[FIXED] Restoration of a new Philips PM3375 stored on a shelf for 25 years

[petercut62]

Hello all,
I'm Pierre, i live in Belgium and i love to fix old electronics stuff when it's possible.

I've found, few months ago, a PM3375 Philips oscilloscope from a auction site. It was stored brand new in his box on a shelf for 25 years. The daughter of the previous owner tried to power up the oscilloscope and guess what : lot of smoke.

As i found the service manual in english on a Dutch site for 35€, i've decided to bought the oscilloscope for 30€ and the service manual too.

Restoration time.
After a rapid check up, I discovered a burned capacitor on the power supply, completely toasted. I've tested all the capacitors and many of these were leaking.

So i've :
- Replaced faulty chemical capacitors
- Checked everything twice before power up

Finally the oscilloscope has booted up without any problem.

Rapid check : Everything is ok, but calibration is out, especially on the digital side.

Yesterday, i was calibrating the oscilloscope when i've touched the plastic film protection in front of the LCD display with my finger and i think that a ESD has killed the display or the controller or both ?

As you can see on the pictures some areas of the display are very dim and a snapshot of the reset screen shows some missing LCD signs on the display.

I'm not too aware with LCD. Do you think it will be possible to fix the LCD device by replacing the 3 PFC8577T on the LCD board (I've found some on ebay for a good price) ? Or should i search a LCD device from broken Philips oscilloscopes ?

Any advice would be appreciated.

Thanks
Pierre

[rf+tech]

Hello Pierre,

A year ago I restored a PM3335 with the same LCD. The scope always hung in self-test. After many days, I discovered five places where traces were open circuit due to corrosion from the leaking batteries, left to sit for many years. This damage was on the SRAM address lines very far from the batteries and occurred where the solder mask was thin or broken.

While attempting to replace the lamp in the LCD (which failed from disassembly/handling) I managed to damage the LCD driver chip(s), which resulted in some missing segments. I was equally surprised as you to find this display is so sensitive to ESD. Replacements from obtained from Digikey: http://www.digikey.com/product-search/en?keywords=568-1082-1-ND. I cannot find a definite answer to the difference between "C" and "T" versions.

The scope has been very pleasant to work with since repair.

Best wishes for success,

RF+ Tech

[rrinker]

 If the LCD connects to the board with an elastomeric connector (zebra stripe), try cleaning both is and the contact points on the LCD and board with isopropyl alcohol. Same if the LCD board connects to the main board in the same manner. It doesn't take much for these to get dirty and have poor conduction, resulting in dim or dead LCD segments. Especially if the thing smoked on power up at one point, and capacitors have been leaking all over the inside. There may be no defective chips, just poor contact.

[petercut62]

Hi guys,
Thank you for your replies.
The LCD was working fine before my touch of death !

I've tried to clean the board and the elastomeric connector ... unsuccessfully
I'm searching about the difference between "C" and "T" versions too. A bit tricky. May be full or half duplex. I don't know yet. I will call a specialist tomorrow.

@rf+tech : did you use a model C or T to fix your damaged LCD ?

Many thanks

Pierre

[rf+tech]

Pierre,

@rf+tech : did you use a model C or T to fix your damaged LCD ?

The PM3335 manual shows the model T in the parts list. I used the model C that Digikey stocks and the display has been correctly.

RF+ Tech

[petercut62]

Nothing better after changing the PCF8577 same pb

Do you think that the LCD might be broken or damaged inside ?

I'll try to find another LCD Module.

So if someone has a LCD module or screen for a PM 3365A or 67A or 75 or 77 let me know 

[rf+tech]

Pierre,

Sorry to learn the problem is still present.

First thing I would do is test the drive signal to some of the missing segments. No drive may point to bad data sent over I2C to the display. Something tells me the problem may be back on the main board. Can you work out a pattern between the missing segments and the PCF8577 driver pins that may reveal a "binary" problem, like exactly one half the output pins of each 8577 are affected?

Perhaps the real problem lies in the parallel to I2C serial presentation upstream. This is a long stretch and tedious task to rule out.

RF+ Tech

[petercut62]

@RF+Tech :
In use the digits are not missing they are very dimmed see the picture 1v/div for channel A is like ghosted. i can see the parameter values increase or decrease by pushing the buttons.
But when i reload the default settings, the lcd "reboot" showing some missing segments in the digits. the L must be an 8 (see screen 2.jpg)   

I'll do a bunch of tests
- check the voltage levels on the I2C command lines and the missing segments.
- As i have the protocol description and the characters table, i will try to send some I2C commands with an Arduino to test the LCD.
- Etc ...

Thanks for your support. I keep going...

[rf+tech]

Pierre,

We have a saying in the US that a picture is worth 1000 words. The photo of the very dim segments reminds me of a Fluke 12 that I bought cheaply because the rubber pushbuttons no longer worked. Many are familiar with dim DVM LCD segments and recommend disassembly for cleaning. Well...

This is the second Fluke 12 with the same failure of the elastomeric conductive strip. Carefully probing across the strip reveals resistance into the 100 kOhm range where new strips are seen to be less than 1 kOhm. In both Fluke 12 new strips have restored the button functionality.

As for cleaning. My belief is disassemby and cleaning only serves to remove microscopic conductive particles on the exposed edges. Consider also that separating these strips from a pcb/LCD where they have spent years of compression results in some of the conductive particles sticking to the pcb/LCD. Third, relieving years of compression causes some relaxation and separation of particles within the conductive channels. Peeling away strips also introduces stretching and more separation of particles.

This seems to be consistent when an LCD is removed for "cleaning" to fix dim segments, the very problem returns again and again. Replacement is to be preferred.

For a quick test, try exchanging the strips between the two sides of the LCD to see if the pattern of missing/dim segments changes.

RF+ Tech

[petercut62]

I've swapped the elastomeric strips. Same pb. In the service manual, the common 1 of the LCD seems to match with all the missing segments.
With a magnifying glass i've saw a little crack in the upper right corner of the LCD. I 've managed to disassemble the whole screen. The common 1 appears to be not fully connected to the driver because it is broken

I don't know what to do, Pin 1 = common 1 is half broken. i've found the missing part.

Might i glue the missing part and try to fix with conductive glue the other side of pin 1 ?
Or should i find a spare LCD ?

The ref : 5322 216 51207 LCD UNIT for PM 3350-52-65-67-75-77

[petercut62]

Update

I managed to find one LCD unit for 25€ at

http://www.skye-electronics.com/

A very interesting address in europe for Philips and Fluke equipment

 

[rf+tech]

Pierre,

Excellent troubleshooting. Yes, I would attempt to reattach the broken piece of glass. Cyanoacrylate, applied with a sewing needle to use the smallest possible drop, should be sufficient. The risk of too much CA covering the ITO on the main glass is the greatest concern. Better to practice making tiny drops on something else. The ITO conductor can be reconnected using a silver-bearing conductive ink pen, such as CircuitWorks Conductive Pen:

https://www.chemtronics.com/p-693-circuitworks-conductive-pen.aspx

The amount of pressure required to depress the tip in this pen will likely damage the glass piece. For better results, squeeze some ink onto a piece of polyethylene plastic (which is rather chemically resistant), thin the ink with a drop of MEK (methyl ethyl ketone, acetone does not work very well) and use a toothpick to apply the ink over the ITO. MEK evaporates quickly, another drop will soften the ink.

With a bit of skill and patience, there is a very good chance to make this LCD work again.

RF+ Tech

[petercut62]

Miracle, Patience, glue and more 

A picture is worth 1000 words

Thank you all for your support

[GreyWoolfe]

    Enjoy your scope.

[tino_so]

What a cool looking scope. I came across an old Phillips Frequency Counter in an auction (PM6669 120MHz/1.1GHz) a while back and I find Phillips did pay close attention to the User interface/front end layout. I hope your Scope will last you for a very long time.

But I just wanted to let you know that I have a Tektronix 2465B that I picked up in an auction as well, waiting to undergo a similar repair and your post has given me much needed encouragement and even some hope to carry on with it 

It will be my first Scope project of this size that I take on, so it feels a bit daunting at the moment. I have the Service Manual, and I finally started to take it apart. I will start documenting the process and perhaps I will be able to share the experience as you have  done, at some time down the road.

Thanks for sharing your experience Peter! Great job!

[ctln]

Hello Pierre,
I'm Catalin and I live in UK.I've got a Philips pm 3375 as well and I'm struggling to find a service manual.The digital part is not working properly or it does need calibration:I'm not sure.
How did you get yours?