Author Topic: Took the Plunge - GoldStar OS-7020 Oscilloscope  (Read 7944 times)

0 Members and 1 Guest are viewing this topic.

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #25 on: January 21, 2020, 06:13:27 am »
.... is this a DC voltage (service manual doesn't say)

You can safely assume any voltages specified in circuit diagrams will be DC - unless they are very clearly annotated otherwise.  Power supply voltages tend to be the safest bet as being DC.
« Last Edit: January 21, 2020, 06:15:03 am by Brumby »
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #26 on: January 21, 2020, 06:40:21 am »
If it gives you a range of -1,995v to -1,805v and you're seeing only -1,650V that's way low, I'm not sure offhand why it's low but that definitely isn't right.
 

Offline olafwagnerTopic starter

  • Contributor
  • Posts: 17
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #27 on: January 22, 2020, 08:49:20 pm »

I removed Q621 from the board and popped it into my component tester, and it recognized it as being a BJT-NPN transistor (see attached).

Unfortunately I don't know what the hFE, Ube and Ic values mean, but considering the tester recognized this as a NPN transistor, does it still mean that Q621 is bad?

I googled for a data sheet, and found one for a Hitachi 2SC1907, and also attached some of the electrical characteristics below (unfortunately I don't know what to look into)
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #28 on: January 22, 2020, 10:33:43 pm »
Those characteristics are useful if you are designing something around a transistor or trying to find a suitable sub but that isn't important right now. You are correct in assuming that if the tester accurately identifies it as an NPN transistor it is almost certainly just fine. 99% of the time a fault will clearly manifest itself as a short or open between two or more pins.
 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4931
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #29 on: January 23, 2020, 01:48:46 pm »
Agreed, normally the failure mode is a fairly clear open or short.

Does seem strange that the blanking signal ends there though...
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 

Offline olafwagnerTopic starter

  • Contributor
  • Posts: 17
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #30 on: January 23, 2020, 09:48:52 pm »

I have been prodding around the board some more measuring voltages, but I must admit I am at a bit of a loss.

Here is a list of Summarized List of Signs and Symptoms:
- After switching on, it takes 30-35 seconds for the trace to appear (I would expect 10-15 to be more typical, this seems long)
- HV Rail measures -1,650v. (Should be in the range of -1,995v to -1,805v.)
- Intensity Pot (and internal trimmer) cannot get the trace to disappear. (It seems to alter the brightness ever so slightly, but not even close to disappearing.)
- Blanking Signal not getting to the Blanking Amp
- Trace is a 'dash' and not a 'dot' (very obvious in low-speed sweeps or XY mode)


I tried to make some progress on any of these issues by doing the following:

1. Blanking Signal
- I measured a lot of voltages around the blanking amp trying to figure out why the blanking signal never gets 'through' Q621. (all measured voltages appear in red in the attachment below)
- I also pulled Q622 off the board and popped it into my component tester, and it is also recognized correctly.
- I noticed in the Hitachi spec that they also have an oscilloscope diagram for what I assume is the output of the blanking amp (before it hits high-voltage), which I am calling "P3". When I attached my scope there I only get a flat signal which I have measured to be 13.5V
- I also tried to measure diode CR621 (1SS83) in circuit with my DMM in continuity mode, but I do NOT get a 'beep' in either direction. Should I get a beep? Or must these be removed from the board and plugged into a component tester to test?
- Lastly I ordered a couple of replacement 2SC1907's from Ebay, so I could simply replace Q621 and Q622 and see whether that does anything.

2. Intensity Controls
I previously measured all the voltages around the INTEN pot (and included the measured voltages on the schematic)
- Today I measured all the voltages around the internal INTEN ADJ. trimmer. I assume they appear correct. I stopped short of diode CR516, because that is inside the high voltage area of the board that has a nice protective shield over it.


3. HV Rail
Here I am seriously out my depth (and do not want to electrocute myself) so I can only do very little testing.
- I measured a few voltages on the input side that goes to the HV transformer (just to make sure the 18v rail was getting here properly)


The 'simplest' issue that I need to progress on is the blanking signal, so my only hope there at the moment is simply replacing Q621 and Q622 and see what happens.




 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #31 on: January 23, 2020, 10:32:23 pm »
If those are HV diodes then the forward drop will be too high to measure with the diode check on a DMM.

Something else that would be worthwhile to test is the CRT itself, check to make sure there are no shorts between elements that should not be connected.
 

Offline olafwagnerTopic starter

  • Contributor
  • Posts: 17
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #32 on: January 24, 2020, 02:25:52 am »

OK, things just got weird.

I wigged the CRT socket connector a bit (to make sure it is properly connected, and reinstalled transistor Q622, reinstalled the horizontal board into the case and turned it on.

Now all of a sudden I have a VERTICAL problem. The trace is making a vertical line that is the thickest at the center graticule, and if I fiddle with vertical position, only the length of the thin line at the top gets longer and shorter. The fatter part remains in the center of the screen.

Also (and more alarmingly), I re-measured my HV rail, and now I get -2,302v!

What the heck is going on.
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #33 on: January 24, 2020, 03:30:59 am »
That's odd, sounds like the CRT socket and associated wiring needs a careful inspection.
 

Offline olafwagnerTopic starter

  • Contributor
  • Posts: 17
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #34 on: January 31, 2020, 08:51:17 pm »

I carefully unplugged and removed the CRT, and checked all of the pin combinations. The only combination that I got a 'continuity beep' on were pins 1 & 14, which I believe is the heater, so I assume that is correct. Also, with nothing visibly wrong with the CRT and nothing rattling around inside, I carefully installed it again.

I checked the CRT socket board and all the pins and harness, but I cannot find anything wrong here. (All continuity beeps out, nothing seems shorted)

I then turned my attention to the HV circuit (because something there must be causing this spike to -2,302v. I opened up the cover to the HV section, and checked all the resistors (they all check out fine within spec), but noticed resistor R697 was missing. I then removed the PCB and looked underneath.

Firstly I discovered that I am certainly not the first here. There is some liquid residue on the board, which I assume is flux that someone before me didn't clean off. It also looks like a lot of the solder points may have been remelted (perhaps whoever had this scope before me assumed there were dry solder joints.

Secondly I found resistor R697 installed underneath (perhaps by the same person that left the other evidence). However, even though the schematic calls for a 3M resistor, there is a 1M resistor installed. This resistor is part of the FOCUS circuit, and the resistor on the other side is a 16M, so assume the 1 vs 3M won't make a huge difference, and is probably not related to my sudden high voltage issue.

Should I remove and test all the 2KV and 3KV ceramic capacitors and test them? (I assume Ceramic caps are very reliable?) Should I remove and test the diodes? (I assume these are far too high voltage to be able to 'beep' in circuit)

Also, what role do the 2 Neon Lamps (NL601 and NL602, type NE-68 which is either an old or unobtainium model) play? I see they are wired across the Blanking and Cathode rails (in series with one another), is this some kind of voltage regulation? How would I go about checking them? Should I see them 'glow'? (because I don't)

 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #35 on: February 01, 2020, 12:40:09 am »
Ah so it sounds like you're dealing with someone elses failed repair, you're going to have to look very carefully because there's no telling what else they messed with in there. I would change out that 1M resistor for the proper 3M or three 1M resistors in series, that could be increasing the load significantly on the HV which is not able to supply much current. Also clean off the flux, some types of flux are conductive and that also can cause problems with high voltage and high impedance circuits.

The neon lamps are likely functioning as spark gaps, their purpose is to protect the electronics in case an arc occurs in the CRT. It's fairly common to see that arrangement on older CRT displays.

 
The following users thanked this post: AVGresponding

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #36 on: February 02, 2020, 10:59:21 am »
Hope this will help you out.  (I don't think I missed anything  :D )

https://www.dropbox.com/s/rybj78uwiksgupo/Goldstar_OS-7020_Service_Manual.pdf?dl=0

File size is around 19MB - but I feel this is on the high side.  If anyone can recommend some PDF wrangling software that might deliver a smaller file size, I'd be interested.

I've presented it as it was in the hardcopy, so viewing on screen should be fine, but printing it out as is might not give you what you expect.

PDF has been updated.  It is now searchable and just under 11MB in size.  The new version has just been uploaded to KO4BB and Dropbox.

Dropbox link is: https://www.dropbox.com/s/m8robnt2kxczc9g/Goldstar_OS7020_Oscilloscope_Service_Manual_searchable.pdf?dl=0


In a moment of weakness I spent some real money on PDF software.
 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4931
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #37 on: February 02, 2020, 04:20:15 pm »
Hope this will help you out.  (I don't think I missed anything  :D )

https://www.dropbox.com/s/rybj78uwiksgupo/Goldstar_OS-7020_Service_Manual.pdf?dl=0

File size is around 19MB - but I feel this is on the high side.  If anyone can recommend some PDF wrangling software that might deliver a smaller file size, I'd be interested.

I've presented it as it was in the hardcopy, so viewing on screen should be fine, but printing it out as is might not give you what you expect.

PDF has been updated.  It is now searchable and just under 11MB in size.  The new version has just been uploaded to KO4BB and Dropbox.

Dropbox link is: https://www.dropbox.com/s/m8robnt2kxczc9g/Goldstar_OS7020_Oscilloscope_Service_Manual_searchable.pdf?dl=0


In a moment of weakness I spent some real money on PDF software.

Please excuse me if I'm missing something obvious here, but couldn't you just use a compression tool on the file?
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 

Offline Whales

  • Super Contributor
  • ***
  • Posts: 2122
  • Country: au
    • Halestrom
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #38 on: February 03, 2020, 04:23:36 am »
[...]  If anyone can recommend some PDF wrangling software that might deliver a smaller file size, I'd be interested.
Please excuse me if I'm missing something obvious here, but couldn't you just use a compression tool on the file?

Compression background

To be able to answer both of these questions: some background about compression is needed.

There are lossless compression techniques (eg .zip, .rar, .gz, .lzma, .7z, png) and lossy compression techniques (jpeg, most video codecs [h264, mpeg*], most audio codecs [opus, mp3, aac]).   Lossy is a lot better at saving space, but it sacrifices some details.

Lossy is pretty much a default expectation of modern computing and society,  with lossless only used where any loss would be inappropriate (eg text files, written documents).

Lossy compressors are very situation specific so I won't go into their details here.  Most have variable control: more compression (more detail loss) or less compression (less detail loss).  You may have encountered this when saving as JPEG.

Lossless compressors work by looking for repeating patterns in your files and replacing these repeating patterns with just one copy (plus some references).  This is how archive formats like .zip, .rar and .7z work.

Regardless of what compression technique you use: the data bits and bytes in the output compressed file should look mostly indistinguishable from random data.  This means that there are no obvious patterns left that can be exploited for further compression.  ie compressing a file multiple times will not make it smaller (with some technical exceptions), only one layer of compression should be applied or needed (again with some weird exceptions, not relevant here).

Initial inspection

Now let's look at the specific PDF file: Goldstar_OS-7020_Service_Manual.pdf (19MiB).  I'll start my inspecting by extracting all of the images using pdfimages (part of poppler-utils):

Code: [Select]
$ pdfimages ../GoldStar.pdf -all ex
$ ls
ex-000.jp2  ex-039.jpg  ex-078.jpg  ex-117.jpg  ex-156.jpg  ex-195.jpg  ex-234.jpg  ex-273.jpg
ex-001.jpg  ex-040.jpg  ex-079.jpg  ex-118.png  ex-157.jpg  ex-196.jpg  ex-235.jpg  ex-274.jpg
ex-002.jpg  ex-041.jpg  ex-080.jpg  ex-119.jpg  ex-158.jpg  ex-197.jpg  ex-236.jpg  ex-275.jpg
ex-003.jpg  ex-042.jpg  ex-081.jpg  ex-120.jpg  ex-159.jpg  ex-198.jpg  ex-237.jpg  ex-276.jpg
ex-004.jpg  ex-043.jpg  ex-082.jpg  ex-121.jpg  ex-160.jpg  ex-199.jpg  ex-238.jpg  ex-277.jpg
ex-005.jpg  ex-044.jpg  ex-083.jpg  ex-122.jpg  ex-161.jpg  ex-200.jpg  ex-239.jpg  ex-278.jpg
ex-006.jpg  ex-045.jpg  ex-084.jpg  ex-123.jpg  ex-162.jpg  ex-201.jpg  ex-240.jpg  ex-279.jpg
ex-007.jpg  ex-046.jpg  ex-085.jpg  ex-124.png  ex-163.jpg  ex-202.jpg  ex-241.jpg  ex-280.jpg
ex-008.jpg  ex-047.jpg  ex-086.jpg  ex-125.jpg  ex-164.png  ex-203.jpg  ex-242.jpg  ex-281.jpg
ex-009.jpg  ex-048.jpg  ex-087.jpg  ex-126.jpg  ex-165.jpg  ex-204.jpg  ex-243.jpg  ex-282.jpg
ex-010.jpg  ex-049.jpg  ex-088.jpg  ex-127.jpg  ex-166.jpg  ex-205.jpg  ex-244.jpg  ex-283.jpg
ex-011.jpg  ex-050.jpg  ex-089.jpg  ex-128.jpg  ex-167.jpg  ex-206.jpg  ex-245.jpg  ex-284.jpg
ex-012.jpg  ex-051.jpg  ex-090.jpg  ex-129.jpg  ex-168.jpg  ex-207.jpg  ex-246.jpg  ex-285.jpg
ex-013.jpg  ex-052.jpg  ex-091.jpg  ex-130.jpg  ex-169.jpg  ex-208.jpg  ex-247.jpg  ex-286.jpg
ex-014.jpg  ex-053.jpg  ex-092.jpg  ex-131.jpg  ex-170.jpg  ex-209.jpg  ex-248.jpg  ex-287.jpg
ex-015.jpg  ex-054.jpg  ex-093.jpg  ex-132.jpg  ex-171.jpg  ex-210.jpg  ex-249.jpg  ex-288.jpg
ex-016.jpg  ex-055.jpg  ex-094.png  ex-133.jpg  ex-172.jpg  ex-211.jpg  ex-250.jpg  ex-289.jpg
ex-017.jpg  ex-056.jpg  ex-095.jpg  ex-134.jpg  ex-173.jpg  ex-212.jpg  ex-251.jpg  ex-290.jpg
ex-018.jpg  ex-057.jpg  ex-096.jpg  ex-135.jpg  ex-174.jpg  ex-213.jpg  ex-252.jpg  ex-291.jpg
ex-019.jpg  ex-058.jpg  ex-097.jpg  ex-136.jpg  ex-175.jpg  ex-214.jpg  ex-253.jpg  ex-292.jpg
ex-020.jpg  ex-059.jpg  ex-098.jpg  ex-137.jpg  ex-176.jpg  ex-215.jpg  ex-254.jpg  ex-293.jpg
ex-021.jpg  ex-060.jpg  ex-099.jpg  ex-138.jpg  ex-177.jpg  ex-216.jpg  ex-255.jpg  ex-294.jpg
ex-022.jpg  ex-061.jpg  ex-100.jpg  ex-139.png  ex-178.jpg  ex-217.jpg  ex-256.jpg  ex-295.jpg
ex-023.jpg  ex-062.jpg  ex-101.jpg  ex-140.jpg  ex-179.jpg  ex-218.jpg  ex-257.jpg  ex-296.jpg
ex-024.jpg  ex-063.jpg  ex-102.jpg  ex-141.jpg  ex-180.jpg  ex-219.jpg  ex-258.jpg  ex-297.jpg
ex-025.jpg  ex-064.jpg  ex-103.jpg  ex-142.jpg  ex-181.jpg  ex-220.jpg  ex-259.jpg  ex-298.jpg
ex-026.jpg  ex-065.jpg  ex-104.jpg  ex-143.jpg  ex-182.jpg  ex-221.jpg  ex-260.jpg  ex-299.jpg
ex-027.jpg  ex-066.jpg  ex-105.jpg  ex-144.jpg  ex-183.jpg  ex-222.jpg  ex-261.jpg  ex-300.jpg
ex-028.jpg  ex-067.jpg  ex-106.jpg  ex-145.jpg  ex-184.jpg  ex-223.jpg  ex-262.jpg  ex-301.jpg
ex-029.jpg  ex-068.jpg  ex-107.jpg  ex-146.jpg  ex-185.jpg  ex-224.jpg  ex-263.jpg  ex-302.jpg
ex-030.jpg  ex-069.jpg  ex-108.jpg  ex-147.jpg  ex-186.jpg  ex-225.jpg  ex-264.jpg  ex-303.jpg
ex-031.jpg  ex-070.jpg  ex-109.png  ex-148.jpg  ex-187.jpg  ex-226.jpg  ex-265.jpg  ex-304.jpg
ex-032.jpg  ex-071.jpg  ex-110.jpg  ex-149.jpg  ex-188.jpg  ex-227.jpg  ex-266.jpg  ex-305.jpg
ex-033.jpg  ex-072.jpg  ex-111.jpg  ex-150.jpg  ex-189.jpg  ex-228.jpg  ex-267.jpg  ex-306.jp2
ex-034.jpg  ex-073.jpg  ex-112.jpg  ex-151.jpg  ex-190.jpg  ex-229.jpg  ex-268.jpg  ex-307.jp2
ex-035.jpg  ex-074.jpg  ex-113.jpg  ex-152.png  ex-191.jpg  ex-230.jpg  ex-269.jpg  ex-308.jp2
ex-036.jpg  ex-075.jpg  ex-114.jpg  ex-153.jpg  ex-192.jpg  ex-231.jpg  ex-270.jpg
ex-037.jpg  ex-076.jpg  ex-115.jpg  ex-154.jpg  ex-193.jpg  ex-232.jpg  ex-271.jpg
ex-038.jpg  ex-077.jpg  ex-116.jpg  ex-155.jpg  ex-194.jpg  ex-233.jpg  ex-272.jpg

It looks like this file is nothing but jpegs and the occasional jp2.  That means to further (re)compress this file: what we really need to do is focus on the jpegs.  The PDF is nothing but a wrapper around them.  This is basically always the case for scanned PDFs.

We have several options to reduce the file sizes of images:
  • Reduce their size (resolution)
  • Reduce their bitdepth (number of colours)
  • Increase their compression level (throw out more data during lossy/jpeg compression)
  • Change the lossy codec to something else (eg from jpeg to jp2)

Sidenote: file quirks & better software

Unfortunately for us each page of this PDF is not a single jpeg.  Each page has been split into multiple jpegs that are aligned together in a grid:





Some software does this, it's quite annoying :P

To work around this: we're going to use the imagemagick tools from now on.  These are smart enough to treat each page as a single image.  This particular software is very common and well known in the *nix and web-development worlds (many websites use it behind the scenes for image processing), but not well known in the Windows world, even though there is a Windows version available.  The world is weird, this software is the duck's guts for anything that requires mass-editing of multiple files simultaneously.

My method

(1) Remove the very first and last pages.  It's pretty to see the paper texture from the original manual cover, but it's all undulating and hard to compress.  The second page of the PDF is the same as the cover any, just in black and white.

(2) Convert everything to greyscale.  There is no point having colour anywhere in this document, so let's not make the compressor (eg jpeg) think it has to preserve it.  Strip it all out.

(3) Remove ghosted text from other pages by fiddling with brightness/levels.  I'm talking about this stuff (emphasized via editing):



You can't (normally) see this and it's useless, so let's get rid of it.  The fact it's still there in the image means the last person's compressor was wasting space trying to keep it.

(4) Avoid reducing the resolution.  Resolution is nice, esp in technical documents.  There's nothing worse than a blurry technical diagram.  The whole point of these documents is to make people happy, not grumpy.

(5) Try several different compressors (jpeg, jpeg2000, png, etc).  PDFs can house quite a few different formats fine.


First attempt: adjust levels, greyscale, jpeg
Code: [Select]
$ mkdir temp1
$ magick convert -density 100 Goldstar_OS-7020_Service_Manual.pdf -set colorspace Gray -level '25%,75%,0.3' -quality 70 'temp1/page%03d.jpeg'
$ rm  temp1/page000.jpeg  temp1/page099.jpeg
$ du -sh  temp1/
9.1M

9.1MiB is not bad, but we can be squeezier if we're clever.

Second attempt: adjust levels, greyscale, png
Code: [Select]
$ convert -density 100 Goldstar_OS-7020_Service_Manual.pdf -set colorspace Gray -level '25%,75%,0.4' 'temp2/page%03d.png'
$ rm temp2/page000.png temp2/page099.png
$ du -sh temp2/
12M

Eep, wrong way.  This codec doesn't seem useful with this sort of image data.

... but what if we massage the image into something that png would actually like?  Eg black and white (no greyscale)?

Third attempt: threshold into black and white, png
Code: [Select]
$ convert -density 100 Goldstar_OS-7020_Service_Manual.pdf -set colorspace Gray -alpha off -auto-threshold OTSU 'temp3/page%03d.png'
$ rm temp3/page000.png temp3/page099.png
$ du -sh temp3/
1.7M

:D

I think the resolution is a bit low however, so I up it to -density 200 and try again:

Code: [Select]
$ du -sh temp4
3.8M

Not as good filesize wise, but the diagrams are a lot more readable.
« Last Edit: February 03, 2020, 05:19:01 am by Whales »
 
The following users thanked this post: rsjsouza

Offline Whales

  • Super Contributor
  • ***
  • Posts: 2122
  • Country: au
    • Halestrom
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #39 on: February 03, 2020, 04:26:31 am »
Turning the individual image pages back into a single PDF

I think it's possible to use imagemagick to do this, but you have to tell it not to recompress the files a second time.  Instead I used img2pdf (probably not easy to get on Windows):
Code: [Select]
$ img2pdf temp3/* -o Goldstar_OS7020_servicemanual_tiny.pdf
$ img2pdf temp4/* -o Goldstar_OS7020_servicemanual_reasonable.pdf

The results are attached.

I have not spent enough time quality checking: some of the large diagrams have unreadable numbers.  It's also hard to differentiate grid lines and signal on the scope shots; perhaps thresholding in black-and-white wasn't the best option here.

Opinion

Don't throw out the original scans.  Keep both the original scans PDF and a small PDF (like mine) side by side and never separate them.  One for convenience, the other for when the convenient version screws up (unreadable detail).  You can't recover data that has been thrown out.

Text searchability

Also known as OCR (optical character recognition).  Never 100% accurate but useful.  Not going into it here :P  sounds like you already have something that works.

I want a program that can do all of this for me!

No such program will ever exist, just as there will never be a program that writes the manual for you.  Every step of the above has required human analysis and choosing between one of many options.

There will be programs that will claim to compress PDFs, but their performance will be hit and miss because they will guess at what is best to do.  None of these programs try to read the circuit diagrams afterwards and determine if they are readable or not.  None of these programs can tell that ghosted text from other pages is not useful information.  And so on.

We are after all trying to compress against human perception, not machine perception.
« Last Edit: February 03, 2020, 04:40:59 am by Whales »
 
The following users thanked this post: rsjsouza

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #40 on: February 03, 2020, 06:18:55 am »
It's amazing how you can ask for suggestions and get ignored - but put something out there that isn't up to some people's expectations and watch all the experts chime in.

Yes, I know there was colour.  Yes, I know there was some text visible through the back of the page.  Yes, I could have reduced it to greyscale - but my efforts took long enough as it was, scanning each page separately, keeping things as square as possible and as clear as possible, stitching together pages that had to be done in two passes because they were too big for the scanner.

Considering that this was my first foray into PDF wrangling, the result seems useful enough.  I've already admitted it isn't optimal.


Here's the original scan that came straight off the scanner...
https://www.dropbox.com/sh/dqj8qe9jvfodvfz/AACdycgDv4HgFte3aI7KPHLra?dl=0&preview=Goldstar_OS7020_ServMan_OriginalScan.PDF
(There are a couple of pages that were double scanned.)

Knock yourself out.

Edit: Bit of a rant there.... sorry.
« Last Edit: February 03, 2020, 06:47:16 am by Brumby »
 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4931
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #41 on: February 03, 2020, 10:28:34 am »
Haha no worries.

I know the basics regarding lossy and lossless compression; I recently completed an HNC with the OU that included these topics.

I just wasn't aware of the quality of the source material you were working with, that's quite an effort you put in there!
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 

Online rsjsouza

  • Super Contributor
  • ***
  • Posts: 6107
  • Country: us
  • Eternally curious
    • Vbe - vídeo blog eletrônico
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #42 on: February 03, 2020, 11:43:19 am »
Whales, your posts above are pretty well written indeed, thanks for taking the time.

One question I have: does the process of removing the background shade performs yet another decompression/recompression? I suspect not, but wanted to ask.

As for the utilities, I usually use IrfanView and Foxit for this handling. They tend to do quite a reasonable job, but I will check these tools as well.

In my opinion, however, Brumby's grayscale version looks much more appealing to my eyes than your two versions above.

When I got my Keithley 191, its online manual at Tektronix was scanned in BW with very poor detailing overall, with blocky graphics (due to the threshold in BW conversion) and other issues. I got a physical copy, scanned it in uncompressed 300dpi TIFF (schematics and mechanicals at 600dpi) and worked around to compress it to a more reasonable size. Sure, the file size is much larger than theirs (26MB) but disk space is cheap, internet connections are improving and there are no continuous upgrades, thus making it a one time process.

One thing I may do to this and another manual I posted there is the removal of the shading of the other pages - a brightness setting on the scanner would have minimized/solved this.
Vbe - vídeo blog eletrônico http://videos.vbeletronico.com

Oh, the "whys" of the datasheets... The information is there not to be an axiomatic truth, but instead each speck of data must be slowly inhaled while carefully performing a deep search inside oneself to find the true metaphysical sense...
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12413
  • Country: au
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #43 on: February 03, 2020, 12:33:04 pm »
.... was scanned in BW with very poor detailing overall, with blocky graphics (due to the threshold in BW conversion) and other issues.

This is why I avoid threshhold processing for text.  The intermediate intensity pixels add clarity when viewed by our eyes, which - in conjunction with the human brain - have developed algorithms to see things much more clearly.

Reducing the ghosting from the back of the pages would have been useful - so would converting from a colour palette to greyscale.
 
The following users thanked this post: rsjsouza

Online rsjsouza

  • Super Contributor
  • ***
  • Posts: 6107
  • Country: us
  • Eternally curious
    • Vbe - vídeo blog eletrônico
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #44 on: February 03, 2020, 03:06:32 pm »
.... was scanned in BW with very poor detailing overall, with blocky graphics (due to the threshold in BW conversion) and other issues.

This is why I avoid threshhold processing for text.  The intermediate intensity pixels add clarity when viewed by our eyes, which - in conjunction with the human brain - have developed algorithms to see things much more clearly.

Reducing the ghosting from the back of the pages would have been useful - so would converting from a colour palette to greyscale.
Yes, that is the effect also brought by some of these subpixel rasterizing methods. Interesting that Microsoft's ClearType uses RGB, not only greyscale, but I imagine that greyscale is already good enough for these manuals.
Vbe - vídeo blog eletrônico http://videos.vbeletronico.com

Oh, the "whys" of the datasheets... The information is there not to be an axiomatic truth, but instead each speck of data must be slowly inhaled while carefully performing a deep search inside oneself to find the true metaphysical sense...
 

Offline Whales

  • Super Contributor
  • ***
  • Posts: 2122
  • Country: au
    • Halestrom
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #45 on: February 04, 2020, 09:33:05 pm »
One question I have: does the process of removing the background shade performs yet another decompression/recompression? I suspect not, but wanted to ask.

Yes it does.

Original jpegs -> decompress into raw (in-memory) pixels -> adjust levels to remove background ghosts -> compress into new jpeg

There are tools that can edit JPEGs directly (ie without decompression), but they're pretty rare and would be limited.  It should be theoretically possible to keep most of the JPEG DCT blocks and throw out just the ones you don't want, ie 'lossless' editing of a lossy format, but I'm not aware of any easy ways. 

The only mainstream/common lossless JPEG editing available is image rotation, but even then some tools stuff it up and re-compress the image anyway.  You only need to edit some metadata for this one.

The saving grace is that I destroyed a lot of other detail anyway, so having an extract decompress + recompress really didn't do much harm :P

Quote
In my opinion, however, Brumby's grayscale version looks much more appealing to my eyes than your two versions above.

I agree as well. I quested for a compression method small enough to attach the PDF on the forums, too much was thrown out.

In 2020: a 20MiB PDF is harmless to have on disk, just annoying to transfer over some people's internet connections.  Getting it down to the size of a floppy isn't really necessary unless you face arbitrary rules where you want to upload it.


Quote
One thing I may do to this and another manual I posted there is the removal of the shading of the other pages - a brightness setting on the scanner would have minimized/solved this.

N.B. 'brightness' on different devices means different things.  Sometimes it adjusts both the black and white points, others only the white point, in some it even adjusts gamma instead.  As a general rule for monitors and scanners: ignore what the controls are named, fiddle with them and work out what they mean yourself.

The 'Xsane' scanning utility (Linux) is getting old, but it still works and provides lots of level adjustments & preview options.  It was my choice the last time I did some mass scans, but that was a few years back.
« Last Edit: February 04, 2020, 09:35:06 pm by Whales »
 

Offline Whales

  • Super Contributor
  • ***
  • Posts: 2122
  • Country: au
    • Halestrom
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #46 on: February 04, 2020, 09:38:24 pm »
This is why I avoid threshhold processing for text.  The intermediate intensity pixels add clarity when viewed by our eyes, which - in conjunction with the human brain - have developed algorithms to see things much more clearly.

Yep.

In some situations we can cheat: really high res vector art or vector text can be saved in a thresholded format, then when it gets downscaled onto your screen by your image viewer it looks nice and smooth again.  Unfortunately that's not really possible in this situation.

Quote
Reducing the ghosting from the back of the pages would have been useful

N.B. I had a hard time getting this (levels/curves adjustment) to avoid damaging the greys around the text itself.  Text didn't look anywhere as nice, even with some gamma tweaking.

There are some smarter tools out there like unpaper.  Not perfect and the output has sections whited out (read:ugly), but pretty good at preserving the text you want.

Quote
- so would converting from a colour palette to greyscale.

This may not save as much data as you expect.  JPEG already heavily subsamples the colour out.  Not sure of exact %.
« Last Edit: February 04, 2020, 09:40:39 pm by Whales »
 

Offline olafwagnerTopic starter

  • Contributor
  • Posts: 17
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #47 on: February 08, 2020, 01:45:20 am »

I have been quiet for a couple of days waiting for my 3M resistors to arrive.

1. Inspecting the PCB
While waiting for them, I took a closer look at the PCB, and discovered a problem. Firstly I removed the 1M resistor, and when I removed it I saw the pad for the original resistor was missing (probably why it was hacked onto the bottom). Also, I saw that the connector that goes to the 'Focus' pot was loose, and had a broken trace. This immediately gave me a good theory on why the HV rail was first too low, and subsequently too high. My assumption was that focus is connected directly to the HV rail (through the aforementioned R697 in yellow on the attached schematic), and when this resistor was 1M it was 'draining' too much of the HV rail through the focus pot. I believe when I removed the PCB the first time, it may have caused the trace to get damaged (I was so gentle I assume it was already somewhat damaged, but me removing it broke it, and as a result there was no longer ANY current running from the HV rail through R697 and through the Focus pot. Hmmm....

2. 3M Resistors (Part 1)
Eventually the resistors arrived, and I saw they had '3M' written in pen, but fortunately I double-checked with my DMM and they turned out to be incorrectly labeled. They were actually just 3 Ohm resistors (that would have beeen fun had I installed them). I looked under a magnifier and saw the colors were Orange, Black, Black, Silver, Brown, which backed up the DMM.  (3 (Orange) 0 (Black) 0 (Black) * 0.01ohm (Silver) = 3 ohm (with a +-1% tolerance (brown)). I contacted the Ebay seller and to his credit he immediately sent me the proper resistors. Just more time to wait...

3. Checking the Caps 'In-circuit'
While I waited I thought I would turn my attention to all the relevant capacitors in both the HV and blanking circuits. However, I didn't want to remove each and every one to test them, so I thought I'd do a 'cursory' in-circuit ESR test to see whether anything obvious pops up. The problem is that my ESR meter only offically measures down to 1uF (I have a Signstek MESR-100 v2), and many of these caps are much lower, but regardless I soldiered on and marked those that could be measured (and measured good) in green, and those that came back with no resistance measurement at all in yellow) Unfortunately this didn't highlight any caps as being problematic (I guess that doesn't neccessarily mean that they aren't part of the problem, just not obviously so - or are in-circuit ESR measurements somewhat dependable?)

4. 3M Resistors (Part 2)
The correct 3M resistors arrived today, so I immediately installed one (as well as a small bodge wire to address the missing pad and broken trace). I also replaced both transistors Q621 and Q622 (might as well while I have the board out), and then reinstalled everything and powered up the scope.

At last, I can report a 'baby-step' of progress. The HV rail now reads -1,821v, which now falls WITHIN the spec range of -1,995v to -1,805v (only just, but I'll take it). :-) OK, so my theory above on the wrong resistor value and broken trace turns out to be correct.

Unfortunately, all my other issues remain, so there is still some work to do. :-(

Two of the issues which I still feel must be related are:
- Unblanking signal (which I can see at "P1") does not make it to "P2"
- The INTEN Pot (or internal trimmer) are unable to reduce the signal to a blank screen.


5. Next Steps
So I have 3 theories (options?):
Theory #1: Capacitors: There are only 3 caps in this area (C659, C650 and C649.) (While my schematic shows C671 it is not physically on my PCB nor is there a silkscreen marking for it) I could pull those 3 caps from the board and test them properly out of circuit. I don't feel very strong about this because all 3 tested fine in circuit.

Theory #2: Transistors: There are 2 more downstream transistors: Q632 (2SB648A which is a PNP Hitachi Low Frequency High Voltage Amplifier) or Q623 (2SD668A which is the complementary NPN Hitachi Low Frequency Low Voltage Amplifier) Does it make (theoretical) sense that if either of these transistors had a problem it would affect Q622? Doesn't seem logical seeing as the emitter of Q622 goes both to the base of Q623 as well as -12v (through R691), so I would expect Q622 to function regardless of whether its output signal is amplified or not.

Theory #3: Diodes: Another very weak theory, but I am unable to test these diodes in-circuit with my DMM, I just don't get a 'beep' in either direction on ANY of the diodes (my DMM just doesn't crank out enough volts I guess). Is there a safe way to test these diodes in circuit? Or should I pull them and plug them into my component tester? (there are about 4 or 5 in the vicinity).

Those are pretty much all the options I can think of. The only other components are resistors, and I have measured all of them in-circuit and they all read very very close to their expected values.
 

Offline olafwagnerTopic starter

  • Contributor
  • Posts: 17
  • Country: us
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #48 on: February 08, 2020, 04:49:50 am »

I was poking around the board some more taking measurements, and I got 'something' at "P2" by turning the intensity ALL the way up, but what I see doesn't make sense.

I left the scope running, had some dinner and when I came back I switched it off, and heard a lot of static 'pops'. I switched it on again, and while measuring some more, I heard more pops that sounded like they were coming from near or around the CRT. As I switched off the scope and was staring at the CRT I saw a few tiny sparks (accompanied by pops) on the power supply board (on the negative pin of one of the big caps) I immediately discharged all the caps again and removed the power supply board.

When I inspected the board, I saw yet another damaged trace at capacitor C821, which sits between the 160v and 180v rails. I measured the ESR on the cap (and it measured good), so at first I just wanted to fashion another bodge wire, but I thought I am so close, I am going to remove this cap and test it properly.

I heated up the soldering iron, and as soon as I touched the negative pin (where the broken trace is), I heard the classic 'simmering' sound, and I knew, this cap has leaked. I then removed the cap entirely, and it is nice and wet on its underside (see attached pics).

Also when I removed it, (of course), the pad immediately broke loose. The broken trace also means that the 160v rail will not work (because that pad sits inline), so while I didn't retest my 160v rail recently, I assume that it couldn't have worked (or at least not properly)

For giggles I hooked it up to my LCR and it still measured 763uF (should be a 1000uF cap) which is more than I was expecting, but certainly far out of 'spec'.

Now I need to order a replacement (I don't have anything close to this large in my stash). I am curious that this is only a 50v cap, when it sits between the 180v and 160v rails. (even the schematic calls for a 1000uF 50v cap)  Is this because the delta is just 20v? Or should I try to find a cap with a higher voltage rating?

 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4931
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Took the Plunge - GoldStar OS-7020 Oscilloscope
« Reply #49 on: February 08, 2020, 09:30:48 am »
The cap rating of 50V is because in normal operation it doesn't see more than 20V (though if one of the rails it's connected to fails to 0V, it's not gonna be happy!), and a 1000uF cap with a say, 200V rating would be big and expensive.

Since its bum is still wet, I'd guess it's a recent failure.

One problem with testing caps for leakage failure is that it often doesn't manifest until you get close to operating voltage, or at any rate, well above the voltage that most LCR meters put out.

EDIT: Also,   :wtf:  on those 3M resistors! Good thing you did check them before using them!


:-BROKE
« Last Edit: February 08, 2020, 09:39:21 am by ThickPhilM »
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf