Recently I bought a 10-year old Nikon Coolscan 5000ED, complete with SA-21 and SF-210. From the beginning I had problems with the SA-21 doing index scans, as color negative frames were not registered correctly. After cleaning the SA-21, the situation deteriorated to the point that it did no longer accept any film strip, but immediately returned it. I searched the internet for solutions, found that I was not alone with this problem, but could not find a really satisfactory analysis and cure. So I decided to do my own analysis. At the end I found a solution at least for my SA-21, that to my knowledge was not reported anywhere else in the internet. Because it may be of help to other SA-21 users, I want to report of my findings here.
My starting point was a description of a check of the three photo detectors built into the SA-21, given in
http://janburke.de/index.php/blog-en/130-replacing-photo-sensors-on-nikon-sa-21-film-feeder. In this blog it was argued that IR diode wearout was a problem, but it came without a live detector signal analysis. So I decided to do this analysis. I traced the 3 photo transistor collectors to the big blue connector at the rear end of my SA-21, soldered wires to them and brought these wires to the front end of the SA-21, inside its housing. With a storage oscilloscope at hand I was able to trace these signals while inserting a film strip. These were the results:
1) with no film inserted (detectors should detect a “dark” condition), the front and rear detectors showed a 1,9V level, the center detector a 3,5V level
2) during insertion and transport of a film strip, the detectors showed pulse trains with the above low levels (when transport holes passed) and high levels approaching full 5V (apparently when film was detected = “light” condition)
My conclusion was that I had no problem with IR diode wearout or film reflectivity, as the “light” level was perfect. But I had a problem with the “dark” level at least with the center detector, where the signal swing was only 5V-3,5V=1,5V. From the above blog and my cleaning action I knew that the center detector is constructed different from front and rear. Front and rear have open spaces above them so that IR light, passing through film transport holes (“dark”), is not immediately reflected. But in case of the center detector, the spring loaded film holding frame sits above it and offers only a small cavity for IR light to disperse in “dark” condition. Thus the reflectivity of this cavity plays an important role.
To improve the “dark” level of the center detector as an initial trial, I cautiously applied a little of my wife’s mascara to that cavity. As a result, all problems were gone! Film strip feeding and index scan frame registration were without any flaw. I have worked through >300 film strips in this constellation since. By the mascara application, the “dark” level of the center detector improved from the previous 3,5V to 2,4V, obviously enough to offer a robust signal swing to the 5000ED mainframe.
With these findings, the former failure scenario is as follows:
After film strip insertion by the user, the SA-21 begins to pull in, first past the front, then past the center and finally past the rear detector. The front detector signals presence of film, but the center detector fails to do this. As soon as the rear detector again signals presence of film, the 5000ED mainframe determines an implausible condition, aborts and returns the film strip.
I regard my mascara solution as a hot fix. But so far I was unable to find some anti-reflective paint as a permanent solution. My suspicion is that during my cleaning action I inadvertently removed some coating already present in the holding frame cavity. But this former coating must have been insufficient to guarantee a robust signal.
Maybe this helps to revive one or the other SA-21.
Burkhard
