and shiny again.

[carl_lab]

Clean it in an ultrasonic bath with isopropanol/water mixture.
Dry it in oven at 50-60°C.

The engineers at Gossen had placed the battery holders in the X-Tra Series multimeters directly on the main board, without any barrier to the battery compartment.

"X-Tra" for extra cheap...?

[Dachpappe]

At first thank you for your answers!

Sadly I don't have access to an ultrasonic cleaner, but at the university we have a nice hot air soldering station and good flux available.
I will try that tomorrow on the parts that are not solderable by hand (at least for me).

Until now I cleaned the rest of the meter and cleaned/repaired the traces around the dual mosfet. I desoldered it, cleaned the PCB with alcohol and used a glass fiber brush on the difficult spots. Later I used a self-made mixture of rosin and alcohol and coated the traces with tin. On the chip I also had to scrape off the rotten tin before it was solderable again. Soldered the mosfet back on and tadaa: multimeter turned on after hooking it up to the lab power supply! I was able to go through the menu, but I am not able to measure a thing. Only a decimal point is displayed regardless of which range/mode is selected.

Regarding to the 3 "CT2 16Z" I found out that they are presumably multi-LDO packages from Ricoh. Likely a RX5320X series chip. If anyone is able to find the marking specifications for this chip please feel free to share the exact model and voltage! The only promising link I could find with Google is unresponsive (www.aeneas.com.cn/PDF/Ricoh/new/MARKING.pdf).

After further investigation these chips are more likely level shifters

Two things to add:
1st: I made a mistake with the product series, it's actually called the "Star-Series"
2nd: The beloved MetraHit Energy has the same design in accordance to the battery holders. So better get your batteries out there when not using it!

[HighVoltage]

Lay it in a bath of alcohol / water solution and leave it in there for several hours to get the acid out of all areas. Use a soft brush to gently clean all parts of the PCB. Then leave it a while longer in the solution. Finally take it out and dry it in an oven at around 50 degree C for at least a couple hours.
Repeat is needed
Then go in to the details of fixing traces.

I had very good results with this procedure on many PCB's.

[Dachpappe]

Here's a quick little update on the progress of the repair:

I cleaned the PCB some more with the suggested 1:1 alcohol/deionized water solution and it worked quite well! 

Also I desoldered the LDOs and discovered that at least one pin per part was disconnected due to corrosion and a lot of gunk was trapped under them.
The chips themselves look intact, I will try to clean them and will solder them back on later.
Furthermore pin 1 of the HCT573 got eaten up by the acid, so I will have to replace that one as well as the damaged passives.

[HighVoltage]

This looks very well fixable now.
You can use very fine sandpaper on the traces to clean them.

[carl_lab]

You can use a fiberglass eraser to clean the copper tracks:



https://de.wikipedia.org/wiki/Glasfaserradierer

[Dachpappe]

Another quick update:
Due to the lack of time there was no great progress on the project this week.
I had a closer look on the PCB under the microscope and discovered at least one trace that is completely eaten up. Also pin 1 (output enable) of the HCT573 is missing, so that might be the reason for the lack of communication between the main processor and the conversion part. A replacement arrived today from the US (mouser to be more specific) as no matching parts in this form factor were available from any small seller in Europe. I also ordered a replacement for the dual MOSFET near the battery terminal. Interestingly the FDS6894A is obsolete since 10/2011 so I ordered the suggested replacement FDS6890A.

For mechanical cleaning I already used a glass-fiber-brush on some parts and the outcome was satisfying.

Hopefully I will have some more time next week before the lab closes down over christmas.

As always here are some photos for you (this time closeups from under the microscope).

[Dachpappe]

Today I had the time to solder my new 573 and the cleaned CT2 chips back on.
(The CT2-16Z turned out to be SN74LVC2T45 Dual-Bit Dual-Supply Bus Transceiver IC's from TI). Identified by using http://www.chip.tomsk.ru/chip/chipdoc.nsf?OpenDatabase
Soldered everything under the microscope made sure nothing was shorted and powered on the board later. Sadly the issue remains the same.

So the investigation began again.  I traced the data lines connected to the transceivers. One side is always connected to the 430 and the other side is connected to the custom. Looks like Gossen realized a 4 bit communication between them. I could trace 4 lanes to alongside pins on the custom but only 3 to the 430.
The 4th data line can only be traced to a via under the 430 (next to the vias for the other lines). There I lost it. Maybe there is a broken trace underneath the 430?

What do you think? Should I try soldering a bodge-wire fitted with a small resistor (maybe 1k) from the lonely transceiver pin to the suspected pin on the 430? Does anyone of you have another idea?

[Dachpappe]

Well, it has been a while since I last posted an update on this but there are good news!

I figured that the trace I expected to be broken was in fact a broken.
So I replaced the SN74LVC2T45 level shifters and soldered a bugfix-wire from the MSP430 to the level shifter where the trace belonged to.
After testing I glued the wire to the pcb so it does not "flap around in the breeze" 

Turned it on and voila: meter is working fine, and the readings are looking good as far as I can judge it. 

[Florian68]

My multimeter Metrahit X-tra is dead. Can you help me to solve them, please!
I don`t know what is the problem. Everething looks good. Some leaked Battery on the minus pole.

[carl_lab]

First clean it as described in this thread.
Then look for suspicious tracks and components.

[Dachpappe]

Take off the backcover of the meter and have a look how bad the damage is. Dave has a video on his channel on how to disassemble a Metrahit Energy. The other Starline-series meters are quite similar on their mechanical part. If you're lucky and only a bit of juice got around the terminals it might be an easy fix. Directly next to the main negative terminal is a dual mosfet for power control. This gets eaten up first. If there's more damage then you need to get the board out and clean everything very well. When taking the board out be sure to follow Dave's advice on how to unsolder the 4mm terminals. If you bend them more than a few degrees they break.
Maybe you can post some pictures of the damage so we can have a look. Also it might be a good idea to start a new thread and leave a link here

[Twoflower]

Check the marked solder joint I marked in the picture. It looks like it is only tacked on.

[Dachpappe]

Check the marked solder joint I marked in the picture. It looks like it is only tacked on.

You're right, that one looked a bit messy and got touched afterwards. Meter is still working fine!

[Florian68]

Thank you for the answer! Monday when I go to work I will make some photos to see.  I tell you that the board look very very good.
I have seen corrosion traces only a little bit at the base of the negative pole of the battery.

[Florian68]

Hello guys,
This is my dead multimeter! I need your help!

[Dachpappe]

Well that does not look as bad as I had expected. Clean everything that has come in contact with battery juice well. For me a solution of 50% ethanol + 50% deionised warm water in combination with a toothbrush worked well. I discovered that the batteries in my multimeter placed an oily film through the whole meter. If that's the case at least clean the rest of the meter with alcohol. After cleaning let it dry for a few hours at a warm and dry place.
Do you have access to a lab power supply with 2.8 - 3V and 200mA current limit? If that's the case connect it to the battery terminals and try to turn it on after drying. My meter draws around 11mA when in DC volts range but needs a little over 100mA for power up.

[Florian68]

I measured 4.7mA in stand- by and when I push the power button 7.7mA. Mutimeter don`t start.

[Dachpappe]

That's too much current in standby. There's current flowing where it shouldn't. Maybe take another round of cleaning. Try probing components around the battery terminal to see if they're ok. Resistors and diodes often fail after contact with corrosive substances. Especially check the power management circuit as your meter is configured for the external supply. The dual mosfet (mentioned before) should be responsible for choosing the supply and it would make sense if the battery is disconnected when external power is detected. Make sure that the solder joints between components and the pcb are intact. In my case a lot of them were either dead (but rather good looking) or intermittent.
Feel free to post a few close-up pictures of the affected area so we can have a look.

[Florian68]

This is the pictures. In standby is too much 4.7mA? What is the correct value in standby?

[Dachpappe]

The standby current on my meter is well under 1 mA. Also you have to clean some more. There is still residue left. Take a toothbrush and the suggested solution and clean the affected area really well. After this take another measurement and check the surrounding components if the problem still persists.

[Florian68]

Problem stilll persist! Do you have a diagram of this multimeter? I measured at HCT573 in 20 pin and I have 0V. Why?