Lens chip reverse engineering?

[hannu]

Hi,

 

I've been reading the articles about manual Nikon lens compatibility

issues with newer non-pro bodies (and also saw that conversions are

offered). This hasn't been a big deal for me yet, since I put my F65

on a shelf more or less permanently when I got a FM2n and a few

beautiful AI primes. However, now I've started to drool after a D100

(D1x is too expensive for me, especially here in Europe) and it seems

like I come up against the same issue again.

 

Just out of academic curiosity (the sort that kills professors'

cats), has anyone reverse engineered D or non-D chips? From what I've

read it seems that ignoring the distance information, the chip only

passes focal length and maximum aperture information -- for a prime

these both would obviously be constants.

 

My 50mm AF-D has five contacts, one bit more separated. Maybe that's

Vin and the lens mount is ground (just guessing here and I know

nothing of electronics). 4 bits for data would be only 16 different

signals, so I guess some non-binary coding is used.

 

Any theories? With my soldering and metal working skills, I'll

probably just pay for a conversion, but would be interesting to know.

As matrix chips seemingly are not available for all focal

lengths/apertures, this info might even be used in making an ersatz

chip.

 

If there are any Sigma or Tamron employees around... :-)

[nico_.]

Please have a look at these threads:

<p><p>

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=004xlN

<p><p>

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=004HDj

<p><p>

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=004NPv

<p><p>

It's all in the matrix ;-)

[angel_o.]

Well, reverse-engineering is not a problem at all. You only need to know how the hell these signals from the chip are codified - unfortunately Nikon's bricolage support isn't that good, so they don't post their knowledge on the Internet. ;-)<br><br>

This is actually, exactly the problem of 3rd party lens manufacturers! They only can wait for new equipment coming out and test how it behaves and try to find out a working compromise - this compromise will never be a 100% clone of the original. Of course, their researching funds are much higher than of the normal do-it-yourselfer.<br><br>

I think that the only possible homesolution is to buy an original Nikon matrix chip of a similar lens and fit it into the manual lens.<br><br>

Head up!

[todd peach seattle, washi]

There are some pitfalls; the chip can in some instances affect how the camera AFs (or in your examples, 'focus confirms'). Here's a page I wrote up on how Sigma failed to anticipate multi-sensor AF systems, leaving me with a lens that worked on my N90s and not on my F100:<br>

<a href="http://www.thepeaches.com/photography/Sigma.htm">http://www.thepeaches.com/photography/Sigma.htm</a>

[mskovacs]

Read your story about the Sigma Todd. Couldn't be any worse than Nikon "forgetting" to add the aperture ring in its new designs ;)

 

The big missing chip in my mind is the f/2.5 chip that would allow a 105/2.5 conversion. Do the 3rd party guys sell any Nikon AF lenses with f/2.5 aperture? Could you install a f/2.8 chip and dial in a -1/3 stop exposure compensation?

[todd peach seattle, washi]

<I>

Could you install a f/2.8 chip and dial in a -1/3 stop exposure compensation?</I>

<p>

Speaking for myself, I could easily live with a lens that was 1/3 stop off. I'd probably hesitate to use it for Matrix (compensating a matrix meter is 'iffy' in my book), but it would be plenty adequate for enabling the CW meter on a D100.

[mskovacs]

Of course I could indulge my NAS and get a 105/1.8 which I likely can have chipped ;) Violates my 52mm filter rule though...

[ShunCheung]

It sure would have been nice to reverse/repair any chips on the front or rear elements on some old lens and make it optically as good as new. :-)