Thread: "big" ursa 4k magenta screen of death?

Page 2 of 2 FirstFirst 12
Results 11 to 20 of 20
  1. #11 Main board and sensor .. 
    Quote Originally Posted by Robert Hart View Post
    I had another visitation inside the gizzards of the beast and have found its beating heart. So far no signs of burnings of parts or obvious trauma. So it looks like I am going to hit a wall.

    It appears that this motherboard carries a memory battery. I just wonder if the battery goes dead flat over an extended period like the three plus months this thing slept, whether the embedded operating system can become corrupted by the next start-up and the appliance knows no longer its raison d'Ítre, whether it is a hospital IV drip controller, an electric toaster or an URSA camera's guts.

    It was rare but possible for the SI2K camera to lose its mind. It was running on a stripped down embedded Windows XPe and the hardware was up to critical medical equipment standards. If the memory battery went flat, it would do weird stuff to the recorded file order. For it, there was a recovery tool on a USB stick which reconstructed the operating system from scratch.

    If that memory battery on the URSA motherboard is done for, then maybe replacing it, then attempting to put the latest firmware in again might help it mend its ways. Then again if there was doubt it was a brick, there may remain no doubt afterwards.

    There is also a part near the centre of the board which looks like a little push-button momentary microswitch but it may just be some specialised component. If a switch, it may be a manual erase or reset switch to original defaults. I have not been game enough to go anywhere near it to see if that little white plastic centre actually moves.

    I would dearly like to get my greasy fingers on some service notes, which can warn me off such folly or tell me how to go about it.Attachment 28291
    Hello Robert , most of the time , " super Capacitors " look like lowprofile Electrochemical Capacitors .. If there is a Backup Battery on the Board , I would think maybe it's to keep the Internal Clock Running .. Same as on a Computer's Motherboard ,.. in the menu you can enter the Date and Time ..

    I have left BM Cameras unused for very long times , and never had a problem to Start them Again .. For me , the Firmware and Menu Settings are written onto EEflash Memory .. no need for a battery or " Super Capacitor " .. to maintain firmware and settings ..

    Volatile Memories is an old technology .. I have seen that long ago on some equipments .. fi the battery went dead , you had to send the equipment back to the Factory for resetting it .. Some third party companies provided replacement boards which did not need a battery , everything was written " Hard " and non volatile ..

    I have studied the CMV12000 Sensor Datasheet .. It might well be the sensor on the Ursa .. Resolution wise , and Global Shutter .. When you study the Datasheet , you Quickly understand this is not a " Small Component " ..... I would say this is " The Hell of a Component " .. not for Hobbyists .. I am in Admiration and I Congratulate Black Magic Engineers ... they know their job .... Fantastic Device , and Fantastic Job .. 237 Pins ... You have to Interface everything .. you have to Set the Registers .. and more ... Beautiful ..

    When you read PAGE 13 of the Datsheet " ABSOLUTE MAXIMUM RATINGS " , you see that Maximum Junction Temperature is 70 C ( which is not that high .. ) . And you read that Maximum STORAGE Temperature is 40 C... It means you should not forget your Camera inside your Car , under the Australian Sun ... Even here , we had up to 45 C. in the Shadow .. last July ( which I suffered from .. I prefer Rain .... ) .. which could be Harmful for the Sensor ..

    Continuous Power Dissipation 4200 Milliwatts .. It's very important to have a Radiator Behind , and a perfect Thermal Contact ..

    Also Relative Humidity MINIMUM 30 PerCent .. that means no storage in a Completely Dry Environment ... It begins to get Complicated ....

    There are Three Different Voltages Needed : 3.3 , 3.0 , and 1.98 Volts .. On PAGE 21 , we read that " VDD18 " and " VDD_PIX " draw large Peak Currents . Especially VDD18 Draws 1.7 Ampere every Readout Line .. They Mention : " The voltage regulator
    should be able to handle the 1.7 A. " .. It means there is some STRESS on That Regulator ... We have to Note that VDD18 is shared among several different Pins .. There are Three Different Voltages , but there are Five " Power Lines " ..

    1.7 Ampere is not a " Low " Current , so , is the Dedicated Voltage Regulator still Alive ? I don't know what is the Architecture of the Voltage Regulators System .. With very small Drop Voltage Regulators , my personal technique is to Cascade them .. 3.3 then 3.0 then 1.98 .. meaning the first 3.3 Regulator has to hold the Total Current .. It's always good to cascade .. if you input 12 Volts in a 2 Volts Regulator , it means a Ten Volts Drop .. under 2 Amperes , meaning Twenty Watts of Power Dissipation .. Way too Much ..

    I have read that this Sensor is Capable of a 15 Stops DR .. but sure not in Global Shutter Mode ..

    On Page 82 , we can Read : " Image sensors with color filter arrays (CFA) and micro lenses are especially sensitive to high
    temperatures. Prolonged heating at elevated temperatures may result in deterioration of the optical
    performance of the sensor. "

    So , what if the Lens is Directed to the Sun ? Very Dangerous ...

    All That confirms what I have Always Thought ... Our Cameras are Fantastic Equipments ... but they need much Care .. when using Them , and also when Storing them ..

    I have read on another Thread that some Filmmakers " Wake Up " their Cameras from Time to Time , because they are Frightened they could not start again .... Like Flowers ..
    you have to give them some Love and Care , or they Fade and Die ... it's Very POETIC ....

    I Think Reality is Less Poetic .. Nothing Magical .. Pure Technology .. if something does not work , there is a Reason ... and I hope you'll find the Reason ..

    Good Luck Robert , let us be informed ..

    With my Best Regards from France ..

    Mike ..
    Reply With Quote  
     

  2. #12  
    Senior Member
    Join Date
    Sep 2014
    Posts
    143
    Michel59.


    Thank you again for your further information. It begins to lead me from the motherboard to the sensor/turret. But then there is that totally white image being generated from the motherboard, whether the turret is connected or not.

    In-camera mic audio also comes into the motherboard from small mics in the bottom of the turret via the ribbon cable to the turret dock strip it seems. There is no audio input showing on the display waveform.

    We typically have high temperatures and periods of low humidity at certain times of the year. The camera has seen two 40C+ plus days in storage here. It has not been operated in hot conditions by myself.

    When I extracted the turret, there was no resistance to it coming out. There was also a broken-off piece of grey hard thermal paste in the fan at the rear which compares to other attached extrusions around the thermal pickup for the heat from the peltier heat exchanger. The thermal bond to the peltier heat exchanger on the back of the turret may have been disrupted by violence when the package was being shipped. This likely has compromised sensor cooling.

    The thermal bond between the two Kintex-7 processors and their heatsink was good. The motherboard was well and truly attached.

    I have a lot of homework to do searching the internet regarding the components.

    As for dynamic range, the real-world performance seems comparable to the SI2K at about 10.5 stops. The whites seem more forgiving but overcooking the whites seems to introduce that same magenta cast as the SI2K does when the green channel clips. If the sun gets in the shot, the sun is not black but magenta.

    One accompanying image is of what appears to be a memory battery retained by a spring clip. URSA MEMBATT 1.jpgThe ability to remove the battery has been defeated by the clip being pinwelded to the battery body or it may be variety of the battery with a tag attached to enable a solder point. The battery is delivering 0.7V.

    On the reverse face of the motherboard but not exactly in place behind the battery is this component shown in the second image. Two conductors may feed through the PCB to the battery from one end of the chip.URSA MEMBATT 2.jpg It appears to be connected in series with another component on the same side of the PCB as the battery and appears to be a diode. This PCB is an assembly of two double-sided boards bonded together
    Last edited by Robert Hart; 10-07-2019 at 11:48 AM. Reason: error
    Reply With Quote  
     

  3. #13 Components .. 
    Quote Originally Posted by Robert Hart View Post
    Michel59.


    Thank you again for your further information. It begins to lead me from the motherboard to the sensor/turret. But then there is that totally white image being generated from the motherboard, whether the turret is connected or not.

    In-camera mic audio also comes into the motherboard from small mics in the bottom of the turret via the ribbon cable to the turret dock strip it seems. There is no audio input showing on the display waveform.

    We typically have high temperatures and periods of low humidity at certain times of the year. The camera has seen two 40C+ plus days in storage here. It has not been operated in hot conditions by myself.

    When I extracted the turret, there was no resistance to it coming out. There was also a broken-off piece of grey hard thermal paste in the fan at the rear which compares to other attached extrusions around the thermal pickup for the heat from the peltier heat exchanger. The thermal bond to the peltier heat exchanger on the back of the turret may have been disrupted by violence when the package was being shipped. This likely has compromised sensor cooling.

    The thermal bond between the two Kintex-7 processors and their heatsink was good. The motherboard was well and truly attached.

    I have a lot of homework to do searching the internet regarding the components.

    As for dynamic range, the real-world performance seems comparable to the SI2K at about 10.5 stops. The whites seem more forgiving but overcooking the whites seems to introduce that same magenta cast as the SI2K does when the green channel clips. If the sun gets in the shot, the sun is not black but magenta.

    One accompanying image is of what appears to be a memory battery retained by a spring clip. URSA MEMBATT 1.jpgThe ability to remove the battery has been defeated by the clip being pinwelded to the battery body or it may be variety of the battery with a tag attached to enable a solder point. The battery is delivering 0.7V.

    On the reverse face of the motherboard but not exactly in place behind the battery is this component shown in the second image. Two conductors may feed through the PCB to the battery from one end of the chip.URSA MEMBATT 2.jpg It appears to be connected in series with another component on the same side of the PCB as the battery and appears to be a diode. This PCB is an assembly of two double-sided boards bonded together
    Hello Robert , I looked at your Pictures . Yes , it seems it is a Battery , and we can see two " Point-Weldings" on the upper part .. For the Component on the other side , we can Read " ESR6" which means " Equivalent Series Resistance 6 " .. So , it seems it is a Capacitor .. we know that , but not really useful I think ..

    You write that , when you remove the Turrett , you feel no Resistance at all , that seems strange .. I would think maybe there is some mechanical failure or problem on the Internal Part of the Connector .. .. ?

    We should have access to the Schematics and a Service Manual ..

    i Read on another Forum someone suggesting that you Loose your Time with Technics , instead of Making Films .. I do not totally agree .. I think that The More You Know the Better it is .. and even if what you have discovered is not useful this time , you have learnt things and that is a good thing for the future ..

    If you are Good at Making Films , and Also have knowledge about Technics , That's Perfect ..

    Anyway , Good Luck Robert , and let us know ..

    With my Best Regards from France ..

    Mike ..
    Reply With Quote  
     

  4. #14  
    Senior Member
    Join Date
    Sep 2014
    Posts
    143
    Thanks again Mike.

    When I first pulled the turret out, there was no resistance. I would have expected there to be a lot if that thermal compound was stuck on. So maybe there has been reduced thermal conductivity from the peltier output pad to the cooling pad behind.

    The way it is set up, the strip plug should bottom out before the peltier pad bottoms out. That will be why the application of thermal compound has been very generous. The material appears to be intended to set firm and not slump out of a clearance gap and break the thermal connection.

    With the motherboard and all components out of the body and the strip plug inserted fully home, the image remained the same which suggests that the strip plug and its socket are making connection. So it comes down to the turret and sensor being dead or the motherboard being faulty.

    My first move should be to change out that memory battery and see if that fixes anything.

    The next move should be to attempt to re-install the last firmware upgrade on the guess that maybe the base operating system is in non-evaporative memory but the actual imaging firmware is in evaporative memory and maintained by the memory battery.

    If the whole thing is resident in evaporative memory then I am sunk.

    Then again, something may have gone west on the motherboard.


    That little white button switch, is a soft power-on switch. I guess it is there for test purposes or bench calibration of the system before assembly into the camera body and connection of the controller touch screen and button assemblies. I gave way to temptation and pressed it to see what would happen and the camera started up.
    Last edited by Robert Hart; 10-09-2019 at 05:21 AM. Reason: errors
    Reply With Quote  
     

  5. #15  
    Senior Member
    Join Date
    Sep 2014
    Posts
    143
    I have carefully prised the spotwelded tag off the battery body. It turns out the "battery" is an ELNA 3.3v 22f supercapacitor. It would have been very helpful to have shop information but BM keeps its intellectual property very close. It would have been handy not to go down this particular dead-end. Getting it desoldered is going to be a bit of a mission. It now has to come out because the spotwelds have bulged the metal case of the supercapacitor and it likely is now ruined.
    Reply With Quote  
     

  6. #16  
    Senior Member Zolac's Avatar
    Join Date
    Jul 2013
    Location
    Canada
    Posts
    169
    I rooting for you to succeed in your efforts to get your URSA running again.

    While you're at it, seeing as you've torn it down very extensively. Can you ( if you haven't already ) document in some finer detail the entire/construction/assembly/components/pcb of the camera?

    Pictures of PCB in High res, with good lighting ( to show the IC marking, etc... ) The sensor block, etc.. cooling system.

    The pictures you've posted so far aren't the greatest quality, hard to read any detailed information on any of the PCB.

    You can add to the existing thread I have here: http://www.bmcuser.com/showthread.ph...ing-Blackmagic
    ( looks like you were already posting about your URSA there )

    I'm sure it would serve well to those interested in the topic, even if just for pure academic purposes.
    Reply With Quote  
     

  7. #17  
    Senior Member
    Join Date
    Sep 2014
    Posts
    143
    Zolac.


    Except for film stills camera, I don't have anything with better resolution than the picturephone. There is a DoP freind who has a 7D. I might get him to take some happy snaps of the post-mortem for that is what it now is. A tiny 1mm x 4mm conductive pad on the motherboard must have roasted off with the heat of desoldering the supercapacitor which was thought to be a coin battery. If BM was not so secretive, I would not have gone after a failed memory battery and instead left the supercapacitor where it was. The pad rose up a little and became detached. The little piece does not appear to have any connection to surface tracks on the PCB so I would have epxected it to be connected to other layers by feedthrough conductors. There are no signs of feedthroughs into the circuit boards below it so I have no idea how the current travels from the capacitor -ve to anywhere else. If there are feed-throughs in the PCB to other layers they must be spiderweb thin. How to install a replacement supercapacitor is going to be the challenge if at all possible. The system still switches on and the image displays are the same as before. A chill spray over the board and components and dry-off changes the behaviour of the faint pixellation pattern which becomes very sightly more apparent. Like other tests, the pattern goes away on subsequent startups, maybe due to temperatures reaching an operating range. The device cannot be operated for more than a few seconds because I have no heatsinking on the Xilinx processors and they will be damaged if I power up for longer.
    Reply With Quote  
     

  8. #18  
    Senior Member Zolac's Avatar
    Join Date
    Jul 2013
    Location
    Canada
    Posts
    169
    For what it's worth, I did also have an old BMPCC go bad, to the point where the sensor output was similar to what I think you're experiencing.

    I had a bad power button on the BMPCC ( among other things like a broken HDMI port ) which meant I had to disassemble the camera and make repairs a few times.
    In the process of doing that, I believe I imposed too much stress/bend on the sensor flex cable that is bonded between the sensor board and the main PCB.

    Upon re-assembling it the last time I did this was the image output I was greeted with...

    IMG_1428.jpg

    The explanation I got from BMD was that this is caused ( most-likely ) by a fault/damage to the sensor connection.
    Which totally lines up with what was happening over the numerous times I was in and out of the camera.

    So I'm not sure your diagnosis of the super-capacitor as the culprit of your problem is necessarily the right one.
    Obviously lots could be at play, but I'm going to draw some similarities between these two and say It's an issue with the sensor connection ( possible damage to it. )
    and/or any sub-systems associated with it.
    Reply With Quote  
     

  9. #19  
    Senior Member
    Join Date
    Sep 2014
    Posts
    143
    Yes. That is near-identical to the screens on the big URSA. - Dead and buried then at a guess. The ribbon cables are sandwiched between the multilayered PCBs on the turret and the motherboard. There are no obvious test points one can use to check continuity across the in-line slot plug. One would have to be really careful when checking at other points on the boards. Without info from BM that cannot happen systematically. Time for crying in one's cups or dropping down on the ground in a rage and kicking up dust.
    Reply With Quote  
     

  10. #20  
    Senior Member
    Join Date
    Sep 2014
    Posts
    143
    Further to above, the phantom power and external mic inputs work and the travelling waveform on the controller screen displays correctly.
    Reply With Quote  
     

Similar Threads

  1. Replies: 4
    Last Post: 05-30-2015, 07:29 PM
  2. how to mark the thread "SOLD" or "BOUGHT"
    By Bubble in forum Off Topic
    Replies: 0
    Last Post: 11-27-2013, 11:35 PM
  3. SD Card "Not Ready" when Pocket Cinema Camera is set to record "RAW".
    By StarCoreFilms in forum General Discussion
    Replies: 9
    Last Post: 11-12-2013, 04:12 PM
  4. Pocket Camera "blooming" and "Blackhole" Thread
    By sfpuk in forum General Discussion
    Replies: 158
    Last Post: 09-05-2013, 07:13 AM
  5. "smart" mft lenses working on "dumb" mft mount?
    By SuperWhiteFlame in forum Lens Discussion
    Replies: 7
    Last Post: 08-14-2013, 02:33 PM
Bookmarks
Bookmarks
Posting Permissions
  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •