Howtek D4000 vertical streaks

[georgl]

I recently bought a used Howtek D4000 drum scanner. I scanned two slides which looked fine. On the following scans I started to notice "vertical streaks" over the whole image. They are aligned with the drum rotation, so they are not related to the lead screw!?

I added a sample with some sharpening to enhance the effect. These streaks are visible all over the image.

It took me one month to get my hands on the lubricant for the lead screw, the scanning material and fitting belts - Aztek send a holiday notification mail, I repeated my request after the holidays, but still no answer...

- I cleaned the scanner
- I used the clean and nearly-new drum
- put the slides on different positions of the drum
- replaced the belt (it was lengthened)
- lubricated the lead screw

But I still have these streaks! I'm going crazy! What else can I do?

I don't know what to do! Help! Thanks.... :o

[georgl]

The forum reduced the image size... So just the crop @ 200%

[Edwin Beckenbach]

Did you replace the drum pads?

Does the bearing sound smooth as it spins?

[uniB]

My D4000 has developed exactly the same problem, I've been meaning to get in touch with PS Electronics who I bought mine from to find out what's going on - just been waiting to get some money together to get it fixed. I'm pretty sure these guys will know what the problem is http://www.pselectronics.co.uk/

[georgl]

@uniB
Thanks! Could you keep me updated?

The pads are fine. It's even stranger: the very first scans without streaking were made with the worse drum already missing two pads - so this effect doesn't seem to be very sensitive about drum alignment! Some scans further I noticed the streaks and switched to the better drum.
The bearings sound fine, but I have no idea how it should sound!?

[IanMazursky]

There are a few things to look at.
1. Lead Screw lubrication. Since you just received it, i would clean and re lubricate it a few more times.
You would be surprise what they pickup over the years. Check the bearing all the way at the end of the right side, most of the excess lube ends up there.
Tons of stuff ie. dust and pieces of tape end up there. It may also be a greenish color which seems to be fine.
2. Room temperature. It should be between 65-75F. Colder then that, the scanner starts behaving weird!
My 7500 started acting up and producing line like that, the ambient temperature has dropped to 50F in my scanner room.
Both of my scanner (Howtek 7500 and Aztek Premier) don’t like it at all. I placed a heater at the air intake and after an hour she’s a happy girl again!
No more lines or errors.
3. Wear pads and location of film. Worn pads or the location of film can unbalance the drum.
This can cause areas of softness and a sound like the drum is trying to fly away. Try to place film in the middle of the drum and replace the pads when they don’t have a rounded top to them.
4. Drum belts. Those probably wont have much effect on the scan as they are connected to the rotation encoder and that compensates to some degree of wear.
But they should have very little slop and should not be marred, cracked or chipped in any way. If they are then they can start to disintegrate and its no fun to clean up.
5. Linear Encoder. This is something that is very dangerous to check (not to you but the scanner). This is the encoder that senses the position of the carriage on the lead screw.
Dust can clog the aluminum encoder disk and mess with the positioning of carriage. This can cause any number of issues.
You will have to disassemble the scanner case to get to it and its probably all the way on the left side of the lead screw.
A very light blast of air should dislodge the dust but be very careful, if you bend or damage the disk at all it will be ruined and its expensive to replace.
This is a last resort of things to check.
6. Scsi cables. They probably have nothing to do with the problem but make sure they are good cables and a good active terminator.
Bad cable can cause a whole host of issues, even the scsi card can contribute to problems.

Hope this helps and good luck!

[IanMazursky]

Forgot to add, try scanning the same image at different resolutions and apertures.
See if the lines show up at 4000spi and at lower resolutions.

[onnect17]

George,
I have a D4000 and I had the same problem at 4000 dpi. The only way I can get clean images is with a very soft rubber belt. Try the old belt.
Armando

[georgl]

First thanks for the suggestions, but so far nothing worked.

I compared the old (lengthened by >1cm) belt and compared it to the new [belt.jpg] looks identical to me.

I did a test with various resolutions (3000/4000/8000ppi) - while 3000ppi hardly shows any streaks, it gets worse at 4000ppi and horrible at 8000ppi - which is strange, because 8000ppi is just interpolated (same scan time) but sharper ? [streak2.jpg]

- streaks are always vertical -> it has something to do with the drum rotation, not with the lead screw ?

- it doesn't change with the belt

- it doesn't relate to the drum (pads, calibration strip)

- I warmed it up over several hours, turned up the room temperature (>20°C) - no change whatsoever

- it isn't affected by the film location on the drum - as far as I can tell

Can I somehow clean the encoder?

[uniB]

I'm waiting on PS Electronics to phone me back as Pete wasn't around today - should get back to me in the morning so hopefully I'll be able to shed some light on the problem tomorrow. I suspect it's got something to do with the Linear Encoder (I'm afraid!).

[IanMazursky]

The rotational encoder is located on the back of the drum spindle and it is sealed.
Do not try to open it or you will probably kill it. They are almost never the problem and last a long time.

The linear encoder on the other hand is open and as i mentioned in my last post, you can clean it.
But be very careful, a strong blast of air can damage the metal encoder wheel.
Its most likely the carriage motor and or the linear encoder that is attached to it.

Aztek may have the replacement parts. I would give them a call, they are amazing people to work with.
Evan might have more insight on the problem, he is Azteks service and support guru.

[onnect17]

The image shows the problem I had. It's like a shifting of the line.

The movement of the carriage in the D4000 is different than other howtek scanners.
Quoting Phil from Aztek,
...The D4000 and 4500 are more likely to vibrate because they
capture one raster line move the lead screw and then capture the next.
The Premier and HR8000 to synchronous capture thus moving and capturing
in on motion...
Unfortunately the drum drive pulley receives too much pressure from the drum in one of the sides, getting the pulley out of alignment and affecting the traction of the belt (it changes from the center to the sides, changing the transmission rate). The vibration associated with moving the drum to the next raster line does not help. The only way I'm able to reduce the problem is with a softer rubber belt.

[eric black]

I had a similar problem with my HR8000 which I sent to Evan at Aztek for diagnosis and fixing. Evan spent quite a bit of time troubleshooting and I dont recall all of the details but the verdict was new Mother Board needed- an expensive repair at over $5K but worth it for me as I now have essentially a new scanner and hope to get many years more of use out of it. Good luck with your troubleshooting and diagnosis- I hope your problem isnt as severe as mine.

[IanMazursky]

I could be wrong but i think the 7500, 4500 and 4000 use the same scan then move system.
At least thats how it sounds when my carriage on my 7500 moves.
My Premier sounds much different, its more fluid and doesn’t have move/stop sounds when its scanning like my 7500 does.

But it sounds like something is wrong with your scanner, i would try to let the scanner warm up for 4 or 5 hours and see if anything changes.
There is a lot of metal in these scanners and it takes a long time for them to warm up if its really cold.
You can also try reseating the main boards (turn the scanner off first), cleaning out the scanner of dust bunnies and reloading the firmware.
You never know what might help. Evan at Aztek might be able to tell you more.

[onnect17]

Ian,
Glad you mentioned the firmware. This one of the first things George should check.
I think the latest is the 535. Because my investment on the scanner was low I was free to almost totally dismantle it, and got familiar with it.

Check the 7500. Perhaps the drum has two dents or orifice. The 4000 has only one dent and the drum puts too much pressure in one side of the drum drive pulley. Also, every time the screw moves to the next raster line the momentum of the drum affects the match between pin and dent (I think they fixed the design later using orifices instead of dents and two pins instead of one).

[IanMazursky]

The firmware is always a good thing to re flash every now and then.
I have no idea why or how it gets corrupted but it does and will produce the strangest errors that will drive you crazy.

My 7500 drums have an through hole for the center hub to go into and another one off to the side for the pin.
I think this is a much more stable design then the 4000.
Also the clamping pressure of the 7500 is high, it could take your finger off!
I think my Premier is also designed roughly the same way.

[onnect17]

Ian, I sent you a PM. Thanks.

[georgl]

I've tried to install the new firmware but Trident says the scanner doesn't accept it and the scanner gives the error-code F717 ?
Since Trident doesnt start the scanner, either, I guess I have install WIN98? Does Win2000 also work?

I've tried some very flexible rubber bands - it seemed to help in the very first scan when I compared it to the scan made the day before with the new belt. But it introduced new artifacts, then the rubber band felt apart and spread all around the scanner! After close inspection of the scans, they showed the same streaks... It drives me mad!

[aluncrockford]

I have had exactly that problem,it was caused by a lose connection to the bulb causing the slightest of flicker, though having a unstable base might also not help

I also would recommend Pete, he serviced my Howtek and I have not had a problem since, If you have not tried it then Silverfast scanning software has made my life a lot easier.

The only problem I have got is trying to wet mount 10X8 negs, and the least said about that the better

[IanMazursky]

Speaking of the bulb, use the front panel and access the testing functions.
Turn on the transmissive lamp, open the lamp access door.
Look at the light coming out of the bulb and see where it collimates onto the fiber bundle.
The center brightest part should hit the middle of the fiber bundle, if not unscrew the lamp assembly and adjust it.
Just be careful they get hot!

Onto the base, i have placed all of my scanners on rubber vibration isolation pads.
This has helped tremendously with shaking. Even though our tables are stable the vibration from the scanner would still it to shake lot.
The pads are really necessary to reduce it to a manageable level. I bought some from McMaster Carr.

[georgl]

I will do further testing within the next days and keep you updated!

Thanks to "onnect17" as well for his help!

[georgl]

I have time again to play with my Howtek. Any news from those who are dealing with a similar problems?

Thanks for those who tried to give me advice - but so far nothing really helped!?

- the lamp works fine, no calibration, no flicker-issue
- the newer drum sits fine, the pads are like new, no rattling, nothing.

I cannot imagine that it has something to do with the axial movement - the alignment of the "columns" is fine but their height (along the rotation of the drum) varies - I've tried to paint the problem.

So the problem must be within the rotation - or am I completely wrong? Driving the rotation with a rubber belt is a pretty stupid design, but it might also be the bearings. These are standard-bearings, that can be replaced?

I will try something else: I will add weight to the drum and therefore stabilize the speed of the rotation.

Let's see what happens...

[Don Hutton]

I had some issues with my 4500 a while ago and cured the problems by removing all the boards and using Deoxit D5 on all connections. I also placed the machine on a "cleaner" power supply (by running it through a decent UPS unit).

[georgl]

I've tried it with adding weight (and therefore "smoothing" the rotation) and it didn't work.

I've found a "stupid" way to deal with the problem: multisampling with Silverfast. With 4x multisampling it's nearly gone, with 16x hardly detectable - now I need 4h to scan one 35mm slide and propably one week for a 8x10"-slide ;-) !!!

http://img27.imageshack.us/img27/5186/streakmulti.jpg (http://img27.imageshack.us/i/streakmulti.jpg/)

But Silverfast went completely nuts, I've reinstalled it (even a newer version), deleted the pref-files - nothing works - it "forgets" to add the adjustments from the pre-scan to the final-scan. The pre-scan looks fine, the final scan is completely off! Therefore I cannot scan negative material anymore, it looks like this:

http://img825.imageshack.us/img825/2677/ektar1000.jpg (http://img825.imageshack.us/i/ektar1000.jpg/)

Alternatives like Trident don't work, I've tried Win2000 - it killed my XP and I had to reinstall my whole system and I can't go beyond 16 colors because there are no pre-XP-graphic-drivers available...

Occassionally, I get the "F715"-error while it's calibrating - does that mean something?

I hate this machine!

Any news from the other Howtek-"patients"?