#apertus IRC Channel Logs

taking the right side lid off right now

then we can speculate which part is what and what it's used for

think it can be fixed in post?

the calibration data will be on a flash on the sensor board - which is through that flex.. so low chances of fixing it

what I know is that it's static

the curve doesn't change

the 2k5 bmcc has a 256mbit flash, for the 5MP sensor (which is used even in crop mode at ~ 3.5MP)

they'll probably be in the same place

I'd really rather not take it apart further until I'm certain it can't be fixed in post

reason is, that the data is already cooked from wrong components

so these merge artifacts are in the area wher the lg switches to hg

off to bed now ... have a good one everyone!

danieel, ripping the camera apart... gently

danieel, yup

it's a spansion flash that appears to have the teenage problem of 0201 caps growing around it

https://i.spirit.re/KCWD0.jpg

well at least I think it is

64MB

it might be hackable to be honest

then buy another one, flash it in

.5mm pitch it looks like

integrity may be compromised if the package body is exposed to temperatures above 150°C for prolonged periods of time.

danieel, I don't plan on preheating it

also I just realized that you're Daniel Rozsnyo, and we know each other

Does the board have buried vias or are there visible vias beneath the BGA?

I won't be removing it

https://i.spirit.re/LkF0N.jpg

I shouldn't overheat it

well, cooling down the board is nice but doesn't help much

Bertl, it's a thermal camera

you need to heat up the BGA balls and the pads

the best way to do that for both reseating as removing is to use a thin copper foil directly on the part

I'm heating it up until the solder just melts

attach with some flux and heat directly with the solder iron

so I can't do ultrasonic

so no flux

-"need to" :)

:P

any flux which is labeled 'no clean' doesn't need to be removed

it can be rinsed off

of course you can remove excessive amounts with a special q-tip

so what should I set it to for unleaded solder in this case

heat gun or solder iron?

you very likely need above 300 for that, better 320-350

yup

okay, I got it to float

https://i.spirit.re/6UZr4.png

this thing soaks up heat like a magnet

sensor glass did not delaminate

bubble*

nope, I've decided to assemble the whole thing

but first, to photograph the whole thing, for future reference and a stupid article

well, the multilayer boards are lot of copper, especially with the shielding planes

yeah this thing looks like at least 8 layers

off for now .... bbl

:P

the fpga is a z7030 / z7045 ?

will take a look in a bit once I get to photographing the pcb

I'm not very keen on taking the heatsink off

not the heatsink.. just the paste

where does the paste reside?

https://i.spirit.re/9fae3.jpg

the spansion chip has spots now that I look at it

is anyone here registered on bmcuser for a few months?

it's ridiculously tight even without screws

do you wanna share it in the teardowns section of the apertus wiki? https://wiki.apertus.org/index.php/Teardowns

absolutely

at this point I'd love to own an axiom beta

buuut it doesn't do internal recording(yet?)

no internal recording on the beta yet correct

any major improvements over the last year-ish?

I guess it can be a very expensive machine vision camera :P

I have a cmv2k usb3 cam for that

that's 4k30

the hardware has dual 6g sdi outputs and single 6g sdi input

https://github.com/felixheld/AXIOM-photonSDI-hw

it'd be nice to have a single 12G-SDI output and a 3G in

ah

or is that usually just done with an fpga

the ultrazed-eg or ev are very nice

but with the new mpsoc

I don't think they're pin-compatible

it's a lot more expensive though

$440 in 500

but since there's two 6G gtps, it would be interesting to see dual SATA3 instead of dual 6G-SDI

the same as it does on the ursas

when the 6g sdi module is finished and the gateware developed, i'm planning to make a sata board. that's mostly removing stuff from the board and add the sata connectors

:P

for the sata gateware, litesata can probably be extended without too much hassle

reflowing the chip did nothing

this sucks.

nothing unexpected :)

Spirit532: I'd suggest to make high-res scans as well

not as usefull as 2kdpi scans

I don't have a scanner

money, mostly