Fluke 187 DMM upgrade to 189

[shakalnokturn]

Hello,

I recently got my hands on a cheap working Fluke 187 multimeter, I haven't opened it yet but from photos I've seen there is about a 4 component difference with the 189 model. (20k resistor, ceramic decoupling cap, SRAM, leaky supercap.)
I don't really need the extra functions on the 189 but if it can be done easily it will be hard to resist.

Firmware version is 2.02

A couple questions the photos I found could not answer:

Is the SRAM a 32k x8bits model or smaller still?
How is the function switch rotation limited? (Will I be cutting plastic away to access the logging position?)

Edit:
A couple of related topics:

https://www.eevblog.com/forum/testgear/fluke-87-4-to-89-4/

https://www.eevblog.com/forum/testgear/fluke-87iv-please-whisper-in-my-ir-p/

[shakalnokturn]

For the cutting plastic away, the answer is "yes" the rotation limit is on the underside of the grey rotary knob, and the 187 will display "LOG" when turned beyond the µA DC position.
I have not been able to find a good enough photo of a 189 on the LCD side of the PCB to read the SRAM reference. I think it's a HY62256B-LLT1-70,  it would be great if someone could check though, as that's a 5V Vcc part but I only ever measure 3V Vcc to this IC in the Fluke 187 it doesn't make sense.

[shakalnokturn]

Auto-updating...

SRAM should preferably be a low standby current version. I realized a little late that most manufacturers use "LL" in the suffix for that option but not Mitsubishi, that should have been a M5M52256-CVP70XL.
I'll swap that out if there's any noticeable draw on the batteries. I find it can be a good thing having to change batteries anyway, as now you can't trust many of them to not leak on the long term.

Next step will be fitting a new (FC0H224ZFTBR24) backup EDLC in there and finding/building an IR interface.

[ogdento]

You ever make any progress on this project?

[shakalnokturn]

I'm not quick at any projects, I still have to see to adding a stopper to the function knob, change the backup capacitor.
I don't think I'll be swapping the SRAM for a lower power version. The IR interface has not been tried either yet.
However as it is today my 187 does have all the 189 functionality.

If you're thinking of doing likewise you'll need either a 189 top half case or to cut the CW rotation limit off the selector and find a way to place another one position further.
Backup cap, HY62U256B  LLT1-10I (original part confirmed by coromondalix), 20k 0805, 100nF 0805, BAT54S.

[shakalnokturn]

I finally got down to replacing the backup EDLC on my 189, the funny thing is that it lost the logging function after that.
Turning to "VIEW MEM" would display normally as if memory was empty.
The suspected reason is SRAM lockup due to power glitches during the EDLC replacement. (I have experienced similar problems hacking T.I. graphic calculators.)

For those experiencing troubles after replacement of the backup capacitor:
Removing batteries, restoring factory defaults was no help, the solution was to discharge the new EDLC. (I did this using another DMM on mA range.)
After this the save feature was restored but displayed "FULL" after storing only 3 values, this was solved by doing a "CLEAR MEM"

Edit: With new EDLC and not quite perfect SRAM reference, OFF power consumption is approx 115µA @ 5.8Vbat.

[shakalnokturn]

For the new rotation limit I used a blob of "Fermacell" dry wall glue because the stuff is tough, has high viscosity and sticks like s#!t. I does take a long time to set though... A blob of epoxy should do the trick too.

[ogdento]

Very nice work on the knob, I'd never heard of Fermacell before but it looks like some good stuff if it will stop that knob!

I just did a similar thing to another Fluke by drilling a small hole in the case and gluing in a small plastic "pin" I cut from scrap... the pin is securely set into the case and the very top bit became the new knob-stop.

[masterx81]

Really interesting, someone know if this apply also to an 87 IV?

[masterx81]

I can answer myself. Actually, it's easier, as the sram is already placed on the board. Only need to move the stop. Thanks for this post!

[shakalnokturn]

Small update:

I was lucky masterx81 turned up with his 87IV, curiosity dragged me in to realising that my multimeter wasn't working as reliably as I thought with a 5V RAM where a 3V model is needed. OTOH he was lucky I fitted the wrong RAM to start off with, it made things easier finding out where the difference was between a 87IV and 89IV.

The 187/189 F.W. V2.02 will check RAM and automatically update model number in EEPROM, the 87/89IV F.W. V1.01 however doesn't (this makes sense if all the 87V already have RAM populated on PCB). Maybe situation is as with 187/189 on later revisions: F.W. can autodetect but RAM physically missing?

If you're looking to upgrade your 87IV to a 89IV, other than the switch that needs modifying to turn to the "View Log" position the EEPROM address 0x0069 needs contents changing from 0x0037 to 0x0039.
It should be possible to do this over I.R. RS232, unfortunately I don't know the commands that must be issued.

Attached a EEPROM dump from my DMM while it was a 187, and a spreadsheet of the EEPROM bus activity during 3 different power-on cases.