Somewhat OT, but ... (Slide scanners)

I have on of those scanners and it has developed a fault with a collar that connects the stepper motor to the mechanism that advances the sensor. The collar has split so the push-fit onto the motor shaft sometimes slips.

I fixed it by putting tape round the collar to hold it to the correct size and then wrapped a bit of tap round the motor shaft to make it slightly larger. Together, these make sure that when the motor turns a few steps, the advance mechanism moves the correct amount.

That's clear, but I'm uncertain if mine has the same problem, so here's the sequence I followed and where I think it fails:

1) I hit the ON switch. The green light flashes a bit and then stays on

2) Then, when Vuescan is launched I hear the scanned run a stepper motor fairly fast three times as Vuecan gets synced with it

3) When I put the slide holder in, I push it in about 20mm against some resistance until I feel it 'click', at which point the stepper pulls it in past the first slide, running a lot slower than it did during initialisation.

4) Then hitting scan does the business, producing a good quality image which is then saved as a JPG. This is the latest version of Vuescan and differs from the copy I last used back in 2011, when scanning was a three stage process: scan in a low resolution preview, use it to set framing etc, do the high resolution scan and only then save the image.

From what my sister said the scanner no longer grabs the slide holder when it is pushed in (step 3). I was on the phone during the last attempt and clearly heard the triple set of stepper runs during initialisation, but did not hear anything when the slide holder was pushed in.

I realise its probably a long time since you had a problem with your scanner, but how does this more detailed account compare with your experience?

--
Martin    | martin at 
Gregorie  | gregorie dot org

Sadly I think it's different. In my case the slide holder was being sucked into the machine OK, but then the very slow pixel-by-pixel advance during a scan was slipping. I wonder if two different motors are used for the coarse feed for loading the slide holder and the fine feed during scanning.

I think I'm correct in this: it was a year or so ago when I had the problem. I'll see if I can find the scanner and its PSU (it's always the PSU that goes missing with appliances!) and check my memory of the symptoms that I had.

I remember that my fix with tape round the motor spindle affected the aspect ratio slightly, so I had to punch a hole in a bit of film leader with a hole punch and measure the exact pixel dimensions on two axes, and apply a correction to all scans. If only Minolta were still in business I would get a new plastic collar to replace the one that's split.

Ah! Many apologies. I've just checked my scanner and it's a Scan Elite II, not a Scan Dual IV. My experiences and remedies may well not apply to your model.

It looks as if on my model there is only one motor (apart from the focus motor) so evidently this must both load the slide holder and feed it at low speed for scanning.

Here are my instructions, on the off-chance that they are relevant to your model.

1. Remove single screw on reverse (by USB and power sockets).
2. Remove four screws on base beside rubber feet, but not two on underside of front panel.
3. Unscrew the two screws on the metal plate in
   formatting link
4. Remove the two ribbon cables from their sockets
   formatting link
   (wiggle and pull out; maybe label them first to make sure they go back in the correct sockets)
5. formatting link
   is the collar that connects at the left hand end to the stepper motor and the right hand end to the feeder advance for loading the slide holder and for moving the slide past the sensor
6. If the motor is slipping on the collar, unscrew the motor
   formatting link
   and wiggle to remove its spindle from the collar (push fit, although I wonder if there are also Allen key grub screws on the collar)
7. Wrap a bit of tape round the motor spindle and then refit the motor - hopefully there is now a tight fit between the spindle and the collar. Looks as if I didn't try binding the outside of the collar to repair the split - I'd remembered that bit wrongly.
8. Refit everything by following Steps 6-1 in that order.

This may all be totally irrelevant to your scanner, but it may help you...

Good point: mine initially had an overheating problem which meant it would need rebooting toward the end of the third box of slides in a scanning session so I had it apart to see if I could fit a fan.

I don't remember how many steppers I saw in it.

FWIW I didn't fit the fan, mainly because it would have involved chopping a fan-sized hole in the case and, probably, adding a baffled outlet passage on the outside to stop light leaking in and degrading the scanned image's contrast ratio.

--
Martin    | martin at 
Gregorie  | gregorie dot org

Thanks for that.

Repairwise, I've an aeromodelling background, so if I'd had your problem I probably would have run a drill through the collar and shaft and put a small machine screw and Nylock nut through to lock them together. An 8 or

10BA screw would do the job.

--
Martin    | martin at 
Gregorie  | gregorie dot org

I did wonder about that, but the shaft is only about 2 mm diameter so it would be difficult to drill through it without cutting it in half. Grub screws to bite onto the shaft should be a good way of fastening collar to shaft in that case.

I did wonder about trying to file a flat onto the shaft for a screw in the collar to get a better purchase on.

I'm suspicious of any precision mechanical fittings that are plastic rather than metal. Plastic (the collar was probably nylon) is too inclined to crack under load or as it gets old.

Still, my admittedly bodged repairs was good enough to scan the photos of my parents' wedding in time for their 50th wedding anniversary :-) Getting good scans from 50-year-old slides, some of which had faded and others of which were underexposed, was another matter. Slide film was so f-ing slow in the early 60s that even in good daylight, camera shake was a problem with some of them. Probably Kodachrome 25 (not even 64), and certainly not Ektachrome 200 or 400 :-)

As a matter of interest, have you had any success scanning negatives as opposed to slides? I find I cannot get natural-looking colours. They all look rather flat and with garish colours, like the coloured "plates" illustrations that you got in 1930s books. And dark objects against a light sky (a light object against a dark background on the neg) tend to suffer from bloom - the light diffuses sideways and giving a halo around the object. Having said that, there's a *lot* more shadow and highlight detail that a scan of a negative can reveal, compared with a scan of a print from a high street shop.

Minolta's software gave better results than VueScan, but it won't install/run on Win 7, only older version of Windows. VueScan has lots of presets but none which properly match the edge-markings on the film - mostly fairly standard Kodacolor print film, so nothing unusual.

Scans of B&W neg film are fine, though I find I get better results scanning either as a positive and reversing or else scanning as colour neg and then reducing to B&W. Does VueScan have presets for Ilford FP4 and HP5? I got fantastic results with the Ilford B&W film that uses colour chemistry, despite rather dense negatives (overdeveloped?) - sharpness was probably limited more by the camera lens than the film grain. I got less noticeable grain from that film at 400 ASA than for FP4 at 100 ASA.

I didn't realise it was that thin - thought it would be more like 3-3.5mm!

I might still have drilled it, say at 0.7mm which is a push fit for

22swg, and put a piece of piano wire or copper wire through with the ends bent over to stop it falling out.

The problem would be getting the enough threads on the collar - unless it was at least 2mm thick.

Yes, I'm with you there.

Tell me about it! the stuff I scanned was 1977/78 vintage Ektachrome - don't remember the ISO rating, but some of the sliders had a fine black speckling, v. noticeable on sky. Took me a long time to get that cleaned up with the GIMP but I think the effort was well worth it.

Much of that slide collection, which covered an overland trip to India and back, was shot with a Leica IIIb through f3.5 Elmar 50mm and 35mm lenses and by waving a Weston Master IV meter at the scenery. The rest was shot on a 2nd hand Zenit with a cheap as chips East German f3.5 135 lens that worked surprisingly well until about 18 months later when its iris disintegrated. Thats when it became obvious that the lenses were mounted in what looked suspiciously like Plaster of Paris(!).

I've never tried, mainly because I've only ever used B&W roll film negs.

I went with Vuescan mainly because I'd converted entirely to Linux about

5 years earlier and at that stage WINE was still pretty much rubbish. I used Vuescan because it was about all I could find that could run on Linux and drive the Minolta.

Both WINE and Vuescan seem to have improved greatly since then.

I haven't looked - sorry and don't currently have a runnable copy to check with. I was a bit surprised to find that I still have the 2011 copy installed but it won't run because at least one support library has moved on. I'll get hold of the latest version when the Minolta comes back and, hopefully, gets fixed.

--
Martin    | martin at 
Gregorie  | gregorie dot org

[Snip]

Please take this to email, as only you two have any interest in this off topic conversation.

---druck

Even more OT. I never throw away Wallwart PSUs etc. unless they are faulty. This is why I have collected two boxes of PSUs in the last 35 years since moving into my own home. At the last count there were over

70 of the buggers! The adoption of USB mini A or micro B based chargers means very few have been added to the collection in the last few years.

Either that, or a more appropriate group, like sci.electronics.repair

I may only be lurking but find the discussion interesting

--
Good judgement comes from experience.  Experience comes from bad  
judgement. 
		-- Jim Horning

Thanks for the hint. Subscribed.

--
Martin    | martin at 
Gregorie  | gregorie dot org