Hameg HM8118 repair

[TRN]

Hi,
I recently purchased a defect HM8118 on eBay. This unit has several issues, one of which is that only half of the display is working,as you can see from the attached pictures.
I managed to update the firmware from V1.47 to V1.57, but of course this did not fix the LCD problem.
 
The LCD is a COG-LCM138-06 with serial number: 110111408422, and I presume that the one of the controllers on the LCD is faulty, and thus does not drive part of the LCD anymore.
Does anybody have a spare working display for sale, or else can me provide a datasheet, because I searched the internet, and cannot find any information about the LCD; not even who manufactured it.

I also could not find a service manual for the HM8118. If somebody has this a copy would be greatly appreciated.

Here is all the info what is written on the LCD, and also on the separate backlight unit. (see also the pictures)

FPC-LCM138-01, 2004.05.11
3810
SHON58 VL

COG-LCM138-06

WU621A-01
LBL-LCM138-01A
210/11/06

Thanks

[Hamelec]

have a look into the comments of
Maybe Michel.Waleczek (at) gmail.com can help you out with a manual...

[shakalnokturn]

Good luck, sourcing this kind of display is a nightmare.
Of course you could kindly ask Rhode & Schwartz for some information or spares.
If you're on your own in the end, it can help searching to know the display resolution, number of pins on FPC and bus data format.

[Hamelec]

cant believe that it is a failure of the display themselfes, more of the flexboard/connector or driver unit

[coromonadalix]

as for the problem,  could be hw failure of the lcd matrix glass or on the flex lcd assembly  or even the lcd controller

yes  ask the company  under an non disclosure agrrement if you can order the SV Manual   or get that part ?

Unless it's a customized one because of the 2x  pcb traces groups on each side,  i never see theses in lcd's ??

[jchw4]

cant believe that it is a failure of the display themselfes, more of the flexboard/connector or driver unit

Exactly.

The flex bonding looks very suspicious:


Does it normally fold the flex cable back behind the display? I would try pressing the bond with pencil eraser and see whether it helps.

My personal experience with trying to heat up and re-bond the flex cables under stress has been negative so far. It works until you fold the cable in place and then detaches again.
You may be able to fold a piece of sheet metal to make a clip and attach it to the top to hold the flex cable. It worked for me once...

[coromonadalix]

they are glued on the glass with conductive glue ....   and sometimes having some hardware who push / press on theses  could  help

its a gamble

[TRN]

Good luck, sourcing this kind of display is a nightmare.
Of course you could kindly ask Rhode & Schwartz for some information or spares.
If you're on your own in the end, it can help searching to know the display resolution, number of pins on FPC and bus data format.

Well I have tried all angles to find any information and /or replacement of this display. But so far no luck. There is no information whatsoever available on the internet, so that seems to be a dead end.
I also send an email to Sky-Messtechnik.de, which is the R&S authorized service company for Hameg equipment, and although they conduct component level repairs, they do not sell spare parts for the HM8118, nor do they handout service documentation for this unit.
Just like R&S they have standard repair prices, which for the HM8118 is:

Standard Repair Price is  € 405,00 net   incl. adjustment   
(repair on component level, re-adjustment without measurement protocol)

If display or boards needed:

Display              + €   70,00 net
Main Board        + € 225,00 net
Calibration         + € 250,00 net  (incl. measurement protocol from external calibration lab)

So a repair would minimal cost me a hefty ( 405+70) +23% VAT = 584,00 EUR excluding the shipping cost from Portugal to Germany

[TRN]

they are glued on the glass with conductive glue ....   and sometimes having some hardware who push / press on theses  could  help

its a gamble

OK I will try this

[coromonadalix]

the last resort would be to snif out the communication protocols with an logic analyser and figure out wich controller by chance they are using and finally finding the right lcd ??

[TRN]

have a look into the comments of
Maybe Michel.Waleczek (at) gmail.com can help you out with a manual...

Hamelec,

Thanks for this info, and I have contacted Michel, who responded that he is currently enjoying his vacation
I will contact him again once he is back home

[TRN]

cant believe that it is a failure of the display themselfes, more of the flexboard/connector or driver unit

Exactly.

The flex bonding looks very suspicious:
(Attachment Link)

Does it normally fold the flex cable back behind the display? I would try pressing the bond with pencil eraser and see whether it helps.

My personal experience with trying to heat up and re-bond the flex cables under stress has been negative so far. It works until you fold the cable in place and then detaches again.
You may be able to fold a piece of sheet metal to make a clip and attach it to the top to hold the flex cable. It worked for me once...

Your hunch about the suspect flex bonding was correct.👍

With the lcd installed in the unit, I flexed the cable in different positions, while booting the unit, and it booted once with a beautiful full working display.
I then removed the lcd, put some light pressure on the bonding area, and reinstalled the lcd
Unfortunately after this no more luck, so I must have screwed something up👎
Anyway this proved that the lcd matrix and 2 controlers are ok.
Does anybody know any company (preferably in Europe) which repairs flex bondings, and if so, what would be the estimate repair cost?

[TRN]

the last resort would be to snif out the communication protocols with an logic analyser and figure out wich controller by chance they are using and finally finding the right lcd ??

Sniffing out the communication protocol, is probably doable, but trying to find the controller type will be a hard nut to crack

[coromonadalix]

well i checked the video   since some comments are in french (im native french)

there was a substitution with an LCM 219  with a different bootloader  for this lcd ....   and this one too is a tough one to get


R&S  get  bad points here for repairability, and for sure they know it

it would mean to trace down  lcd pinout on the main board and try to figure out the data lines etc ...  i would be able  if i have this lcr  ... wich i dont


look like a 40x8 characters lcd for sure .... if i count them in their user manual images  ....  i dont think its a graphical one ??

[CJay]

look like a 40x8 characters lcd for sure .... if i count them in their user manual images  ....  i dont think its a graphical one ??

I'm pretty convinced it's a graphical LCD, the third picture has an uninterrupted line of pixels down the display, character LCDs generally have 'gaps' between rows of characters.

Looking at the includes and source files in the video linked give the impression it's a graphic one too, LCDDRAW.C and LCDTEXT.C, plus the Epson.c might indicate it could an Epson (compatible?) controller?

The pictures are nearly good enough to count the pixels, I think it's almost definitely 64V just by counting but obvs counting pixels that aren't visible on the RH side makes it tricky, maybe 128H making it a  128x64 (maybe 138 which would mean custom?).

There's 2 chips mounted on the glass and the flat flex looks as though there's a bunch of vias so they're 'in parallel' for the data and probably most control lines so there'll be a 'chip select' line somewhere.

It might be useful to know the physical dimensions of the display.

****Also might be interesting to see a little more of the CPU and memory, if it's got external program memory or there have been firmware updates it might give some clues to examine the contents?****

Interesting, uses a PSD chip so it's 8052 based, firmware update is likely encrypted though

[CJay]

SED controllers likely given the Epson.c in the source tree

[coromonadalix]

goood catch   haven't tought of that  lolll  should have checked more  ....

[TRN]

look like a 40x8 characters lcd for sure .... if i count them in their user manual images  ....  i dont think its a graphical one ??

I'm pretty convinced it's a graphical LCD, the third picture has an uninterrupted line of pixels down the display, character LCDs generally have 'gaps' between rows of characters.

Looking at the includes and source files in the video linked give the impression it's a graphic one too, LCDDRAW.C and LCDTEXT.C, plus the Epson.c might indicate it could an Epson (compatible?) controller?

The pictures are nearly good enough to count the pixels, I think it's almost definitely 64V just by counting but obvs counting pixels that aren't visible on the RH side makes it tricky, maybe 128H making it a  128x64 (maybe 138 which would mean custom?).

There's 2 chips mounted on the glass and the flat flex looks as though there's a bunch of vias so they're 'in parallel' for the data and probably most control lines so there'll be a 'chip select' line somewhere.

It might be useful to know the physical dimensions of the display.

****Also might be interesting to see a little more of the CPU and memory, if it's got external program memory or there have been firmware updates it might give some clues to examine the contents?****

Interesting, uses a PSD chip so it's 8052 based, firmware update is likely encrypted though

Good catch indeed, it is a graphics display.

The owner of the video was so friendly to send me just now a copy of the module specifications.

It is a Varitronixs COG LCM138 240 x 64 Dot graphics display which has 2 x Epson S1D10605 COG's

For who's interested I have attached copies of the spec sheet and S1D10605 data sheet.

[coromonadalix]

just send them an email  if they have a substitution ??  at least a 30 pins flex  with all the signals  ...and the controllers 

yep thats what i tought   2x chipselect,1x Enable pin ...



dual S1D10605  lcd's ?

http://www.3t-lcd.com/MechanicalSpec/Graphical/TMGG24064-XX.pdf                   22 pins lcd ?

COG-BT240064-05
https://www.farnell.com/datasheets/31889.pdf   22 pins lcd ?

TMGG24064-XX
http://www.3t-lcd.com/MechanicalSpec/Graphical/TMGG24064-XX.pdf    22 pins lcd ?

the TMGG24064   is the one we get the most while searching ?

https://i-lcd.com/LCD/Graphic/240x64.php

[CJay]

I like a result like that, finding and getting help from the designer is just great.

Interesting solution to limited resolution controller chips too, I really hope you can find an alternative display.

If you can't then it might be worth trying to track down one of the car LCD dashboard repairers, some of them have the equipment to rebond/replace the flat flex.

[TRN]

just send them an email  if they have a substitution ??  at least a 30 pins flex  with all the signals  ...and the controllers 

yep thats what i tought   2x chipselect,1x Enable pin ...



dual S1D10605  lcd's ?

http://www.3t-lcd.com/MechanicalSpec/Graphical/TMGG24064-XX.pdf                   22 pins lcd ?

COG-BT240064-05
https://www.farnell.com/datasheets/31889.pdf   22 pins lcd ?

TMGG24064-XX
http://www.3t-lcd.com/MechanicalSpec/Graphical/TMGG24064-XX.pdf    22 pins lcd ?

the TMGG24064   is the one we get the most while searching ?,   it would need some pcb adapters to reduce the pin count ??

https://i-lcd.com/LCD/Graphic/240x64.php

22pins means a serial interface
Unfortunately this LCD has parallel interface

[coromonadalix]

I think you should check you received  pdf from the designer  before answering ....

Check the pinouts and you'll learn that your has SPI  lines, see SCL SI  etc ....  unless specifically written parallel operation and or 8088 interfacing etc ....  witch the pdf doesn't tell


Well thats all for me, it seem you have some stubbornness,   i'll keep  my comments for now and read only

Now that you seem to have some connection from the designer,  if he still wants to answer, ask him for substitutes,  and do some search now that you got some answers on the 2x controller blob

I think you may, in the end do some magic in the hardware and maybe firmware too    R&S  should be made aware of this problem, unless they already know it, and not selling parts is not good even if you end up with a non calibrated unit, this is abusive behaviour.

Theses lcd's are old or unobtanium and the other suggested are more recent, but still old in some models ...

So sad to have an expensive lcr pludered by a now cheap lcd problem,   take some courage and made it repaired ?

[TRN]

I think you should check you received  pdf from the designer  before answering ....

Check the pinouts and you'll learn that your has SPI  lines, see SCL SI  etc ....  unless specifically written parallel operation and or 8088 interfacing etc ....  witch the pdf doesn't tell


Well thats all for me, it seem you have some stubbornness,   i'll keep  my comments for now and read only

Now that you seem to have some connection from the designer,  if he still wants to answer, ask him for substitutes,  and do some search now that you got some answers on the 2x controller blob

I think you may, in the end do some magic in the hardware and maybe firmware too    R&S  should be made aware of this problem, unless they already know it, and not selling parts is not good even if you end up with a non calibrated unit, this is abusive behaviour.

Theses lcd's are old or unobtanium and the other suggested are more recent, but still old in some models ...

So sad to have an expensive lcr pludered by a now cheap lcd problem,   take some courage and made it repaired ?

Hello coromonadalix,
All your comments are well received and appreciated, so keep them coming , and I don't think I am stubborn

I might be mistaken but the reason I mentioned that the protocol is parallel, because tracing the PCB, all 8 data lines are connected to the MCU via 51ohm resistors, and I figured that if SPI was used they would not have connected D0-D5, but only D6 & D7

[coromonadalix]

ok i admit  i'm in the wrong  sorry   

check  pin2   p/s  the answer is there

maybe find some projects  with similar displays ....  and you have  other 240x64  dual drivers like yours    but in the standard physical metal frame

some specs  found today

https://www.crystalfontz.com/controllers/Epson/S1D10605/80/   attached here

from

https://www.crystalfontz.com/controllers/Epson


possible substitutions   ??  but 8088 mode and not the same controller  .... would be nice to have help from the designer
https://www.winstar.com.tw/products/graphic-lcd-display-module/240x64-lcd-display.html

https://www.raystar-optronics.com/graphic-lcd-display-module/240x64-lcd-display.html
https://www.raystar-optronics.com/upload_files/monochrome-lcd-module/240x64-graphic-lcd-dispaly/RG24064A2-datasheet.pdf

https://www.orientdisplay.com/products/amg24064p-g-w6wfdw-4-3-240x64-cog-lcd-module/

now i understand   when you have 22 pins  you loose the parallel mode ....

so many driver ic's and possibilities .... but no access to the cpu code   

[TRN]

No problem, everybody makes mistakes once and a while

Yep, pin 2 is connected to +3.3V, so the bus is parallel