Hitachi V680 oscilloscope water damage repair

[gkmaia]

I got a V680 with some issues. But going over the service manual the TP signal plots are in terrible resolution. Pretty much useless... will be too hard to fix it without the signals.

Has anyone ever come a cross a V680 manual with better resolution where these graphics can be read?

[Chris56000]

Hi!

The trouble with that manual was the waveforms were originally printed in a shade of pale blue and unless you're exceptionally careful at scanning the thing, details like this simply vanish!

The Hitachi scope Service Manuals were not unfortunately, packed in with these scopes when new and what few original colour–printed originals there were ended up in calibration houses who refuse point blank to let them outside their premises, so the chances of getting one are almost nil!

If you've got a fault you're stuck with, please post as much as you can about it and I can help you go thro' it without the waveform details!

The only other possibillity is if another Member has a V680 in good working order & can retake the waveforms again – (I can't afford one myself or I could help with this!) – the arrows pointing to each part of the circuit they were taken are clear enough!

Chris Williams

[gkmaia]

Thanks Chris. That would be awesome.

It only produces a dot. So I suspect the sweep board. The other boards may be able to produce the signals so I can update the manual with those.

When it is all fixed I can update the missing graphics save a new PDF and share it here.

[Chris56000]

Hi!

If you're only getting a dot, this could be the timebase stalled, the auto–trigger circuit not working or failure of the horizontal  (X) deflection amplifier!

We need to begin with the X amplifier to begin with, so first of all try the following simple test:–

Connect one end of a 10k resistor to the positive terminal of a 9V battery, connect the free end of the resistor to the centre of the CH2 input of your 'scope, then connect the negative side of the battery to the outer side of the ch2 input, switch ch2 to 2V/div, and operate the "XY" function – if the X amplifier after the XY pickoff is operational, the spot should move about five divisions to the right.

If this test passes, everything after the XY pickoff and before the X deflection plates is operational!

If there is no movement, either the XY selection or the X amplifier has failed, so please try this test first and get back, whilst I run off printouts of the schematics and study them for you!

Chris Williams

PS!

How anybody could make such an absolutely atrocious bloody job of scanning what was originally a high-quality colour-printed document is beyond my comprehension - I did get to see an original factory Hitachi-Denshi scope service manual (not this model unfortunately) about 8 years ago, and I know they were very high-quality documents produced with coloured pcb layouts and coloured waveform diagrams!

The only hope of being able to resurrect the information lost in that atrocious scan would be to have one of those scopes, de-populate and scan all the PCBs, trace over them or scan both sides of the bare PCBs, then take and draw all the original waveforms again from a working model!

[Chris56000]

Hi!

Another quickie tip – most older analogue oscilloscopes I've used that have a multi–position rotary timebase switch nearly always produce a slight movement of the spot horizontally whilst switching thro' between the speeds due to the charging current of the timing capacitors producing a small voltage pulse as each time–constant is changed by operation of the TB speed–select switch, if you see this movement of the spot, then the signal chain from the Miller rundown to the X plates is basically complete!

Chris Williams

PS!

Have you tried operating all the front–panel TB controls back and forth rapidly over their operating ranges to see if any of them produce any horizontal pulses or movements of the spot, or operation/partial operation of the TB in one or more positions? If there's a dicky switch or variable pot amongst the customer TB & trigger functions then this test will help show which one it is!

If the spot is a total fixture with no control producing any movement whatsoever,  then I'd definitely be thinking X amplifier failure!

[gkmaia]

Hi Chris,

Not yet. I will probably do more towards the end of the week.

[Chris56000]

Hi!

Please keep us updated with all your findings!

Chris Williams

[gkmaia]

Hi Chris,

Did not forget about the project. I went to pickup the V680 and the seller has misplaced one of the covers. I got to wait till he finds it to pick it up. Keep you posted.

[Chris56000]

Hi!

No worries – I've printed out what there is of the circuit diagrams for study when I get back from holiday this week, and I'll post again when you've got the 'scope setup for repair and you've updated me on what's working and what isn't!

Chris Williams

[gkmaia]

I got the scope yesterday. The previous owner said there was a leak in his house and some water got through the top vents.

First thing I did was to check how bad it was. Just one board got water. All other boards are fine. Not rust at all.

The exposed board was the A/B sweep generator board. I attached some photos. Although it looks bad I probed the worst ones and could not find one broken trace initially. Seems the rust is still superficial. But before turning it on I would like to take care of the oxidation best way possible given the complexity of the board.

As you can see on the dark spots the oxidation created pocket between the copper and the UV mask. When traces are severed or the board is not so packed I would usually mechanically remove the UV mask to expose the pockets, clean the copper to bare, then do the repair and apply a new coat of uv mask.

But in this case there a few things to consider:
- way too many traces
- traces are too small and delicate
- removing the uv mask will remove the components markings

What would you do?

Leave the oxidation pockets and hope for the best long term and just fix the broken traces if I find any?

Other option would be to use the WD rust remover soak as per the post bellow. Then we do the repair. When repair is done remove pots and connectors and do a dip of conformal coating. Huge work... what you think? Overkill? Effective?

https://www.eevblog.com/forum/repair/repairing-corroded-pcb/
https://au.rs-online.com/web/p/conformal-coatings/1558156/

[gkmaia]

The pot knob of one of the vertical controls was broken. Had to machine a new shaft and fit it. Turned out quite good!

I need to do just a few small repair on that rusted PCB then will put it together and make a video on how it goes electronically speaking.

[gkmaia]

Hi!

No worries – I've printed out what there is of the circuit diagrams for study when I get back from holiday this week, and I'll post again when you've got the 'scope setup for repair and you've updated me on what's working and what isn't!

Chris Williams

Initial test done Chris!

Most of the controls on the sweep board work. These are the first two rows of pots and switches from top to bottom.

[Chris56000]

Hi!

That's looking very promising - the full amount of X - scan amplitude is there and the readout figures are clear and not squashed or stretched in the horizontal direction, and the x10 mag appears to work as expected, so the X amplifier output stage is operating correctly and reasonably accurate regarding it's gain, and the sweep generator circuits are responding to the auto-trigger generator!

It wasn't immediately apparent your instrument had suffered water-damage from the initial post as your query was about the manual initially!

All I can recommend is to carry on looking as carefully as you can at the top (component) side of the pcb for any further signs of water ingress, and remove all components around the most badly affected copper. Any components showing any signs of green corrosion on their leads should, ideally, be replaced with new if possible!

If there is suggestions water/liquid might have got into any of the multi-contact selector switches, you might gave to remove the PEF-711 pcb from the oscilloscope so you can get in their contacts with De-oxit, etc!

One thing I can give you confidence in, from the video you provided, is that the readout control generator/micro circuits appear to have been unaffected by the damage, but please free to remove and inspect the PEF-770 Readout PCB if you wish!

As it's not immediately obvious from your video, please could you re-post a list of which functions or switch positions still seem to be inoperative or otherwise faulty, and I'll post exactly which components to look for & check!

Chris Williams

PS!

Did you use a.c. coupling to take the wave-forms A to H on the circuit-extracts you provided? Not withstanding this, apart from Waveform B which is inverted because of the particular switch settings you chose (triggering on -ve slope) they correspond almost exactly with the totally illegible remnants of what was printed on the original drawings so I think virtually everything on the Sweep Generator(2) diagram appears to be functioning perfectly!

PPS!

Do you have a 'scope with a "HF trig" or a "TV Line" sync. function? Oscilloscopes offering "TV Line" sync. functions are designed to trigger on the types of waveform you've annotated as "hard to sync properly" - failing that Hitachi might have set up a storage-type scope to record them!

[gkmaia]

Did update the schematic images on the previous post with latest scans. Added some comments to each graph (as per the image example bellow) with info about what controls affected the wave form and how. If a control is not documented there it is because it had no affect on the wave form. AC/DC/gnd had no effect in any measured wave form so far.

PEF770 and PEF711 are intact. Not a sign of oxidation.

What I would like to do before removing any components it is to identify which block of the topology may be faulty to then remove the components around that area and inspect. Reason being by removing the switches to check rust under them I noticed the board is very sensible to heat. It delaminates very easily. I would like to keep it to a minimum if possible.

Videos of points using a TV sync scope:

N


J


K

[gkmaia]

I could finally manage to get the scope to trigger a signal. Only channel 2 is working at the moment and with more attenuation than normal.

But at least the trigger board is now operational!

[Chris56000]

Hi !

Good show, it looks like you've got most of it working now!

I've now been able to buy a V680 at a price I could afford, so I will start writing a new colour–published Combined Instructions & Service Manual from the scope itself and what data is readable from the atrocious bloody Elektrotanya scan!

This will be a long project tho' that will take several months to complete!

Chris Williams

[gkmaia]

Nice!!!

I still have mine and it is 100% working now. In the end was just one faulty IC.

[andy3055]

When it comes to scanning diagrams with very lightly colored areas such as PCB traces and screened areas, the problem is that most scanners of those days did not pick them well. There was a time when copy machines would not even print light color like blue at all. These technologies have now improved so much that if the same diagrams are scanned today, on most machines the traces will be picked up quite sharp.

[gkmaia]

yes, it is a pain to fix these things without some plots... the OS1100 I am working on now has a few voltages here and there and that is it...

I've got two new project now.

A philips 3543 logic scope that had a faulty logic board but it is now fixed.

And a Gould Advance OS1100 that has multiple issues. Power supply that is now fixed. Missing rear panel, fabricated a new one. Vertical amp that is now fixed. But still to manage it to sweep... has been quite a challenge.