Minolta DiMAGE Scan Elite II troubleshooting

[ttyR2]

Hello!

First time poster here.

I've been on a journey reviving some classic Apple Power Mac computers. My current project is a beige G3. I wanted to get a negative scanner to enable me to scan and edit photos on this machine.

This is going to be a bit of a story, but here it is.

I found a Minolta DIMAGE Scan Elite II on ebay Canada and bought it. It was listed as untested so I wasn't expecting much. It arrived with all of the accessories. I plugged it in and the power LED and controls didn't do anything. I opened up the case and found the ribbon cable that connects the front control panel disconnected. I reconnected that cable and then the buttons started working.

The Mac running OS 9.2.2 (again, this is just fun hobby stuff for me) and the Minolta DIMAGE scanning software (version 1.0.1, I believe) couldn't find the scanner at all via firewire. Next I installed the trial version of VueScan on a Windows 10 machine and connected the scanner via USB. The first time I launched it, it didn't recognize the scanner. I reseated all of the cables inside of the scanner and noticed that one of the proximity sensors, the one for the upper-most carriage, had the little circuit board tweaked. I'm surprised the internal sensor wires didn't break. I reflowed solder on that board and put it all back together. With the scanner powered on, I launched VueScan again and this time it detected the scanner by name and asked me to close the door on the front of the scanner. I did that and the scanner made a few more noises and then went silent. I ran the calibration operation and heard various motors in the scanner doing their thing and the calibration failed and locked up the software completely. I force-quit VueScan and power-cycled the scanner. Trying it all again, it then passed the calibration. Next it told me to insert the negative holder strip. When I did that, it grabbed it and pulled it in to position. I clicked Preview, and it made some noises, and started scanning the first negative. It stopped about half-way through, locked up the software, and the power LED on the front of the scanner was blinking rapidly in a 5-blink pattern. That was the only time I got the scanner to go that far. It then proceeded to stop doing anything other than making a high-pitch whirring noise (sounded like the small motor that's evidently inside the actual scanner head) for a couple seconds and then stopped. After reseating things again and getting slightly desperate, I tapped on the scanner head gently with a plastic screwdriver butt, and it successfully went through the calibration process but failed when trying to do a Preview.

I don't really have much to lose, and am thinking about opening up the scanner head but it has screws that have red paint on them which usually means something has been factory calibrated or adjusted. Do you guys think there's anything inside I might possibly be able to "unstuck" or get working? The mildly frustrating part is there are so many sensors/limit switches on this thing that I can't run it completely disassembled to observe what's happening, only partially disassembled.

I looked over all of the capacitors with a magnifying glass and none of them look like they have leaked or are corroded...doesn't mean they aren't bad though.

If I can get it working reliably on the PC, I'll be moving it back to the Mac for use with Photoshop 7.

P.S. I will add that I jogged all three of the exposes stepper motors "off station" with the power off, and when I powered the unit back up, they all reset and stopped at what looked like a precise position based on observation of the proximity sensors, so that part of the system appears to be ok. All of the guide rails also have a layer of silicone grease that is still in good condition (liquid, slippery).

 

[koraks]

I have a Scan Dual IV, so a rather different model. The only real problem I've had so far is a failing power supply. The scanner would still turn on, but it would show erratic behavior, like failed scans, garbage data etc. Since it's easy enough to try, I'd give it a go with a different PSU (perhaps a bench power supply just for testing) and see if that helps any.

What's not so great is that by the sound of it, you're not the first to root around inside the machine. That's generally bad news as you can never tell what they might have broken in attempts to fix whatever problem. It's also an indication that there have been problems already.

PS: the areas where I'd expect problems on these scanners are:
Motors and film transport; look for lack of lubrication and worn gears.
Light source; the CFL itself gets dim or dies and the HV power supply is also likely to fail.

 

[ttyR2]

I spent some more time trying to figure out which motor was making the whining noise during initialization a few seconds after launching VueScan. It looks like the scanner carriage/motor is not working as it should. The stepper motor was actually a bit warm, and when I gently try to rotate the plastic coupler that couples the motor to the screw, it feels stiffer than it should for such a small motor to turn it. Can you guys give me any idea how free the screw should be? It has what I think is a spring-loaded backlash compensator so I'm sure that adds some friction, but it sure feels stiff to me. I'll do some more looking around at the carriage...the one guide rod feels pretty slick from the silicone (or whatever it is) grease.

I should add that previously during troubleshooting this motor drive was working perfectly. At some point it broke and I just kept assuming that it was working fine. That was a mistake.

 

[ttyR2]

Well...waddayaknow. The motor-end of the plastic coupler is cracked. I tried to gently unscrew the setscrew and found it turned freely and the end of the coupler was cracked. Will see if a good superglue will hold long enough for testing purposes.

 

[ttyR2]

So...I can't find an identical plastic "accordion" style coupler and identical aluminum couplers are North of $65. Do you think it would be really detrimental to the operation of the drive if I used a solid aluminum coupler and spent some time very carefully aligning the stepper motor so it didn't put any sideload on the screw shaft?

The coupler is 20mm long, 12mm diameter, with 3mm holes in either end.

 

[koraks]

If it's just a static coupler/drive rod I don't see a problem with substituting aluminum. I hope this does indeed fix the issue!

 

[ttyR2]

I'm going to try embedding a small metal ring using heat into the end of that coupler. If I'm careful, I think it would end up stronger than the plastic was when new. The plastic is acetal or Nylon and I don't know of anything that really bonds that stuff well.

 

[koraks]

That sounds like a promising fix, especially if you can ensure that the metal collar inside won't be able to rotate in relation to the plastic part. Perhaps attach a fin on it somewhere that embeds in the plastic? The torque of the motor will of course be transferred through the collar onto the plastic part, which makes it important to ensure they bond well.

 

[ttyR2]

Well...when I say ring, I mean like a thin wire in a circle to just act as reinforcement. I'm planning on using the existing setscrews as the threaded holes seem fine. The wire ring is just to keep the plastic from flexing. We'll see if it works.

I'll post a picture when done.

....actually...I was going to post a picture of how the other end of the coupler was molded differently and has a plastic ring around the end for reinforcement. Then I noticed it too was cracked, but it looks like the molded ring has kept it together. Hrm...maybe I'll just replace the whole part.

 

[ttyR2]

It's fixed!

I would have liked to replace it with an identical plastic coupler but for now, this is my solution. Steel replacements that include the flexible portion start at $60 and go up from there. The red part is a clamp normally used for fuel line. I ground the tangs down so they wouldn't catch on anything, re-assembled the drive mechanism, screwed the two set screws in on the cracked part just until the plastic started to flex a touch, and then slid the clamp over them. It seems to be working great and the scanner is 100% functional!

One question...so I did pull the sensor out of the sensor housing underneath thinking there was a motor in there. There wasn't. I tried to align it the best I could and the scans I've done look fine, but if I want to be thorough, how would I do an alignment? I assume mostly what I'm worried about is if the sensor is skewed. Is there a negative I can buy with a test pattern pre-printed on it or something?

 

[koraks]

Great!
For the sensor alignment, you could try to make a negative (doesn't even have to involve a camera) that has a cross-hatched pattern with horizontal and vertical lines. If the sensor is askew, I'd expect this pattern to render lozene-shaped boxes.
You could inkjet print a 'negative', or even mechanically/manually fashion it by somehow marking a piece of transparent material.

 

[brbo]

Camera scanning people lately use inexpensive Vlads test targets. I would put something like that in a glass slide mount to ensure the negative stays flat during adjustment.