attention EE engineers and/or Nikon insiders

[darren_cokin]

I wonder if there's anyone out there, that either has access to internal Nikon

specifications, and wouldn't mind leaking a little bit of info, perhaps an

electrical engineer with an oscilloscope and a knack for reverse engineering.

<br><br>

 

I would like to know what electrical signals need to be sent to a VR lens to

turn the VR feature on and off. I assume it's more complicated than just

applying a given voltage across two pins (but wouldn't that be great if it was).

Rather, I assume a constant voltage is applied across two pins, and a binary

code with a "VR on" or "VR off" code is sent to other pins. I want to know the

appropriate pins, voltages, and on/off codes. <br><br>

 

Bonus points if you can also supply a circuit diagram on how to replicate that

signal with radio shack parts! <br><br>

 

What's this about? Read my last thread: <br>

<a

href="http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00HfRR">http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00HfRR</a><br><br>

 

 

I built myself a Lens Scope Converter, but wouldn't it be so much nicer to have

a Scope with Vibration Reduction? It bugs me that I've got this really useful

feature built into my lens, but I don't know how to switch it on... <br><br>

 

I considered buying an old N65 and ripping the electronics out, but I found the

service manual, and became discouraged when I saw the electronics are

interspersed over several flexible circuit boards that would be difficult to

remove and repackage. Leaving the electronics in place, and removing the

shutter box and installing the eyepiece in it's place would be easier, but it

wouldn't be a very elegant looking contraption, now would it? <br><br>

 

Darren

[work-page]

It's not likely that a simple on/off signal suffices. The VR funtion is only available on Nikons with at least 5 focus points. Makes a lot of sense, as the AF sensors are presumably used to measure the amount of movement in the vertical and horizontal direction (with complementary sensitive sensors). Given the limited number of contacts, the corrective signals are probably coded and multiplexed,

just like those for aperture, focal length, distance, AF-S and whatnot are..

[roto]

The lens movement (or, better, acceleration) is measured by solid state accelerometers in the lens itself. The AF sensors are not used for that and they could not.

 

Some time ago I found an exploded view of the 70-200 VR where the accelerometers where highlighted. I can't find it any more. If I do, I'll post a link.

[briany]

Makes sense, Robert, but then why the need for 5 AF points? Regardless, I think Darren is swimming upstream, but it never hurts to try... Even with your N65 without the shutter idea, if the VR is dependent in some way on the AF sensors, don't the AF sensors rely on the mirror?

[joseph_wisniewski]

The "5 AF point" rule indicates the age of the camera. It's just a coincidence that Nikon was able to find a dividing line like that.

 

N2020, 8008, 4004, 5005, F4, N90 are too old to have the proper message set, and they all happen to be single AF zone.

 

F100, F5, N75, N80, F6, D2X, etc. are new enough to have the right message set, and they all happen to be 5 or more AF point.

[work-page]

The "5 AF point" rule indicates the age of the camera. It's just a coincidence that Nikon was able to find a dividing line like that.

 

If only things were that simple, the F55/N55 is decidedly more recent than the F80 and F65 cameras. It's limited to 3 AF sensors, and doesn't do VR...

[darren_cokin]

Peter,<br>

I doubt it's a multiplexed signal. Even if it were, what I'm proposing wouldn't be impossible, just a bit more difficult. But considering that really not all that much information needs to pass between the camera and lens, I doubt it needs to be that complicated. (Lens tells camera it's focal length, and focus distance. Camera tell lens to adjust focus fore or aft, to turn VR on or off, and provides power. Anything else?)<br><br>

 

And I believe Roberto is right, the VR and AF systems must work independently of one another. (Note that VR does not stop working in low contrast situations, when AF does.)<br><br>

 

A few months ago, I studied the repair manual for the N65, which includes a full schematic and circuit diagram. The Front Body FPC (flexible printed circuit) seems to have 8 connectors for the lens. They are identified as follows:<br><br>

 

VCC<br>

R/W1<br>

SCK<br>

SI/O<br>

R/W2<br>

LBAT<br>

LGND<br>

DGND<br><br>

 

But, I don't know what all those stand for. Any EE engineers recognize standard abbreviations? One could assume that LBAT and LGND are battery power and ground, and DGND is a digital ground for the signals, but I can't hazard guesses for the others.<br><br>

 

You can view the N65 repair manual, parts list, and plenty of other old Nikon documents here:<br>

<a href="http://www.micro-tools.com/pdf/Nikon/pdf/">http://www.micro-tools.com/pdf/Nikon/pdf/</a><br><br>

 

That would be a good place to start, if you're interested in reverse engineering.<br><br>

 

Darren

[work-page]

The lens tells at least

 

- focal length, but probably also the range for everything ranging from programmed exposure to speedlight zoom settings

 

- distance (AF-D/AF-G)

 

- aperture range, to allow P and S mode to work

 

- aperture correction, to display correct A on variable aperture zooms and macro lenses

 

- supported functions such as AF or AF-S to tell the camera whether or not to drive any of these, or whether they're disengaged on the lens (can't have shutter lock in manual focus mode)..

 

Both from the manual and from peeking inside my camera, there seem to be 7 connections, not 8. And I couldn't make out whether those quoted were correct. But the most probable meaning of the mentioned connections is as follows:

 

VCC = supply voltage for integrated circuits (regulated voltage)

 

DGND = ground voltage for integrated circuits

 

LBAT = battery power for driving motors/actuators (unregulated

voltage)

 

LGND = battery power ground voltage

 

SIO = serial input/output for data/control, bidirectional

 

SCK = serial clock for timing SIO signals

 

R/W = don't know. Could be anything e.g. read/write to control the direction of SIO

[work-page]

The F55 (no AF-S, no VR) repair manual page 75 shows the following 7 connections to the lens:

 

VCC, R/W1, SCK, SI/O, R/W2 (Unconnected), ? (Unconnected), GND

 

called respectively VCC, LENSB, LENSC, LENSD, nothing, nothing, DGND on the flex cable..

 

From this one can imagine that the '?' pin on the F65 and upwards serves as battery power to power the motors and actuators in AFS and VR lenses..

 

I've found some on the accelerometers on the VR lenses:

http://www.nikon.co.jp/main/eng/portfolio/about/technology/nikon_technology/vr_e/index.htm

 

Apparently VR operates in different ways depending on whether the shutter button is pressed halfway (smooth), just before exposure (center) or during exposure (correct all).. So, it's not just a simple matter of on/off that the camera sends to the VR..

[darren_cokin]

<i>there seem to be 7 connections, not 8.</i><br><br>

 

Agreed, my D100 has 7 visible connectors too. (Maybe the 8th I saw in the schematic is a ground connected to the mounting ring?) <br>

 

My least advanced lens (50mm f1.8) only uses 5 of these connectors. But my most advanced lenses have 10 pins on them! Obviously, there's some future capability Nikon has planned for.<br><br>

 

<i>Apparently VR operates in different ways depending on whether...</i><br><br>

 

It depends on which lens. My oldest VR lens is the 80-400, which has a 3 way switch - VR off, VR on momentarily when you take a picture, and VR on continuously while you press the shutter halfway. But I doubt the camera knows or cares which way the switch is set, and sends the same signals to the lens either way. <br><br>

 

The newest models just have a simple on/off switch for the VR feature (or on/off plus active/normal in the case of the 18-200). <br><br>

 

So, maybe there isn't s special signal from the camera dedicated to communicating "turn on VR now". But there probably <i>is</i> one that says "the shutter button is now being held down half way". <i><b>That is the signal I'd like to replicate.</i></b><br><br>

 

Darren

[joseph_wisniewski]

If you're really going to try this, track down photo.net member "Ro Ro", who has done a lot of work on decoding the messages.

 

http://www.photo.net/shared/community-member?user_id=1831021

 

And you might want to have a look at these.

 

http://forums.dpreview.com/forums/read.asp?forum=1030&message=14971871

http://forums.dpreview.com/forums/read.asp?forum=1021&message=10974743

[rffffffff]

You might be barking up the wrong tree greatly, but I know nothing about lenses specifically.

 

The likely difference in the ability of the body to handle VR is in physical lens contacts and/or current handling capability.

 

The VR fuction is powered from the camera, but likely handled completely inside the lens.

 

I bet, but wouldn't personally try myself on a $1500 lens, that the only thing you need to get VR working is the correct battery voltage on the correct pins of the lens. I would think, but I am not sure, that the 'switch on' that the VR does when you press the shutter button for the first time is the camera applying voltage to two pins.

 

If I am correct, its likely that you could figure this out with a digital meter. Figure out if there is a pin that goes hot (referenced to the ground pin) in the corresponding time that VR would work. If so, all you need is a battery and a switch.

 

The alternative is that there may be a hot pin and an active pin separately... the hot is always on and only when the active pin goes high or low the VR works.

 

Its probably that simple, but who knows... A battery and a switch might be all you need, but then you may also have to deal with some current limiting or conditioning circuit to prevent the VR circuitry from breaking...

 

On the other hand, I seriously doubt that there would be any benefit from VR in this context. VR almost hurts my eyes when I am shooting with its wacky movements and jumps.. I think I would prefer a scope without!

[darren_cokin]

Thank you Joseph, those old posts of yours you referenced were extremely helpful. I never considered that all the info would be in their publicly available patents.<br><br>

 

I don't really have a lot of time to delve into this project full steam right now, but I was very encouraged browsing the AF-S patent quickly. Lots of references to "motion compensation". According to figure 8, command 34H is what I'm after. Or, maybe that's a red herring and wasn't implemented (you can't disable the VR from the body after all), in which case probably all I need to do is close "switch 34" (the shutter button halfway down). According to Figure 6, pushing this button sends regulated voltage directly to the lens via the VCC pin (and also the camera's processor). The lens also gets unregulated battery power to another pin continuously. <br><br>

 

Figures 11 and 12 show various input voltages and waveforms, which would be necessary to send a command to the lens. (It doesn't seem to include an index of commands anywhere though.) But maybe I'm lucky, the "motion compensation" command isn't really needed, and all that would be necessary is sending those two voltages. <br><br>

 

I don't really have the equipment or experience to actually measure the signals/voltages coming out of my camera. I'm a MechE, not a SparkE... <br><br>

 

If anyone else is interested, you can view the full patent here: <br>

<a href="http://www.freepatentsonline.com/5727241.pdf">http://www.freepatentsonline.com/5727241.pdf</a><br><br>

 

But, as that's a raster file, you can't search the text. But you could here: <br>

<a href=" http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=5,727,241.PN.&OS=PN/5,727,241&RS=PN/5,727,241">official US patent office website</a><br>

(That <i>should</i> include all the images too, but it's buggy on my computer, I only see the top inch of each page.) <br><br><br>

 

Making progress! <br><br><br>

 

<i>Robert wrote:<br>

VR almost hurts my eyes when I am shooting with its wacky movements and jumps.. I think I would prefer a scope without!</i><br><br>

 

To each his own... <br><br>

 

Darren

[work-page]

Hi Darren,

 

nice find on the patent! This explains a lot, and also tells about some stuff to expect in the future..

 

The R/W signals determine which way the communication between lens and body go.. The H/S(B) is from body to lens, initiating data transfer, the H/S(L) is from lens to body, enabling transfer. The command 34H needs to be sent to the lens to activate the VR feature according to fig.12.. So, you'd have to send a low pulse to the lens, wait for the signal low back from the lens, clock clk and simultaneously send 34hex on the si/o pin.

 

Something that's mentioned, but not present in any lens I know of is power zoom. This should get the rumour mill started..

[darren_cokin]

<i>The command 34H needs to be sent to the lens to activate the VR feature according to fig.12.. So, you'd have to send a low pulse to the lens, wait for the signal low back from the lens, clock clk and simultaneously send 34hex on the si/o pin.</i><br><br>

 

I'm thinking (hoping) that isn't not going to be that complicated after all. Considering that the "power zoom" feature spelled out in the patent was never implemented, I bet the "motion control" features weren't either. The patent just covers lens-to-body communication. But the VR on/off switch is on the lens, not on the camera body. So I don't think the body knows/cares whether the lens has VR, and doesnﾒt have to command it on. (There's no EXIF tag for VR status, is there?)<br><br>

 

I think all I need to do is send the appropriate conditioned voltages. If the lens gets power, and the VR switch is on, it should work. (Or so I'm guessing.)<br><br>

 

I'll need to send one voltage between VCC (1st pin) and GND (mounting ring), and another between LBAT (6th pin) and LGND (7th pin). Probably okay to intermingle the grounds too. (On my kenko extension tubes, the 7th pin is connected to the mounting ring.)<br><br>

 

Remember above there was discussion about why VR and AF-S work on an N65 but not an N55? It doesn't have anything to do with # of focus points. The N55 only has 5 pins to the lens, instead of 7. It's missing pins 5 & 6, which includes LBAT. (Well, it may have the physical pins, but according to the circuit diagram, they're not connected to anything.) So that's why VR won't work, it's not getting power.<br><br>

 

Still, I haven't found definitive documentation on what the appropriate voltages and currents should be, I don't have the right equipment to measure those values coming out of my camera, and also, I don't have the knowledge to design and construct a voltage conditioner... So I've still got a bit of work to do before I can test my theory. Maybe this fall I'll have some time for further research.<br><br>

 

But if anyone else wants to step in in the meanwhile... Could be a fun project. <br><br>

 

Darren

[work-page]

The F55 indeed doesn't support AFS and VR, and isn't fully connected, while the F65 is. But so is the F90, which I believe does work with AFS lenses, yet not with VR...

[gary_mayo1]

<p>Any interest in again pursuing this project? I would be interested in making a Nikon Lens Scope Converter.</p>

[mike_halliwell]

Doesn't look like it sadly...:(

[rodeo_joe1]

9 or 14 years on.... so maybe the protocols have changed. Nikon have certainly added more pins to their lens connector, and now we have E lenses to contend with.

 

Still, a bit more openness about the communication between camera and lens wouldn't come amiss.

[BeBu Lamar]

Late to the post but if the OP wanted to use VR lens for a lenscope then turning the VR on alone won't be sufficient. The camera has to continuously send data from focus sensors for VR to work.

[Andrew Garrard]

Does it? I didn't know that. I assumed it was just an active "on" signal - VR works with the F5, but wasn't really an active technology at the time, so I assumed the gyro and drive stuff was all in the lens. I know nothing, of course.

[rodeo_joe1]

The camera has to continuously send data from focus sensors for VR to work.

Huh? What have the Z axis, but very XY insensitive focus sensors got to do with the very sensitive XY accelerometer sensors (within the lens) needed for VR?

[mike_halliwell]

Does it? And if so, why?

 

I thought it was lens-mounted accelerometer driven?

 

If you turn off AF, VR still works.

[joÃobraga]

I came across this post as I would like to do the same...

I found this site witch explains the VR keeps checking if camera is in use to save battery, but the control logic is in the camera, so applying voltage should be enough when using off-camera. A spring type switch witch is only on when button is pressed could prevent battery drainage when not in use.

 

http://nicmosis.as.arizona.edu:8000/ECLIPSE_WEB/ECLIPSE_05/VR_KEEPALIVE.html

 

As for the pins, I found this one: F-Mount - Nikon Hacker

 

So according to these two sources, applying a tension of 6.2 Volt to pins 6 and 7 should put the VR to work and if all goes well enable a VR spotting scope.

 

Next question is what battery should be used? My first thought was to use the camera battery, in my case the Nikon EN-EL15, but this is rated 7 Volt, so I guess there is something in the circuit limiting the tension to 6.2V. Doesn't seem a huge difference, but not risking it on a good expensive lens.

 

Using a different battery would be better if it could fit inside the lens converter (not much space inside). But I only know the voltage (6.2 V), what other characteristics would be important to make it last at least 3 hours ? The EN-EL15 has 1900mAh and 14Wh, I don't know what to look for in a new battery. Anyone has a hint?

 

I will probably buy a broken lens on e-bay to test this stuff before putting to use in a good lens, but if you have ideas / experience I would be grateful to hear

[rodeo_joe1]

You could use a 7.4 v Li-ion battery plus 2 power diodes in series to get approximately 6.2 volts out the other end.

 

Any tighter voltage regulation would need a custom circuit, since most off-the-shelf regulator ICs need a 'head' of 2 to 3 volts to work properly.

 

There's no documentation I can find of how much current the VR circuit consumes. You need to know this before you can predict how long a particular battery will last while powering the VR.

 

Power consumption will also probably vary considerably from lens to lens, and with the amount of vibration that has to be compensated.

 

I suspect that bigger lenses, with heavier VR prisms, will take more current.

 

Current (in milliamps) x time in hours gives you the mAh capacity of the battery, and inversely, mAh capacity/mA drain gives you the usage time in hours.