Author Topic: Gossen MetraHit 22M / 26M internal photos wanted! (Read 3743 times)

shakalnokturn · « **on:** February 07, 2019, 01:36:03 pm »

After recently getting hold of a battery contaminated MetraHit 24S DMM and spending a few hours cleaning it out, I'm looking into upgrading it with the logging functionality.
I have already turned my Fluke 187 into a 189, there's probably not much more to it for the Gossen other than the lack of photos to help me.

One side of the PCB has two empty I.C. footprints, the TSOP one is going to be a 128kB SRAM, the other likely some 74HC logic.
Under the LCD there are 11 empty passive footprints, I'm hoping the same µC is used over 22 to 26 models, and that one of these empty footprints is for a level pulling resistor to enable the logging option.

Anyone out there have photos of a 22M or 26M they would like to share?

coromonadalix · « **Reply #1 on:** February 08, 2019, 10:15:52 am »

You love challenges loll

Good luck

shakalnokturn · « **Reply #2 on:** February 08, 2019, 08:36:21 pm »

I do like a little challenge. All the same without the photos this is a whole lot more of a challenge than I want right now.
A quick guess at the missing SOP16 footprint would be that the µC was lacking pins to address the 128kB SRAM directly, so depending on number of missing bits could be a SIPO (74HC595 ?) or just a hex buffer.
The multilayer PCB makes tracing around a nuisance.

_Wim_ · « **Reply #3 on:** February 08, 2019, 09:30:29 pm »

I have a 22m, will try to take some pictures tomorrow.

shakalnokturn · « **Reply #4 on:** February 09, 2019, 09:09:59 am »

That would be great!
If you disassemble the LCD, take care to not put too much torque on the screws reassembling them as they are threaded into the PCB.

_Wim_ · « **Reply #5 on:** February 10, 2019, 10:26:56 am »

Hi,, with some delay (was EEVBlog down yesterday, or was my PC acting up?) here some pictures. For the moment no backside of the pcb, because the pcb seems quite stuck, and did not have time to desolder the terminals. If necesarry I can do this later on, but with these you can already check if it is feasable or not.

coromonadalix · « **Reply #6 on:** February 10, 2019, 04:49:16 pm »

@ Wim

Dont desolder the input jacks, if you're not careful or fast enough it may melt the plastic around them, thrust me i know learned this on a old damaged xx metrawatt meter, the pcb snap into the case, you have 4 little black plastic retainers to gently push while lifting the board .....

I know you may want to help but the metrawatt are kinda fragile while the case is opened be careful, you may affect the calibration etc .... use anti static protection gear.

With the photos you see the ic's markings.

The main plcc 44 pins 01 01 ... is a custom asic, and the cpus are Hitachi or toshiba sama family if i recall correctly, they need special adpaters to try a content read of them, last time i checked, the socket adapters costed 220$ usd ........

shakalnokturn · « **Reply #7 on:** February 10, 2019, 05:27:37 pm »

Thanks a bunch ! Really helpful of you.
I'm a little curious, are you the BIOS guy?

Yes, EEVblog was down yesterday.

So I was wrong, it was a 4040 counter... That's already enough to make me want to give it a try as I must have the RAM and 4040 in my stash.

The footprints under the LCD must be level pulling resistors, when I get round to it I'll check what they link to.

The PCB is held into the front half of the case by the 4 input wires and 2 plastic clips on each of the long sides.

I'll update when I find a bit of quiet time to work on this again.

_Wim_ · « **Reply #8 on:** February 10, 2019, 07:26:02 pm »

No, not the BIOS guy

I have a serial adaptor also. Got both the meter and the adaptor together with some other gear, did not buy it new luckaly. Let me know it you need some other pictures.

shakalnokturn · « **Reply #9 on:** February 14, 2019, 04:04:56 pm »

I have been through my stash of pulled components, have the 74HC4040 at hand. The problem is the BS62LV1024 SRAM being a very low standby current model 1µA max @ 3V.
The best I have is a HY628100B with a standby draw of 30µA max @ 5V, that should actually be a lot lower at 3V, anyway I've ordered a CY62128EV30 (4µA max @ 3V).

I've been around google and EEVblog looking at other MetraHit's from the 22-26 range for photos. My 24S has a µC with firmware version 7.2 (date code 154.02) this is the same version as _Wim_ has on his 22M, I've also seen the same version on some 24S and 26S models so I'm quite confident this can be done.

@_Wim_: Once I have traced where the level pulling resistors link (if you have another DMM at hand) would you mind making some resistance measurements on the rear side of the PCB to check the ones that are mounted on your 22M without having to desolder the input wires and remove the LCD?

_Wim_ · « **Reply #10 on:** February 14, 2019, 07:01:09 pm »

Hi shakalnokturn,

No problem to make some meausements, just ask. If desired, you could also borrow my one for a couple of months to perform your reverse engineering, I do not use it so often. Just contact me via pm.

MichalZ · « **Reply #11 on:** February 14, 2019, 07:24:41 pm »

Hello, some photo of Metrahit 23S meters:

shakalnokturn · « **Reply #12 on:** February 15, 2019, 01:14:20 pm »

Quote from: _Wim_ on February 14, 2019, 07:01:09 pm

Hi shakalnokturn,

No problem to make some meausements, just ask. If desired, you could also borrow my one for a couple of months to perform your reverse engineering, I do not use it so often. Just contact me via pm.

Thanks for the offer, I will consider it if really stuck but would like to avoid that as much as I can.

Here are the photos with a little reverse engineering, I may get round to cleaning the red numbering into actual names later.

The 4040 counter addresses RAM A0 to A11 the remaining A12 to A16 are straight from the µ controller.
Of the 11 empty footprints under the LCD, one is for a ceramic decoupling cap, for the others I expect resistors if these are needed at all.

Resistance measurements on the points labelled 1 to 9 WRT Vcc are mostly in the 2.2M $\$\Omega\$$ or 4.8M $\$\Omega\$$ depending on ohmmeter's polarity.

_Wim_ could you measure those points (1-10 on photo) WRT Vcc on your MetraHit 22M when you have a little time for that?

A few side notes:

Current current consumptions (before adding anything to the 24S) and on Lithium batteries (Vbat=3.4V) are: 9 to 12mA depending on function selected, 184µA in standby.
Funnily I was expecting at least the SRAM would be powered straight from the batteries in standby, as lowering the supply voltage also reduces the current draw. This is not the case the MAX856 supplies the SRAM with 5V even when the meter is in standby.

OTOH the 93LC66 is not powered straight from 5V Vcc but gets power through the µ controller (@4.9V) and is not supplied in standby, this could be either to reduce standby power or as a measure to not corrupt memory on power transients (?)

Another thought on the "M" option for these meters is that it may be software enabled in the 93LC66. I'll dump the 93LC66 soon that will at least get me a backup of calibration data "just in case".

@ MichalZ: Thanks for the photos, they further confirm that different FW revisions have been used throughout the range and likely indifferently on "S" and "M" models. V7.2 is the highest revision I have spotted so far.

shakalnokturn · « **Reply #13 on:** February 16, 2019, 08:54:18 am »

Here goes the 93LC66B EEPROM dump.

I'd be curious to compare it with other dumps from the 22-26 series M or S models if anyone has any available.

_Wim_ · « **Reply #14 on:** February 16, 2019, 12:06:45 pm »

Here are the measurements.

1=510K
2=510K
3=510K
4=10K
5=100K
6=100K
7=510K
8=510K
9= 10K
10=100K

I also disassembled my meter further (and successfully assembled it again), so also attached are the pictures of the backside.

shakalnokturn · « **Reply #15 on:** February 17, 2019, 12:26:43 pm »

Thanks a lot _Wim_ for going to the trouble of taking the extra pictures and measurements, I now have the logging function enabled on my MetraHit but I doubt I would have succeeded without your photos.

I got a little impatient to try and soldered the HY628100B SRAM I had, added the missing resistors (didn't have any 510k at hand, used 499k instead...) and 100nF decoupling caps.
I'll have to clean that rosin mess some time, it looks terrible on the TSOP package!

The good news is that I didn't have to play around with the 93LC66B, there is a difference other than the missing components between my photos and Wim's on the LCD side of the PCB, it actually took me a while to see it.
Does anyone else notice it?

Although the logging function is very complete with advanced triggering adjustments, I'm actually disappointed to see that you can't even browse through the logs on the DMM. The only way to make use of the logging is through the serial link?
Does the DMM identify it's model number to the Metra10Win software? That could be somewhere in the 93LC66B.

I tried to measure the current draw after the modifications... Well I knew I should have done that with a lab power supply from the start, my battery voltage has dropped a little so it turns out I'm measuring less current draw with the extra components.
Anyway I doubt I'll even bother swapping the SRAM out for a (theoretically) lower power CY62128EV30 when it arrives.

coromonadalix · « **Reply #16 on:** February 17, 2019, 04:58:04 pm »

congrats

Falkra · « **Reply #17 on:** February 17, 2019, 05:14:00 pm »

Very nice job !

shakalnokturn · « **Reply #18 on:** February 17, 2019, 08:14:55 pm »

Thanks!
Another case of doing it "just because it can be done".

Ok, I'll give the detail away for those who are tempted to follow the example: On Wim's 22M there is a 33k mini-Melf resistor that is non mounted while mounted on my 24S (see photo). It links to pin 40 on the µ controller, I believe this is the one that enables the logging option.

I was also considering turning this 24M into a true RMS meter, don't have the courage right now, have other things to see to.

During my search for other models of the range I noticed something odd:
The back half of the case has a shiny metal disc as if some shielding was embedded into the rear case, the PCB's plastic cover has a hole corresponding to this bear metal, most models have the precision resistor network in the way but some models don't. Instead they have a large via that looks as if it could have been intended to have a spring soldered to for making contact with the back case.

coromonadalix · « **Reply #19 on:** February 18, 2019, 02:38:44 am »

I think the hole is for the frequency compensation on the 28s and 29s. On mine, when the main case is removed, you see an adjustment thru the secondary plastic shield ??

I dont have any metal spring who make an shield contact on the main rear case, only a metallic washer with the screw for holding the secondary plastic shield, not sure if the plastic is conductive in some way

shakalnokturn · « **Reply #20 on:** September 25, 2019, 10:20:19 pm »

I'm bringing this one back-up to correct a mistake: My meter was giving false current values, that turned out to be because of my removing of the previously mentioned 33k mini-melf resistor. So apparently this is removed for the 22S/M models with no specific "mA" range.
Memory option is still present.

On reassembling I managed to trap the battery holder wires under the plastic cover screw hole and shorted the batteries, to I dragged that out to the other end and fastened it tight.

Then had a little check on how well my GMC's agree.
Which one should I give away?


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Gossen MetraHit 22M / 26M internal photos wanted! (Read 3743 times)

Share me