Metrahit 25S Teardown + Component Analysis

[Felicitus]

Hi,

this is a short teardown of my newly arrived Metrahit 25S multimeter. It has not much in it; basically they seem to do everything with their custom IC.

Notable things:
- metal can TrueRMS converter (AD636KH)
- Isabellenhütte 50ppm/K 0.5? resistor (probably shunt? notable as the manufacturer Isabellenhütte seems to be little known)
- Seems that the main microcontroller (TMP87PP23F) has dual clock inputs; in that case 2MHz and the usual 32kHz watch crystal
- The IR diode has no opening in the case; instead, it actually transmits through the closed case

There seems to be nothing much on the other side of the PCB, no actives at all, only some passives and the LCD. As one has to desolder the input pins to flip it completely over, I skipped that.

A hi-res version (PDF, 15MB) of the attached image can be found here: https://dl.dropboxusercontent.com/u/3343223/Metrahit25S.pdf

[Wytnucls]

Good picture and component labels. Curious to see if my 26s, which I will receive next week, has the same layout.
Do you know if the SIBA fuses were fitted at the factory on the original series 2 Gossen meters?

[Felicitus]

Good picture and component labels. Curious to see if my 26s, which I will receive next week, has the same layout.

I do believe so - the unfitted component footprints look suspiciously like a flash and some support circuitry

Do you know if the SIBA fuses were fitted at the factory on the original series 2 Gossen meters?

I have no idea, my hands-on gossen metrahit experience is less than 12 hours old

[Vgkid]

Thanks for the teardown, it is rather strange that the AD 636 line is still all through hole.

[retiredcaps]

Just curious who makes those 5 electrolytic caps?  Can you list the brand and series?

[Felicitus]

Just curious who makes those 5 electrolytic caps?  Can you list the brand and series?

No, already put the meter together, sorry. Next time I open it up, or maybe Wytnucls can give some insight when his meter arrives.

[Wytnucls]

Sure, I'll have a look.

[TMM]

I'm intrigued to know what the purpose of those large holes/vias are:

Surely, a whole lot of smaller vias would be cheaper and lower impedance

[Felicitus]

I'm intrigued to know what the purpose of those large holes/vias are:

Surely, a whole lot of smaller vias would be cheaper and lower impedance

I have spotted those, too, and one of those actually has a matching reflecting ring in the case (air distance about 1cm) - not sure what this is about, but as I haven't looked deep on the other PCB side, I have no idea what this is supposed to do.

[Vgkid]

Do the holes match up with screws or metal standoffs?

[Felicitus]

Do the holes match up with screws or metal standoffs?

No.

[peter.mitchell]

Isabellenhütte 50ppm/K 0.5? resistor (probably shunt? notable as the manufacturer Isabellenhütte seems to be little known)

I think the large metal sheet is the shunt, not this.

[Len]

Isabellenhütte 50ppm/K 0.5? resistor (probably shunt? notable as the manufacturer Isabellenhütte seems to be little known)

I think the large metal sheet is the shunt, not this.

Could be both, for different current ranges?

[Felicitus]

Could be both, for different current ranges?

I think so, too. Why else would you need a precision resistor in that spot?

A few notes about the usability of the meter:
- The bargraf actually displays the range above the major ticks (e.g. 100, 200, 300, 400 or 10, 20, 30, 40), and not "1 2 3 4" with a multiplier on the right side as on the Fluke 87. I like that very much, it reduces confusion when you have peaks of some sort
- The min/max/avg functions are a bit odd to use, one should really read the manual, especially on the timed min/max function
- You can specify a low/high event trigger for voltage as well as counting zero crossings
- It has a signal generator which has an amplitude of 3V and a frequency between 1 and 1000Hz. It's possible to configure delays between the pulses. Not sure if this could come in handy some day…
- The continuity tester is not the fastest, but also isn't slow (about half as quick as the Fluke 87 one), but it's possible to configure the resistance when to trigger the beeper
- The data hold function is a bit slow, it needs a relatively stable measurement for 2 seconds

So far I can highly recommend that MetraHit meter, and it certainly plays one league above the Fluke 87.

[Wytnucls]

Capacitors seem to be made by United Chemi-Con, from the KMA series 10V 105deg C.

http://www.datasheetarchive.com/dl/Datasheet-026/DSA00459677.pdf

[Wytnucls]

The 26s, with AC+DC and higher 20kHz capability, has a few more components, using the same PCB as the 25s.

[Wytnucls]

I'm intrigued to know what the purpose of those large holes/vias are:

Surely, a whole lot of smaller vias would be cheaper and lower impedance

The one next to the 16A fuse is the receptacle for a screw that holds the internal PCB cover in place.
The other one takes an adjustable component fitted for the 26s. Something to do with the higher frequency or AC+DC averaging, I would imagine. (Looks like an adjustable capacitor)

[Fraser]

I briefly owned a Metrahit 16S but that was a used unit and the LCD display just started to loose its contrast over time. I found the contrast adjustment pot on the PCB and it was already almost at MAX. The meter eventually became unreadable unless viewed at a very specific angle. Some background checking suggested that these meters can suffer significant LCD contrast degradation over time. For that reason I have avoided them since, and stuck to my Fluke 87-Mk3. I like the Metrahit meters so would consider getting another if they now use a decent LCD.

Has anyone else knowledge of this 'issue' and whether it has been resolved since the 16S was produced ?

[Felicitus]

Fuse types found in my 25S:

Siba FF 1,6A 700V (70 125 40)
Siba ultra-rapid 16A 1000V 30kA (50 199 06)

Maybe not original, but I doubt it, even if the Siba 1,6A has only 700V printed on it.

It's interesting that my 25S has no adjustment pots at all, compared to the 26S.

[Wytnucls]

Strange, I haven't seen faded LCDs on MetraHit meters yet.
I have a second series 26s which dates from 2006, according to the calibration certificate. The LCD is sharp and clear as if it was fitted yesterday.

[Wytnucls]

Fuse types found in my 25S:

Siba FF 1,6A 700V (70 125 40)
Siba ultra-rapid 16A 1000V 30kA (50 199 06)

Maybe not original, but I doubt it, even if the Siba 1,6A has only 700V printed on it.

It's interesting that my 25S has no adjustment pots at all, compared to the 26S.

I'm pretty sure, the original fuses would have been 1000V. They're not pots, but adjustable caps.

[Lunasix]

For comparison, I've posted photos of a Metrahit 28S.

https://www.eevblog.com/forum/testgear/gossen-metrahit-28s/

I like precision of this meter but functions are not easy to use.

[Wytnucls]

I know what you mean... took me a whole day to make sense of the 26S user manual. They are still a couple of features I'm unsure of. 

[Lunasix]

But I don't really use or need extra functions, this meter is on my desk and as replaced (for a small price  ) a bench Schlumberger Solartron meter. I like the big rotary switch, and precision is very good : I have a lot of 0.1 and 0.01% precision resistors, and measure (and so resistor) is always better than value + possible error indicated on resistor. Capable of measuring 1uV or 1nA, temperature reading with thermocouple, PT100 and PT1000. I use this function to select resistors used to calibrate PT1000 converters.
I've compared with other meters, like Amprobe-Beckmann-Meterman... 
I also have a Metrix MX53C (50000 counts and 0.1% precision), good !
No mistery  : cheap multimeters are rarely good, even if they are on paper.
I prefer an already used good multimeter as a new one but not good.

[Wytnucls]

I checked the accuracy of my recently acquired second-hand MetraHit 26S (last calibrated Oct 2006) against the MetraHit 30M (last calibrated Oct 12). I also plugged in the UT71 and UT61E for comparison. (DC voltage accuracy)
I'm pleased with the results, except for the UT61E, which is way out of whack.