Disclaimer: if you happen to attempt reproducing what I've described below, any resulting damage, material or immaterial, is your own responsibility. In any case your warranties for the equipment involved will probably be voided. Otherwise, have fun!
Recently I've become interested in macro photography. I started out with a Soligor extension tube set, but when offered a PB-4 bellows for a decent price I couldn't resist.
While I could use the bellows with my manual focus Nikon body without a problem, the fact that the digital body which I have, the D100, was intentionally crippled by Nikon (as it doesn't do any sort of exposure metering without a CPU-equipped lens) discouraged me from trying out the bellows with it. I'm still so much an amateur that I have my hands full as it is, never mind trying to calculate macro exposure in my head when taking a picture! I wanted TTL measuring, just like I have with my FM2n. However, getting one of the pro digital bodies, such as D1X, which can do spot and centerweight measurement with manual lenses, was out of the question due to their price.
As it happens, Rolland Elliott found out some years ago that you can insert a CPU chip to an old manual focus lens and have a perfectly functioning exposure measurement even with Nikon's feature-stunted consumer bodies. He offers a cheap service to convert many, but not all, lenses, and he doesn't do bellows at all as far as I know. Then, I started to read Bjørn Rørslett's pages. He had got the idea of installing a chip into an extension ring, to use some very esoteric lenses -- and bellows just as well. Now this started to look like something I could do just as well...
First, I had to do the gathering of the necessary parts. For the chip, any chip would do, as the camera would not be in control of the lens aperture on the other side of the bellows -- any reported aperture will be fine, aperture will be controlled by manually stopping down (the lens will just make the camera think it suddenly got darker and lengthen the exposure time correspondingly). As I'm not a pro photographer or a camera repairman, I didn't even bother to try to order parts from Nikon directly. I went around the local camera shops, as well as repair shops, trying to look for cheap, broken AF lenses. Absolutely no luck there -- only a friendly, or in some cases unfriendly (after explaining what I wanted the thing for) 'no, we don't have them around'. Dang, don't those plastic Nikkors ever break?
Finally, I gave up and bought a functioning IX-Nikkor zoom (a 60-180 f4.5-5.6) for about 40 euros. A bit expensive, but I wanted to go ahead with the project! When I got home, I showed it the screwdriver and it promptly fell to pieces (no, really, one of the major construction elements was sticky black tape!). The chip print was actually surprisingly large, due to the two long contact strips which would convey the focal length and the focussing distance. That worried me a bit, but they seemed to be foldable.
I decided I'd sacrifice one of the Soligor tubes (the 20 mm one) for the purpose; it'd be easy enough to get a replacement if I needed it. I decided on the 20 mm tube for two reasons. First, the less extra extension would be introduced, the better -- you can always add some more, but to take any away is impossible. Second, from the three-tube set the 20 mm was the one which seemed it would actually manage to house the chip without much violence to the electronics! Then, off to work it was!
First I took the selected ring apart non-destructively to see whether the chip really would fit in. A problem with the Soligor set is that it's made of plastic -- to enhance the structural strength, the inside of the tube is not at all smooth but full of little strengthening ribs. However, it still seemed that with a bit of tucking and pushing in the chip would fit.
I decided to discard all of the mechanics which I possibly could, meaning the AF screw, the aperture lever, and the lever which conveys the lens speed index. All of these would be useless on the bellows. I had to leave the meter coupling ridge bit in place, because it also blocks light from entering the ring.
After that, the insides of the ring contained a lot of unneeded metal which probably just induces vignetting (heh) so I decided to grind it away with a Dremel.
This was, in my opinion, the only tricky part in the whole operation. You have to get the electronic contacts of the chip aligned just right, so that they mate with the camera properly when the bayonet locks into place. I decided to use a decidedly low-tech technique for this: mount the loose male bayonet into the camera, press the chip into position with fingers and mark the position on the bayonet with sticky tape. Then, take the bayonet out and drill the holes for the screws. Worked like a charm -- except that the first time 'round, I managed to slip the chip 2 mm to the side before marking the position. Groan. Fortunately, I managed to make the second set of holes without problems.
The main electronics of the chip just fits when the extension ring is reassembled, but the aforementioned contact strips required a bit of work. I managed to tuck them here and there, though not very cleanly.
Below you can see a couple of pictures of the result.
 Fig.1. Here you can see one of the contact strips, held in place by sticky tape, what else. To reduce internal reflections, I inted to cover the inside with matte black material. The grinding marks are a bit ugly, but then again, it's not a beauty contest here. |
 Fig.2. Another view, showing more removed brass grind marks and the installed contact block more clearly. |
After putting everything together, I used -- cautiously as ever -- my F65 as a guinea pig. Mount the ring, turn it on, and YAY! My neighbours must've wondered at my rejoicing. An aperture showed on the camera, the light meter would respond -- perfect! And on the D100 it worked just as well!
The maximum aperture that happens to show is f5.3. After checking from the EXIF information, the focal length reported by the camera is 95 mm. Both are, as far as I can say, due to the fact I did not short the zoom contact strip at any particular position; apparently when the chip gets all zero bits from the strips, the above numbers are what it maps to. If you really want to, you can probably debug what gives what values by connecting the chip to the camera and poking the metal forks, which slide over the contacts, from the lens into different positions. I didn't consider it worth the trouble; the only problem that might arise would be in conjuction with flash (distance information for D-TTL purposes would be all wrong) by I've yet to try that out.
Note that the aperture on the camera has to be at the maximum value, as there is no way the camera can control the aperture of the lens at the other end of the bellows (and I took the lever coupling away, anyway). Otherwise the pictures would be overexposed. Good old stop-down together with TTL metering will determine the right exposure when the camera doesn't try to close the aperture. Note also that the same reason, lack of aperture control, will prevent using S or P modes -- just M and A are available, but that's what I always use, anyway.
 Fig.3. Playing the accordion on the F65. The 'E' on the camera indicates lack of film, in case you were wondering... |
 Fig.4. Finally, a demo picture; taken with a 80-200 at 80 mm position and an added extension of 20+60 mm. The subject is a detail from a 20 euro banknote. White balance is a bit off (evening sun plus incandescent lamp) but nevermind. |
I took one or two evenings to examine the parts, and another two evenings to actually do the work. I intentionally worked slowly not to make too many mistakes; the one I did was not fatal, which is an accomplishment in itself since when it comes to mechanics I'm basically all thumbs! Add to that 40 euros for the IX lens, and about the same for the 20 mm ring; in the end it's not such a huge investment. After all, what photographic equipment do you get for 80 euros? A couple of filters, maybe. So, in my opinion, it was well worth the trouble!
Copyright (c) Hannu Mallat < hmallat at gmail dot com > 2004. Last updated Jun 11, 2004.