Nikon Coolscan 4000 intermittent LED self test failure

I've acquired a Coolscan 4000 in questionable condition.

It successfully starts about half the time and I can see the LED RGB self test running. And the other half of the time it makes the same sounds but I don't see the RGB's running and then it errors after the second movement sound.

Just wondering if anyone has any insight about whether this likely indicates imminent total failure of the light source, or can be explained by an adjustment problem like alignment or connection re-seating.

I have not got it set up with a computer for a test scan yet but will in the next day or two and will report the result of that.

It was represented as in used but working condition, and also represented as intact but arrived with a totally destroyed faceplate. So I have the option to get a full refund but it was cheaper than the going rate for a good condition model so I am still interested to see if it can be salvaged into functional operation.

My guess is that it took a fall during transport that broke the front and an internal plastic part.

I think my CS 4000 had the same symptoms when I got it. I couldn't get a refund so i "fixed it". It's not like new, but it does scan.

hwy17 said:

So I have the option to get a full refund

Click to expand...

Frankly, that sounds like the way to go. This unit evidently suffered major damage. You might be able to fix an intermittent light source problem, but what's next? The moment you start fixing things on the device, it basically implies you've accepted the unit and you can't go back to the refund.

I agree that a refund is the reasonable way to go.

Having said that, these scanners can be hard to find at a fair price, so I understand if you hesitate to let it go.

I now remember this page, it has good instructions for disassembly:

This image shows where a small plastic past can break if you drop the scanner. It's 24 steps to replace it. But you don't have go many steps before you can see the part and check if that's your problem.

Yeah I'm fine navigating my options to proceed as far as the purchase, just wondered if anyone had encountered the specific symptom of intermittent light source self test failure.

I don't remember the order of the steps in the calibration, but it will abort when it comes to a step that fails. Have you established that it is the LED test that fails and not a test before that?

FotoD said:

I don't remember the order of the steps in the calibration, but it will abort when it comes to a step that fails. Have you established that it is the LED test that fails and not a test before that?

Click to expand...

I could still study it more carefully but it does seem to me like it completes the home point detection successfully and then only errors out at the LED testing step.

Ok, that was a good webpage. My scanner failed on step 3. It wasn't the optical interruptor. It was a mechanical problem that prevented the carriage to trigger the interruptor. And it was possible to fix.

If your scanner has problems on step 4 then it could be more serious. Hard to know.

FotoD said:

Ok, that was a good webpage. My scanner failed on step 3. It wasn't the optical interruptor. It was a mechanical problem that prevented the carriage to trigger the interruptor. And it was possible to fix.

If your scanner has problems on step 4 then it could be more serious. Hard to know.

View attachment 401471

Click to expand...

That's interesting about that source naming failure at this step as either LED failure or motherboard VRM. Because it seems less likely to me that an LED would failure intermittently, but a VRM certainly would. I hope they're surface mounted rather than pad mounted.

To satisfy people's curiosity though, the way I'm considering proceeding is with a partial refund to accept it.

I know a lot of people would rather just spend more on one good item, but I'd enjoy more to have several questionable examples if I can frankenstein them into a working one and be left with a parts library.

Sounds like a good compromise.

Verifying the different power supply voltages sounds like a good idea. I do wonder if some are activated only as needed; this can be confusing if you're measuring voltages and they magically disappear and/or re-appear, which does not necessarily indicate failure.

But really, I'd start by very carefully checking for any apparent mechanical damage. The unit has dropped hard, evidently, so the most probable cause of this init failure IMO is some kind of mechanical damage resulting from the same fall.

Wow the Apple Thunderbolt 2 to Firewire 800 adapter is going for $150 on Ebay. I got an order in on Amazon for a much more reasonable price, but I'm doubting that the seller is actually going to fulfill it.

This might have an impact on the market value of Firewire Coolscans if they become as expensive and difficult to connect as the SCSI models.

SCSI in today's digital world is a downright b*tch. I can definitely see how this makes SCSI Coolscans less attractive. Most people won't bother even trying to hook them up to their new PC's and Macs, even if there is a way to pull it off to begin with.

koraks said:

SCSI in today's digital world is a downright b*tch. I can definitely see how this makes SCSI Coolscans less attractive. Most people won't bother even trying to hook them up to their new PC's and Macs, even if there is a way to pull it off to begin with.

Click to expand...

At this point I'd almost rather deal with SCSI than firewire, haha. Maybe not if I actually tried it. I had a motherboard with Texas Instruments Firewire 400 on it but I tried to get a windows install going on it yesterday and it seems like it might have died in the process. Startech PCIe Firewire card with the TI chip is next.

Here's my adventure trying to coax Windows 10 to play nice with a SCSI Imacon scanner: https://tinker.koraks.nl/photograph...tight-precision-ii-scanner-on-windows-10-x64/

hwy17 said:

At this point I'd almost rather deal with SCSI than firewire, haha. Maybe not if I actually tried it. I had a motherboard with Texas Instruments Firewire 400 on it but I tried to get a windows install going on it yesterday and it seems like it might have died in the process. Startech PCIe Firewire card with the TI chip is next.

Click to expand...

To make matters (slightly) worse, firewire stack is going away with this year's macOS 26 and I doubt Apple will go out of their way to keep FW around (semi-officially) the way Windows has. So, even the expensive Thunderbolt2-FW800-FW400 adapter chain will become useless for Mac people (unless you stop upgrading or keep a dedicated scanning machine around).

I purchased a Startech low profile 1394 PCIe express card. Didn't expect to have that expense but it's a nice product with a good Texas Instruments Firewire controller that works automatically in Windows 11 without any manual driver installation.

I am using Nikon Scan 4.03 on Windows 11. The app installs and runs fine without any intentionally applied compatibility settings. I got that installer from an ISO of the installer disk hosted by Internet Archive as I could not find the download link on Nikon's website, even though I found the page that looks like it would contain that link. In addition to that the only particular trick required is I had to install the vanced driver for the LS-4000 from Josh Waller which required disabling windows driver signature verification.



The great news is that the bashed up LS-4000 works! At least this time. Perhaps it still has an intermittent startup failure but when it starts it works. I did get my partial refund for the damage.

Also great news is the SA-21 purchased in parts/repair condition works too!

It does not work with an SA-20 that I also purchased in order to have multiple carriers for troubleshooting.

It does not work with an MA-20 that came with my other Coolscan IV.

The Coolscan IV I purchased in the meantime somewhat out of haste. Purchased in parts/repair condition with a side bash and a faceplate crack. It looks like the LED tries to self test but I only see some of the RGB colors not all and it reliably fails every time. This purchase feels a bit regrettable but I did say I wanted to acquire a parts library so there way go, maybe this one will be the donor to save the 4000 one day.

I think I will go ahead and modify the SA-21 now to see if I can get it to scan whole rolls in the 4000.