How the F35 Camera Dock Communicates - Unlocking the Missing Features

V

Viddovation

Veteran
So, I've been digging into how the Sony F35 dock works, and while the information I have to share is preliminary, it's encouraging. In short, with some development effort, it should be possible to link communication between the camera and some of the newer recorders by using an intermediate microprocessor. Exactly how this would work, and the degree of limitation is going to depend on the recorder you want to sync to, but enabling record triggering from camera and enabling tally lights seems in the realm of feasibility. For recorders like the SR-R1 complete record integration with timecode fed back to the camera as well as a playback signal seem feasible as well.

Please understand this is all preliminary, and based primarily on reading, mixed with a bit of monkeying with the parts. But here are some of the basics:

  • The camera dock offers power, some clocking signals, up to 4 SDI outputs, and two SDI inputs, as well as a serial digital interface, based on IIC.
  • The IIC or I2C serial protocol is really the centerpiece of the conversation here. It's a very common protocol that allows integrated circuits to talk to one another. It only uses two cables!
  • While I haven't done so yet, it's apparently fairly straightforward to passively intercept I2C bus communications. From there, they can be recorded, and reverse engineered.
  • It's not a very complicated protocol, is unlikely to have any encryption or obfuscations, and should be easy to simulate in cheap microchip hardware like the Arduino, Pic16, Pic24, or Pic32.

So what can this do for us? Well, at the very minimum, by replacing the existing circuit board in the Interface box, we could potentially lie to the camera and tell it that Yes! there is a SRW-1 recorder attached. Even if in fact all our microprocessor did is lie, and had no communication with our actual recorder, this would cheat the camera into enabling it's own auxillary SDI record signal out of the monitor port!

But if our microprocessor acted as a bridge between the camera and our recorder, we can do one better - let's take a recorder like the Atomos Samurai that has a simple LANC record protocol. If our microprocessor can talk I2C and talk LANC, it can act as a translator, passing the record signals on, and confirming receipt of messages back to the camera. To the extent the camera asks for information the recorder can't provide, the microprocessor can lie if necessary.

If your recorder is like the Sony SR-R1 and has a remote port for full deck control, the potentially available options are even cooler. In theory, virtually complete integration with the camera, including record button usage, tally lights, timecode return to the camera, playback of recorded video to the viewfinder and monitor out ports is all possible, simply by using a microprocessor to bridge the I2C communication with the Sony 9-pin remote deck protocol.

It may also be possible to use this research as part of a process to get the camera to offer quad link SDI from a single dock port.

So what's this thread for if nothing's been made yet? Well, for starters, if anyone else has explored this road, and would like to share what they've gleaned, that would be awesome. I'd rather not spend development time on processes that someone else has already researched it.

Beyond that I'm simply happy to have a conversation if anyone has general thoughts or questions about this.

Downside? Now I pretty much will need to get my hands on an SRW-1 so I can start the reverse engineering.
 
Sounds exciting. I for one would be very interested in getting power out of the dock, if for no other reason than the 12V out on the side of the camera is insufficient for the Cinedeck.
 
I haven't tested this fully, but I suspect getting unregulated 12v power out of the dock would be exceedingly easy. That said, I have no idea on how much would be safe to draw in that manner, or if it will fare any better than the 12v on the side. Now, the SRW1 with all options draws 62W, so I'd have to say the dock can supply that much safely, so long as the power source for the camera has that much to give.

How are you powering your F35? Because if your power source is insufficient, that would be a determining factor both for the side and the dock.
 
Thanks for all the great info, what are the 2 SDI inputs that you said are in the dock connector? Is this to input timecode or sync for 3d? I wish I had paid more attention in programming class in high school, I actually have a rasberri pi, but I've only used it like a media center, haven't really learned how to come up with custom applications.

I wouldn't stick with the original IF box design, why not keep the Ca joint and build a new box, I feel like a thinner but wider box looks more natural on the back of the F35 anyways, the Sony IF box is too thin and looks funny because the recorder or even battery plate usually surpasses it in size. Of course I'm talking about more work, but I've found a few pre made metal boxes that might do the trick.

I have found a few references to 444 50P being sent over a single dock connection so I think you are right. If we had a breakout to 4 SDI cables, would this work with the 7Q 50P? I think the box that makes the F35 compatible with "modern" workflows is a new IF box with 4 SDI outputs of the 50p 444 signal, (or is there a multiplexer inside the box for all the pix users). There is a lanc connector, timecode input. The box powers the camera internally from a plate on the backside connected through the inside.

My biggest complaint about my F35 setup is constantly cables, I am always loosing HDMI. I think the solution is to find a deal on the good sony viewfinder or get an alphatron with SDI. If you look at any cinema camera setup out there it is covered in cables, and that's acceptable because it's necessary. I suppose it's more of an ENG appeal of a clean solid camera with no cables, no arms that you can pick up and swing around without worrying about catching parts or having things snag.

I've been on the fence about a new recorder for a long time, I really like trying to conform the F35 to the smallest most compact form. I really wish the Ki Pro Quad had the ability to record the 12bit signal because that would be the best recorder to put on the back and connect the Ki Pro right to the dock connection, the ki pro has a nice thick plate that can be used to mount it which would be perfect. Everything could pass through the dock including the power and there would be no cables except for maybe an HDMI or SDI to monitors. I wonder if CD would consider making a version of it's recorder without the screen. I know for many, a 7 inch screen is awesome, and red has adopted the look of a big LCD on the top of the camera for years. However I'd like to emulate an ENG run and gun setup with my F35 and a recorder in-between the camera and battery plate works the best.

I've just got the Codex recorder in, I haven't even powered it. I was waiting to go pickup the F35 plate until I'd tested that it worked at least. Something interesting, I got the Arriraw version of the Codex M recorder, they are almost identical in design, the arriraw is flat on one side that the Codex M has the grooves. I have to check that the one I got has the right licenses otherwise I have to purchase them for the F35. Codex is incredible with customer support, I emailed a few questions on sunday and got a response in 5 mins and then a phone call monday morning. I don't know how far I'm going to pursue the codex, it's still a lot of $$ for all the parts. But if there was any information I could get you you from the codex, I'd be happy to help. I'm actually very interested in what the Codex F35 plate looks like on the inside.
 
yachacha said:
Thanks for all the great info, what are the 2 SDI inputs that you said are in the dock connector? Is this to input timecode or sync for 3d? I wish I had paid more attention in programming class in high school, I actually have a rasberri pi, but I've only used it like a media center, haven't really learned how to come up with custom applications.
Click to expand...
I'd bet good money the two SDI returns are just dual link SDI video returns. There's good reason for this. Linked with the SRW1, this enables viewfinder and monitor port out playback review of recorded images. And remember, this same dock is shared with the SRPC-1, so all the pin-outs have to match. And when linked to the SRPC-1 the SRW1 can be used as an ingest/playback deck. So it absolutely makes sense that it has output feeds. The fact that it only has 2 outputs to the 4 inputs may be an engineering or cost limitation, but it matches with the information I posted in another thread about the SRW-1 not being capable of true RGB 444 50p/60p playback, even though it can record it.

yachacha said:
I wouldn't stick with the original IF box design, why not keep the Ca joint and build a new box, I feel like a thinner but wider box looks more natural on the back of the F35 anyways, the Sony IF box is too thin and looks funny because the recorder or even battery plate usually surpasses it in size. Of course I'm talking about more work, but I've found a few pre made metal boxes that might do the trick.
Click to expand...
You may be right, but the ideal Interface is going to depend on exactly what recorder you're using and other user specific parameters. I started this process for my own desires, which center on the Sony SR-R1. I'm glad to see the potential for other recorders, but the ideal box shape is going to vary. And there are cost issues, etc. If there was ever a move to make a groupbuy, we'd have to be using a design model that is mutually beneficial while keeping cost affordable. Mind you, that's all a long way off and hypothetical. But one really cheap way to do this is to make the new circuit board the dimensions of the circuit board in the IF box, so you could just switch them out.

yachacha said:
I have found a few references to 444 50P being sent over a single dock connection so I think you are right. If we had a breakout to 4 SDI cables, would this work with the 7Q 50P? I think the box that makes the F35 compatible with "modern" workflows is a new IF box with 4 SDI outputs of the 50p 444 signal, (or is there a multiplexer inside the box for all the pix users). There is a lanc connector, timecode input. The box powers the camera internally from a plate on the backside connected through the inside.
Click to expand...
444 50p is available over one dock port when using the SRW-1. If you look at the parts or the schematics, the dock piece for the IF box only has three SDI cable ports, one of which isn't used (the return), but the one for the SRW-1 has 6 (4 in, two return). So if you use the SRW1 dock piece, and supply the right I2C communication, it should be possible.

yachacha said:
My biggest complaint about my F35 setup is constantly cables, I am always loosing HDMI. I think the solution is to find a deal on the good sony viewfinder or get an alphatron with SDI. If you look at any cinema camera setup out there it is covered in cables, and that's acceptable because it's necessary. I suppose it's more of an ENG appeal of a clean solid camera with no cables, no arms that you can pick up and swing around without worrying about catching parts or having things snag.
Click to expand...
My SR-R1 mod is going to eventually resemble the F65 and SR-R4. The power, video out, video return, data, deck control... it will all run internally through the dock port. That's the idea anyway. No only does it reduce cable snags and troubeshooting when a BNC goes bad, it will make the rig pretty idiot proof for the ACs.

yachacha said:
Codex is incredible with customer support, I emailed a few questions on sunday and got a response in 5 mins and then a phone call monday morning. I don't know how far I'm going to pursue the codex, it's still a lot of $$ for all the parts. But if there was any information I could get you you from the codex, I'd be happy to help. I'm actually very interested in what the Codex F35 plate looks like on the inside.
Click to expand...
Well, presumably Codex knows what the I2C protocol commands are for the F35. So I'd love to know those! But I doubt their contracts would permit them to, nor is it particularly in their business best interest to share that info.
 
I power my F35 from a Hawk-Woods Li-Ion battery. It is wired up for 12 and 24v, but the 12v on it is much stronger than the 24v. It should be able to take the load of both easily - it does presently.
 
JMtheDP said:
I power my F35 from a Hawk-Woods Li-Ion battery. It is wired up for 12 and 24v, but the 12v on it is much stronger than the 24v. It should be able to take the load of both easily - it does presently.
Click to expand...
I haven't worked with the Cinedeck, but some quick Google work gives me some numbers.

A review says the Cinedeck EX draws 60-70watts. Meanwhile, the 12volt output on the side of the F35 is rated at 10.5 to 17volts, with a 5amp max. That means your max wattage draw is between 60-85watts. But in practice the voltage on that port isn't going to exceed the voltage the camera reads out for itself - infact, with additional draw the overall voltage will drop. So, if you're actually sending 12-13volts, you're getting max cap on that port of 60-65watts. That's not a lot of headroom for your Cinedeck. Not surprising it doesn't work well on that port.
 
Viddo

I should have elaborated. I am NOT using the 12v out on the camera to power the Cinedeck, because I know there won't be enough. I'm using the 24v output. Unfortunately, my battery only gives about 3a at 24v, which is also cutting it fine.

The battery can provide much more than that at 12v, but obviously the internal wiring in the camera (or something else) is limiting the output to 5a, hence not using the 12v port.

Hope that explains it a bit better.
 
JMtheDP said:
Viddo

I should have elaborated. I am NOT using the 12v out on the camera to power the Cinedeck, because I know there won't be enough. I'm using the 24v output. Unfortunately, my battery only gives about 3a at 24v, which is also cutting it fine.

The battery can provide much more than that at 12v, but obviously the internal wiring in the camera (or something else) is limiting the output to 5a, hence not using the 12v port.

Hope that explains it a bit better.
Click to expand...

Gotcha. Well, I'm pretty sure the camera is outputting 12v unregulated power from the dock connector by default, otherwise the IF box circuitboard wouldn't have operating power.

I think the connector between the dock connector circuitboard and the main board (ENC-105) is a 12 pin Molex Picoblade connector: https://www.elfa.se/elfa3~/elfa/init.do?item=43-007-47&toc=0&q=famid71464

If so, you could buy your own, and carefully replicate the pinout to get your 12v power. This might be a sub $100 solution, depending on implementation.
 
Quick update - I just paid nearly $400 and ordered a Beagle I2C Analyzer from Total Phase. It should permit me to spy on the communication between the camera and the IF box or SRW-1. If it works properly, it becomes a matter of decoding the commands and when/why they are executed.

Properly reverse engineered, the responses could ultimately be emulated by off the shelf hardware costing $50 or less.

Also, I've done some more concrete testing on getting power from the dock. The dock puts out unregulated 12v power, and the manual specifies the pull is not permitted to exceed 10amps. In theory that means it maxes out at 120watts.

Trick is though, the dock only outputs this power if it detects a CA device (Optical Interface, IF Box) or a VTR device (SRW1, OB-1, Codex). So you must have the ENC-105 IF box board attached to your CN-2908 dock port board, or there is no juice. But you can totally tap into it. And the audio board in the IF box can be left out.
 
Well

I have my custom mount made up which is a mount for the cinedeck that uses the guts of the IF box - so the internals of the If box (except the XLRs) are still there. So I just need someone to make me a cable and adapter and I can get an XLR power port wired up for my cinedeck?
 
JMtheDP said:
Well

I have my custom mount made up which is a mount for the cinedeck that uses the guts of the IF box - so the internals of the If box (except the XLRs) are still there. So I just need someone to make me a cable and adapter and I can get an XLR power port wired up for my cinedeck?
Click to expand...

Please understand that I'm essentially a hobbyist when it comes to camera modification, and I don't have an electrical engineering background, and that ultimately, I cannot bear any liability for how you choose to use this information... whatsoever.

There is a 12pin connector cable between the CN-2908 and the ENC105 board. Best as I can tell, the connection style is Molex Picoblade. Pins 11 and 12 are +12v unregulated, and pins 9 and 10 are unregulated ground. Here's the problem though - the Picoblade specification is only rated for 1amp of power per connector. Using only two connectors, the Cinedeck is going to draw over 5.3amps per connector.

I haven't gotten my hands on an SRW1 yet. But the documents I'm looking at indicate that it uses a different circuit board for the dock - CN-2429. I would think that board will have to have better output options. But I don't know if it's possible to drive that board with the ENC-105. Sony sells the CN-2429 for $265 or so, btw. But I need to get an SRW1 to test with before I'll know for sure it's swappable.

In short: if you're using the CN-2908, I would try to power anything exceeding 20watts for the time being.
 
Last edited:
Quick update...

I'm busy with paying clients, so I'm a bit slow on the actual work progess for this, but I did receive the I2C protocol analyzer. I also just purchased an SRW1 & SRPC1 that include the 102 and 103 boards. When I receive it I'll be able to do more testing. And I'll have ramping and Select FPS!

More later.
 
So...

I've done a successful test powering my recorder (Sony SR-R1) off the dock port of the camera. Please keep in mind, this was a workbench test, loose wires everywhere - not a field ready application - but proof of concept.

Between testing and reading documents I've learned some interesting things about the way the dock operates.

By default the dock has most of its pins disabled. This makes a certain amount of sense - if there's not an IF box or VTR attached, there's no good reason to send unregulated 12v power or SDI video signals - afterall the first is an electrical hazard (rain, etc) and the latter is probably a power saving measure.

Further though, the camera breaks connected components into two categories - VTRs, and Interface units (or CA units, perhaps - but this includes the IF Box and CA-F101). So called interface units are essentially video output devices - but without the feedback of a VTR. I suspect that when the CA-F101 is connected a fiber capable SRW-1 it may "turn into" a VTR as far as the camera is concerned, but I don't have the parts to test this theory.

Regardless - the camera uses I2C (or IIC) for communication with IF units, and a 5-pin protocol (TX-/TX+/RX-/RX+/GND) for communicating with VTR units. I haven't been able to "read" the VTR protocol communications yet, but my suspicion is that is a variation of the standard Sony 9-pin VTR control protocol typically seen on DB-9 connectors that is based on RS422. (If you're familiar with RS422, you know it uses TX-/TX+/RX-/RX+, as well as two common grounds and one frame ground - as any unit attached via the dock is on the same electrical ground as the rest of the camera, I suspect the common grounds became non-essential for this application). My reasoning is fairly basic - the wiring matches up correctly, and seeing as Sony had an existing VTR protocol it makes little engineering sense to build a new one from scratch when plenty of time could be saved by reusing and modifying the existing one.

Unfortunately, my initial tests of hooking the dock ports to a Sony VTR failed to produce any results, so I'm going to need to do more reverse engineering, and that will need to wait on parts and available time.

But! I shouldn't forget to mention, I did determine how the camera decides a VTR is docked, even if said VTR is powered off. Essentially, the dock has a pin that acts as a tripwire when an appropriate amount of resistance is added on the line. I think this "pulls low" a pin somewhere in the camera, telling it a VTR is connected. When the CN-2429 circuit board (dock board for the SRW-1, analogous to the CN-2908 board of the IF Box) is attached to the camera - even if no other circuit boards are attached - it supplies this resistor trip, telling the camera that a VTR is docked (viewable in a camera diagnosis menu). Key thing? The camera then switches the supply of 12v power to the dock port on. If you're looking to get power from the dock - this is the most compact solution - unlike the CN-2908 of the IF box it doesn't require a second circuit board, and also if you use the CN-2429 as is, it has a more robust dedicated 12v output connector built onto it.

Unfortunately, this action doesn't unlock the video signals. So it gives you power - but the camera requires the communication responses from the VTR before it unlocks the video signals. This means for a power only dock a very compact solution is possible, but your video signal will either need to come from the other dock port (with IF box) or from the camera monitor port. To get a VTR emulation dock with power and video will still require some work to develop the emulator (some sort of microprocessor).

Disclaimer: I'm a hobbyist as far as this electrical stuff goes. Anything you do to your camera is at your own risk, and I accept no liability whatsoever.

So, technical stuff:
*Dock pin #16 is used to signal that a VTR is connected. The CN-2429 has 6.8k ohms of resistance on this line. I tripped the dock successfully without the circuit board by wiring 6.75k ohms of resistance between pin #16 and the ground pin. Seemed quite reliable.
*With pin #16 tripped, 12v power is available from dock pins #8, #9, #10. They're small pins, so if your load is over 1amp, you'll want to use multiple pins and distribute the pull.
*If you can't figure out which pin is ground - it's the center one, between the six SDI ports. It's not one of the 17 small pins - It's the large center pin.

VTR communication pins are documented as follows (this is for the dock pin-out, not circuit boards):
#11 - TX(-)
#12 - TX(+)
#13 - RX(-)
#14 - RX(+)

Interface Unit (IF Box) Communication Pin-out (dock pin-out, not circuit board):
#5 - SDA (Data)
#7 - SCL (Clock)


*****************
And yes - using the above the information I did power my Sony SR-R1 recorder. Progress!
 
Quick update...

I've got wires everywhere on my workbench. Progress is slow, but I've been able to read from the SRW-1 to my laptop the data stream it sends to the F35. I don't have it fully decoded yet, and may still need to tinker with the connection settings, but I am able to find and read the timecode from the recorder in the data stream - so I know what I'm looking at isn't gibberish anymore.
 
macgregor said:
I have no idea what you're trying to do but it sounds really cool. :DDDD
Click to expand...

Imagine if your Hyperdeck Shuttle, Atomos Samurai, Pix240, or other recorder physically docked to the camera, received power through the dock, received it's SDI signal(s) through the dock, provided video playback through the viewfinder and camera monitor output ports, and made use of the camera Record button, tally lights, and timecode display...


Mind you - some recorders like the Hyperdeck Shuttle may not offer the necessary control signals for all of the above features, but at a minimum (in theory) I should be able to enable for all external recorders:
- Power from dock
- SDI from dock
- Ancillary record SDI flag on monitor port of camera (Not through dock)
- Use of tally lights on camera
- No need for IF box

For recorders with RS422 Sony protocol control, and possibly other feedback protocols, expected additional features:
- Videoplayback to camera through dock - feeding to viewfinder and monitor outputs
- Correct Timecode/Userbits sent to camera
- Record button on camera works through dock
- Pressing record on recorder provides feedback to the camera, which will show as recording.


Just imagine if switching between recorders literally involved docking and undocking them. No cables. Just input media, and press record on the camera.

Mind you, to do a fully internal conversion of a recorder will mean developing a physical dock interface that hides the power, SDI, and communications cables internally.

It's a large project, but in my opinion possibily the most important one to the camera's longevity. Making the camera easy and intuitive to use with new recorders, eliminating cable mess...

Yeah.


Oh, and I figured out how the data packets are encoded. Not sure what all the data in the packets is yet, but working on it! At this point, I think I could send dummy data from my laptop to convince the camera a VTR is attached, enabling the SDI outputs.
 
You must log in or register to reply here.
Back
Top