Backlight stripes in MINOLTA DiMAGE Scan 5400 II or what?

<p>Hello everybody! Some time ago i got a MINOLTA DiMAGE Scan 5400 II film scanner and after some days of tests i begin to notice a strange stripes on photos with constant background fill. Must say that usually those stripes are almost invisible but sometimes they are really annoying. So i tried to disassemble the scanner (i got some skills and clear room) and got luck to look inside the light element system. It consists of 3 LEDs (half-sphere shaped) + one large lens (seems plastic, because it was very electrostatic), few mirrors, IR elements and few very very thin PLASTIC diffusion pads (so thin that i cant call them diffusion glasses). So my guess that this problem may be caused by poor diffusion system (the pairs of stripes are divide scanned image to 3 equal parts). Unfortunately i can't disassemble scanner further as so as main part with CCD element and mirrors are hidden very deep inside and protected by the gate which opens only when the scanner begin to scan.<br>

So now i got the idea to experiment and put a DIY diffusion glass inside scanner light system.<br>

Also a'm very interested if anyone got same issue before. And maybe there is a ready solution for this problem?<br>

See examples of scanned grey frame below (first is original, second with boosted contrast):<br>

<img src="http://f.cl.ly/items/292s0R3e292S3S3p0h2R/light-stripes-orig.jpg" alt="" /><br>

<img src="http://f.cl.ly/items/2c143z2e3k2i0G0N0r2w/light-stripes-curves.jpg" alt="" /></p>

 

<p>And here is another example in addition to stripes there is a color tilt in lower third of the image:<br>

<img src="http://f.cl.ly/items/1h0B3z3z3F1r1E05392R/sky-non-uniform-color-tilt.jpg" alt="" /></p>

<p>Streaks, like the ones you are experiencing, are from residue on the optical elements of the scanner. The contaminates are from components and lubrication gassing out and contaminates in the air of the environment in which they are used and stored.<br>

Cleaning everything in the optical path from the light source to the sensor is required to eliminate them. How often such a cleaning is required varies from one brand to another. A partial cleaning may help for a while.</p>

<p>Here is one more update: i find a mirror on the other side of the scanner, mounted at 45 degree. I remove it, it was almost clear with only one small micro dust particle, i clean it and put back. Nothing has changed. So i believe that lens+sensor which are and probably combined to one single part and hidden way further can not be affected by dust. And those symmetry... There is something with a light source for sure. Maybe diffusion foil was burned a little a little after years of lifetime...<br /> The another strange thing that those stripes appears differently on different sources. For example when i try to scan IT8 target slide there were no stripes on its grey background (even after extreme contrast and saturation boost). Empty frame scan (even when underexposed and grey) don't produces any stripes too. But then i put a darker grey negative B/W film or peace of semi-transparent white plastic the stripes became very visible.</p>

<p>Here is another good example of non uniform backlight color. Negative film, scan taken by previous owner of this scanner (original and saturation boosted images):<img src="http://f.cl.ly/items/2P2U17420G0u1o371V1o/tint-orig.jpg" alt="" /><img src="http://f.cl.ly/items/3H161y3W3p352s0E2J2Z/tint-saturation.jpg" alt="" /></p>

<p> </p>

<p>The sensor is stationary with the film being moved during the scan. It is possible the sensor and lens are a sealed unit and requires replacement when a problem occurs with it. <br>

Other possibilities are the AC power adapter is putting out an unsteady voltage in the scan cycle or the LED light source may be fluctuating.<br>

I use a microfiber lens cleaning cloth and my breath to clean optical devices because it leaves no residue.</p>

 

<p>Hmm... AC adapter, interesting theory. I use 220 to 110 v convertor to power up the scanners AC adapter, and previous owner used adapter too (but he had different adapter).<br>

I done a little more tests today.<br>

First of all here is a <strong>PHOTO SET OF DISASSEMBLED LIGHT BLOCK: <a href="https://www.dropbox.com/sh/dgxlg0uy3zqolzg/SkOYP9HpbO">https://www.dropbox.com/sh/dgxlg0uy3zqolzg/SkOYP9HpbO</a></strong><br>

Second, i done two scans of matte glass. and with removed optical elements. First scan is in usual condition, and when i remove a large lens there were almost no any difference. And the second scan done with removed triple lens located in front of three LEDs, as you can see the stripes became more visible. I also tried to put a DIY plastic diffuser outside the block but it don't change things, i got only longer scan time. And at last i try to put a sheet of tracing paper inside of light block but the scanner can't initiate itself (maybe because the tracing paper was too dense)<br>

<img src="http://f.cl.ly/items/2s2w2c41370k0o2q0P0M/matte-glass.jpg" alt="" /><br>

<img src="http://f.cl.ly/items/2A2X2l2a3x4422071p3l/matte-glass-no-led-lens.jpg" alt="" /></p>

<p>UPDATE.<br>

After some search on Flickr I find that this problem with backlight nonuniformity presents more or less on all 5400II scanners and depends of many different factors and conditions it could be too visible or almost invisible. So in conclusion the problem required some kind global fix of light source itself or post processing workflow based on scan of uniform matte diffusion source. </p>

<p>Thanks to suggestions from Eric, the Scanhancer developer, the problem was partially solved with a small sheet <a href="http://www.leefilters.com/lighting/diffusion-list.html">Lee 400 Lux Filter</a> mounted about 2 mm in front of the LEDs. The shadow stripes are gone but unfortunately magenta (or green in positive mode) color tint from one of the LEDs still there and too visible on the scans of clear blue sky for example.<br>

So the next crazy idea is to modify light source itself with something like this very modern and very uniform <a href="http://dx.com/p/7w-6500k-450lm-white-light-cob-led-rectangle-strip-silver-dc-9-11v-222933">450lm 7W COB LED Rectangle Strip</a> with separate external power supply or battery pack or this <a href="http://www.ebay.com/itm/4-8W-COB-LED-SMD-Strip-Light-Bulb-Lamp-Bead-Chip-6000-6500K-White-/380636904457?pt=US_Light_Bulbs&hash=item589fba3c09">320lm 4.8W COB LED SMD Strip</a><br>

<img src="http://img.dxcdn.com/productimages/sku_222933_1.jpg" alt="" width="600" height="600" /></p>

<p>Hi Dmitry, <br /> I seem to be having the same problem (also Minolta 5400 II). Here is a sample of an empty slide and an empty film holder, both bare output and with contrast enhanced:<br>

http://www.machalik.ic.cz/pics/temp/stripes01.jpg<br>

http://www.machalik.ic.cz/pics/temp/stripes02.jpg<br>

http://www.machalik.ic.cz/pics/temp/stripes03.jpg<br>

http://www.machalik.ic.cz/pics/temp/stripes04.jpg</p>

<p>What's weird is that the scanner was doing alright on 15th November 2013 when I scanned these photos (although the foliage is making the comparison difficult):<br>

http://www.machalik.ic.cz/pics/temp/aiku01.jpg<br>

http://www.machalik.ic.cz/pics/temp/aiku02.jpg<br>

http://www.machalik.ic.cz/pics/temp/aiku03.jpg<br>

http://www.machalik.ic.cz/pics/temp/aiku04.jpg</p>

<p>And for sure the scanner was fine on 4th September 2013 when I scanned this one (which is a good example):<br>

http://www.m5400ii.ic.cz/pics/temp/contest/009.jpg</p>

<p>I first noticed the issue when scanning some more images on 3rd December 2013. Technically, the scanner lamp deteriorated in 3 weeks between 15th Nov and 3rd Dec 2013. Today, it is producing pictures like these:<br>

http://www.machalik.ic.cz/pics/temp/loz01.jpg<br>

http://www.machalik.ic.cz/pics/temp/loz02.jpg<br>

http://www.machalik.ic.cz/pics/temp/loz03.jpg</p>

<p>Did you manage to find a solution for your problem other than replacing the lamp with something completely different? I am not skilled and will not be able to insert such device into the scanner, it is beyond my capabilities. In case you were still visiting this board (I know I am a few months late), I will appreciate any advice.<br>

You can also reach me at jiri.machala@gmail.com which is my personal e-mail.<br>

I want to avoid buying a new scanner and potentially facing the same issue. Mine worked well for about 8 years so I would much rather repair it and keep it.</p>

<p>Thanks a mill for any answer, I know this is a long shot.<br>

Jiri<br>

Czech Republic</p>

<p>hi Jiri and everyone, as so as there are very interesting events in the place where i live (Ukraine), currently i stop my experiments for a while. but let me share some of previous experience. <br>

1. as i told before Lee 400 Lux Filter mounted right after the leds panel is the simplest way to partialy disable the backlight nonuniformity. its rather simple to do, the scanner is very easy to disassemble and it has very nice and simple design inside. take a look here <a href="https://www.dropbox.com/sh/dgxlg0uy3zqolzg/SkOYP9HpbO">https://www.dropbox.com/sh/dgxlg0uy3zqolzg/SkOYP9HpbO</a> but you still will have a color shifts and you got a little longer scan time.<br>

2. forget about those cheap led strips in post Sep 11, 2013; 05:35 p.m. i do a tons of experiments and IT WILL NOT WORK, and if you really want to know i can explain why later, but in short the optical light scheme in minolta will work only with 3 dot lights, but not with uniform light strip.<br>

3. so my current idea is to use these super new hi-cri leds with violet backlight and 3 component luminophor <a href="http://store.yujiintl.com/collections/frontpage/products/5630-high-cri-95-led">http://store.yujiintl.com/collections/frontpage/products/5630-high-cri-95-led</a> <a href="http://www.yujiintl.com/high-cri-led-lighting">http://www.yujiintl.com/high-cri-led-lighting</a> . i measure the real wattage and the woltage of the original minolta leds and it appear that they almost identical to those Yuji leds, so it should work well from original power supply.<br>

minolta:<br />4.3V (working)<br />4,9V (without loading)<br />0,07А<br />0,301W (working)<br />0,343W (loading)<br>

<br />Yuji 5630<br />3-3,4V<br />0,12 A<br />0,36-0,4W<br>

so now i got these leds here, and my next step is to build a board with 3 mounted hi-cri leds+ diffusion lee filter (the assembled board can be ordered from a any local guy who have advanced skills in electronics). then i put it instead of original one and connect directly to scanner power pin. <br>

its all very experimental and untested and i can't guarantee anything. do so at your own risk.</p>

<p>Hi Dmitry!<br /> I wouldn't have hoped that you'd come back to this thread after a half a year. Thanks so much for getting back to me!</p>

<p>Now I have done some investigation myself (without disassembling the unit though) and found the following: my scanner had always had these stripes, but in the past they were prominent only in very, very dark slides, they have been showing as varying color (chroma) in black slides. See a scan from December 2012 of a non-exposed Kodak E100G slide - a totally black slide. The reason why the photo seems gray is that VueScan tries to compensate for the blackness and "pulls" tonality from the darkness, trying to make the picture look 18 % gray. In fact the slide was completely black (well, in fact the base is slightly reddish so it came out with a sort of a pink/magenta cast).</p>

<p>Raw scan: <br /> <img src="http://www.machalik.ic.cz/pics/temp/stripes05.jpg" alt="" /><br /> <br />Contrast enhanced in Photoshop: <br /> <img src="http://www.machalik.ic.cz/pics/temp/stripes06.jpg" alt="" /><br /> <br />Compare with an enhanced scan of an overexposed film from last week: <br /> <img src="http://www.machalik.ic.cz/pics/temp/stripes02.jpg" alt="" /><br /> You can see the stripes were always there, only in the past they were very, very dim and only affecting black areas in their color (which I didn't even notice on my pictures until, after 6 years of scanning my own stuff, I did a super-hires scans of quality macro photography for a friend of mine who had a lot of dark pictures with smooth background). Back in 2012 I thought this was just the limiting noise level of the CCD sensor and didn't really care - you would always be able to get rid of these in Photoshop very easily.</p>

<p>Now the condition has deteriorated and the stripes are visible all the time on all tonalities, dark or light. I developed a Photoshop action that is able to eliminate them (I apply an inverse layer of the empty slide which cancels the stripes in pretty much the same way digital cameras remove dead pixels) but it is by no means a real solution. See the result in an animated GIF: <br /><img src="http://www.machalik.ic.cz/pics/temp/stripes_out.gif" alt="" /></p>

<p>When I saw your images from the scanner disassembly, I got a theory: see that glass piece on the left side of this picture of yours? <img src="https://dl.dropboxusercontent.com/sh/dgxlg0uy3zqolzg/zsl0lBsMYA/IMG_4382.JPG?token_hash=AAHdiZ9iHA9ezhwIxiE98ftkokTwhxg5Y2YRZJULaL9Hbg" alt="" width="1000" /><br /> <br /> Not its two thin parts are exactly in 1/3 and 2/3 of the shape - the same place where the stripes appear for both you and me. I juxtapositioned it on this image so that you see what I mean: <br /><img src="http://www.machalik.ic.cz/pics/temp/stripes07.jpg" alt="" /><br /> In addition, it can be seen that even the top and bottom parts of the scanned image go dim - where there are two more "lows" of that glass piece. Now I might be completely wrong with this but the peculiar shape of the glass and the fact that the dark stripes appear exactly where the glass is thinner suggest that these two might be related. I am not saying it is just a mere dust buildup or whatever in the narrow crevices of the glass that causes this because that would look different, but maybe the glass (is it even glass? Or polyglass? What is it?) is deteriorating, has bent slightly or developed some tint and that is most prominent in its thinnest points.</p>

<p>The reason that leads me to believe this is that Minolta claims that the DiMage 5400 II (not DiMage 5400) uses a single white LED light source and no diffuser, not a cold cathode lamp and a diffuser as its 5400 predecesor (more on this in this forum: http://www.pcreview.co.uk/forums/dimage-scan-elite-5400-vs-dimage-scan-elite-5400-ii-t1939402.html - search for "light source". I can't wrap my head around how a single LED could develop stripes on its own without any further mechanical/optical element in the light path. One LED = one point of light, no striping or banding is possible with one light. Not caused by the light itself anyway. So it should be something further down the stream - and the glass piece looks very suspicions due to its shape.</p>

<p>It is funny though that your stripes are slightly different. While I have two dark stripes in both thirds of the image and both the top and the bottom of the image are dark, too, which is sort of a copy of the glass shape, you have in fact four stripes, also located around the first and second thirds of the image, but your stripes are symmetrically positioned _above and below_ the thirds, they are not positioned _on them_ such as mine. That I don't understand. Look at the overlay of my stripes and your stripes: <br /><img src="http://www.machalik.ic.cz/pics/temp/stripes08.gif" alt="" /><br /> But then I can see my dark stripes are surrounded by some lightness which, if inverted, would resemble your image. I scanned an overexposed Kodak slide = the image is positive, you did the same, right? You didn't scan a negative.</p>

<p>Let me know what your findings are about the new light source your are building but I strongly suspect the whole problem has something to do with that condenser lens or whatever it is.<br /> Thanks a mill for posting again and fingers crossed all will end up well for Ukraine! We are with you in our thoughts.<br /> Jiri</p>

<p>on my scan there is a negative film, not slide. thats why it looks like different to your example.<br>

all the scanner optical elements are made of plastic. regarding other things - i still don't know what is the problem. few days ago i mount a new led board. its very bright. there is no color shift in 3 areas but the dark stripes are present.<br>

p.s. it appears that there is no easy way to connect the leds. in current example the board takes power directly from scanner dc power adaptor and became controlled with wire used by original leds. not perfect but it works. i'll keep testing it.</p>

<p>Hi Dmitry, <br>

I have done some testing of my own and found the following: the stripes _do_ originate in the triple lens. They appear as soon as it is inserted in the path. <br>

Let's have a look at the light path. It starts with 3 white LEDs > then there is a 45° sloped UV filter (I think it is one, it's that little piece of glass that makes purple reflections) > the triple lens > two very thin diffusor plates > then a single lens > a 45° sloped mirror > one more diffuser before the light exits the lamp assembly > the sensor. <br>

When I inserted a piece of white paper between the triple lens and the thin diffusers and looked at it when the lamp was on, I could see two very distinct stripes already on that paper. They were where the thinnest parts of the lens are and slightly brownish/yellowish in colour. The rest of the system tries to dissolve them but is not 100 % successful because when I look at the very last diffusor (after the mirror), I can still see faint brownish stripes there. <br>

I took the triple lens out and washed it in mildly soapy water so as to remove any possible dust or fat or any sort of dirt on the lens. I also carefully cleaned the thin areas with a piece of cotton wool. After re-inserting the lens back to the assembly, the stripes were still there, meaning they are a direct product of the lens, even if it's clean.<br>

I have two theories:<br>

1) Either they always must have been there and are a natural effect of the triple lens, it just physically produces them. It is partially the task of the rest of the optical system to dissolve them but the system cannot be 100 % effective by principle, hence, ultimately, the scanner used to remove them digitally by the Calibration process in the software by simply scanning an empty field, inverting it and applying over all images scanned in that session, keeping the empty field in memory until shut down. Something very similar is described on VueScan page: <a href="http://www.hamrick.com/vuescan/html/vuesc26.htm#scannercalibrate" target="_blank">http://www.hamrick.com/vuescan/html/vuesc26.htm#scannercalibrate</a> <br>

Also, I used the same process in Photoshop when removing them from this image: <br>

<img src="http://www.machalik.ic.cz/pics/temp/stripes_out.gif" alt="" /><br>

I have no idea why the Calibration process stopped working (my VueScan hangs now when I try to run the Calibrate command).</p>

<p>2) Or they haven't been there before but the plastic triple lens has bent slightly or has been otherwise deformed, even in a microscopic way, at some point and started producing them. What used to be a perfect light dissolver, perfectly spreading the light output of the three LEDs, is now a slightly imperfect dissolver, suddenly producing stripes. The deformation could have been caused by heat or otherwise. If theory no. 1 is correct, this should still be correctible by the Calibration process but maybe the unevenness of the light field it is too strong now and the calibration just can't handle it any more.</p>

<p>In any way, I am fairly sure you can't get rid of them without getting rid of the triple lens. It seems that your tests confirm this as you are seeing them even after replacing the light source.</p>

<p>By the way, a gentleman called Dirk from a company RTC Solutions GmbH contacted me last week, they are a service company for these Minoltas, giving me some useful advices about how to determine if it's the lamp or the sensor that is faulty. He advised that I unscrew the lamp assembly from its position so that it doesn't shine at the sensor any more (but don't disconnect it) and use some other lamp to illuminate the sensor and make a scan. If there was anything wrong with the sensor, like stripes, they would get recorded even if a completely different light source was used to illuminate it. I have done so but unfortunately messed my scanner even more (now it's not booting for no reason known to me). Anyway, still I was able to determine the stripes _are_ produced by the lamp, not by sensor, and more precisely by the triple lens itself. Maybe he can join us in this discussion, I have sent him the address, he was super helpful. </p>

<p>Now fingers crossed I can get my scanner boot and scan again - even with stripes. <br>

Thanks, <br>

Jiri</p>

<p> </p>

<p>unfortunately your example witр empty film scan and masking the stripes in photoshop useful only for slide film. to use this method with the negative you need to scan a exactly empty film with orange mask, which is very different from one to another film manufacturer. <br>

i can confirm than cleaning the triple lens has no effect.<br>

now i 99% sure that the color slightly shift of thirds of the frame produced by leds which are slightly different degraded after many years. (as i told before on my new led setup i don't have this problem)<br>

in the first model 5400 there was a little different light scheme. the 3 ir leds and triple lens were used only for IR cleaning, and the fluorescent lamp tube were mounted directly in front of the film without any additional lenses. <br>

regarding VueScan - i don't suggest to use it at all for film scanning, because there is a lot of underwater bugs and quality problems with this software.<br>

btw here is the article about scan workflow based on minolta scanner (use google translater) there are also a lot of scan-related useful links there https://docs.google.com/document/d/1ojF67rhWbAX4l0IF9Epgg2Clgqg1kYamL9ch88NHghs/edit?pli=1</p>

<p>Some bump for the archaic topic here :) <br>

1. When i change the LEDs the colorful stripes are finally gone, so this part of the problem was solved. <br>

2. Yuji LEDs are great, but unfortunately they can't be just soldered instead of old original LEDs. You need to made for them a new board with separate power circuit. <br>

3. Yuji LEDs are much brighter, so you need to put more additional filters to dim them to normal level required by scanner CCD sensor.<br>

4. Sill some problems with light uniformity (not perfect but usable). As so as new light source is very bright, those two dark (light) horizontal shadow stripes became more visible now. Also the correct placement of the LEDs on the board is very important. Yuji LEDs are thinner so the distance from the LEDs board to the lens block must be adjusted too.</p>

<p>Dear Dmitry,<br>

I just bought an used Minolta Dimage Scan Elite 5400 II, that seems to be in "like new" condition, but I got the same banding problem you got.<br>

Did you find a better solution to solve that?<br>

Could be possible, paying for your service, to send you my Led board, and receive the fixed one.<br>

Thanks<br>

Maurizio</p><div></div>

<p>Update:<br /> I just installed the original Minolta Dimage Scan, and the scanner is working perfectly, without any banding/backlight stripes. I think that Vuescan is not capable of calibrate the scanner, so it will not recover backlight uniformity.<br>

Not if I scan an empty slide, the image will come out just white.</p>

<p>Maurizio, its really interesting. I got exact same black stripes results with Vuescan and with DimageScan software. Vuescan only cause additional pixel level aliasing stripes so thats why i don't recommend to use it.<br>

Did you make additional tests during last post? Note that those black stripes are not permanent, they are dependent of negative/slide density and colors. Also they are 100% invisible on empty scans (scans without any film negative) so its better to test it with clear matte glass or list of semy-transparent diffusion filter. Also from my tests they are less visible on film than on clear diffusion glass.<br>

Regarding LED service the problem is Yuji now sell its leds only 100 pcs min and its $95.00 http://store.yujiintl.com/collections/vtc-series/products/vtc-series-smd-5730<br>

Also seems they got new lower brightness leds http://store.yujiintl.com/collections/vtc-series/products/vtc-series-high-cri-led-smd-2835l-100-pcs those probably don't need so many additional diffusion filters to suppress excessive brightness. but they are also also 100 pcs min.<br>

also i still don't get perfect light area with my modification (its ok on film but the stripes are still very slightly visible on empty clear diffusion glass) its because the led board needs to be placed very precise (distance from led to lens) and its boring to disassemble my scanner dozens of times:) so i scan as is and it all looks ok. The only additional idea i got is to add cool 80a filter to light source to get something about 10000k color temp for better handling of negatives orange mask.<br>

p.s. its strange, but i don't get autonotifications to my email from this forum, so sorry for so late answers.</p>

<p>Dear Dmitry,<br>

If I scan a blank slide with Dimage, I got a perfectly white image, so I'm sure the software is doing a white frame calibration to balance all the imperfections.</p>

<p>do you mean it is unexposed transparent blank slide or photo of white wall? if its unexposed its the same as scan empty frame. do you try to scan something like matte white glass or film with clear sky and add saturation to the scan?</p>

<p>It was a blank slide, like the sample with Vuescan I posted before.<br /> So a blank slide scan with Vuescan has yellowish banding, and the same scan is pure white with Dimage.</p>

I know it's an old thread but it does not become outdated. A few years later the same problem is plaguing the 5400II users. I have a large number of different scanners, including several 5400IIs, and all the units I can put my hands on, do exhibit the problem this thread is about. The pattern and intensity varies slightly from unit to unit but all units have it. I spent countless hours trying to find the element that needs to be cleaned in order to get rid of the stripes. Then I tried the original software rather than Vuescan. And guess what? Yes, it seems like just another nuisance of Mr. Hamrick's overrated software piece. Don't get me wrong - I appreciate him supporting tons of scanners and doing mostly good job on them. But it's not the first time when I lost indecent amount of time troubleshooting a non-existent hardware/setup issue only to realise that it's Vuescan's inability of this or that. As for dmitry_shijan (OP) case - it might be that your device actually developed some real problem that can't be compensated by the calibration procedure, this scanner model requires before every session. For the rest of us - switching to the original software may be the only move that's needed. For what it's worth, it looks like sometimes Vuescan's calibration works better and the dreaded stripes are less pronounced. Another session a few days later may produce completely unusable results.

Scanning with Vuescan often requires multi sampling to remove software bugs. Setting Multi Sampling to "2" or "3" in "Professional" Vuescan helps to eliminate all bugs. Perhaps not with every scanner but multi sampling turned my ancient Microtek i900 from a dud to a usable scanner. Vuescan is all I have for it, so I worked at it for months, cleaning, scanning different resolutions and 1600dpi in the Vuescan list was the highest that was acceptable with single sample, even then it was buggy. Eventually I tried everything in Vuescan and when I came to "Multi Sampling" things improved greatly, soft banding was gone and now I can scan at the max 3200dpi in the list, but the scans take longer to do. It was my first scanner, that's why I persisted. I've bought more scanners since but still have the i900 as a back-up.

Only tangentially relevant to the theme of the post, but, I have a pdf of the service manual for this scanner,

and absolutely no idea how to post it here ?????