Current Server Time: 21:40 (Central Europe)

#apertus IRC Channel Logs

2014/12/09

Timezone: UTC


23:36
g3gg0
left the channel
00:26
fsteinel_
joined the channel
00:29
fsteinel
left the channel
00:31
fsteinel_
changed nick to: fsteinel
01:53
mooseboobs
left the channel
01:53
mooseboobs
joined the channel
03:29
mooseboobs
left the channel
04:39
Bertl
off to bed now ... have a good one everyone!
04:39
Bertl
changed nick to: Bertl_zZ
06:11
intracube
left the channel
06:26
g3gg0
joined the channel
08:04
surami
joined the channel
08:16
surami
morning
09:17
se6astian|away
changed nick to: se6astian
09:23
se6astian
good morning
09:38
sb0
joined the channel
09:42
lab-bot
sebastian closed T160: invite all crowd funding backers to the lab as "Resolved". http://lab.apertus.org/T160
09:46
surami
left the channel
09:50
surami
joined the channel
09:50
surami
left the channel
09:53
surami
joined the channel
09:54
surami
left the channel
09:56
surami
joined the channel
09:57
surami
left the channel
09:58
surami
joined the channel
10:00
surami
left the channel
10:08
surami
joined the channel
10:09
surami
#help
10:10
se6astian
grab this rope!
10:10
se6astian
throws safety buoy
10:16
se6astian
changed nick to: se6astian|away
10:22
intracube
joined the channel
10:57
surami
left the channel
10:57
surami
joined the channel
11:01
intracube
control concept v3: https://lh4.googleusercontent.com/-esn_bhh1MXw/VIWbLKA8VkI/AAAAAAAAA-o/nxk0HUotv94/w750-h562-no/apertus_control_v3e_processed4.jpg
11:07
surami
left the channel
11:08
surami
joined the channel
11:14
surami
left the channel
11:14
surami
joined the channel
11:14
surami
left the channel
11:14
surami
joined the channel
11:48
surami
left the channel
11:48
surami
joined the channel
11:58
se6astian|away
changed nick to: se6astian
11:58
se6astian
back
11:59
se6astian
the first perks have been shipped!
12:10
surami
left the channel
12:10
surami
joined the channel
12:42
mooseboobs
joined the channel
12:49
Bertl_zZ
changed nick to: Bertl
12:49
Bertl
morning folks!
12:59
se6astian
good morning
13:01
Konstantin
joined the channel
13:02
Konstantin
Hello everyone
13:02
se6astian
Adobe sent a draft for extending the cinemadng standard with some new tags for camera manufacturers
13:02
se6astian
Bertl: you wanna take a look and provide feedback?
13:03
se6astian
I see for example that they offer tags like cameraPosition v3f (3 floats)
13:03
se6astian
for storing IMU data
13:04
se6astian
I am not familiar with the standard though so I dont know what changed compared to the last approved standard
13:04
se6astian
hi Konstantin
13:05
se6astian
but I am sure we can research the changes if we consider it worth the effort
13:17
Bertl
Hey Konstantin!
13:18
Bertl
se6astian: I'm not familiar with the standard either, you might want to get Gabe involved
13:18
se6astian
good idea
13:18
se6astian
I know we wont write raw cinema dng data that soon
13:19
Bertl
yes, unlikely with a well supported MLV format nearby
13:19
se6astian
but considering its the only widely supported AND open raw standard and they ask us to participate in it we should consider it carefully :)
13:20
se6astian
MLV is another option indeed, but has much less industry support so far
13:20
Bertl
Konstantin: so, IIRC, you want to help with electronic design?
13:20
Konstantin
oh yes i hope i can Bertl
13:21
Bertl
well, sounds good to me, are you already familiar with the current designs?
13:23
Konstantin
i've looked into some of the parts
13:24
Bertl
let's hear what you figured out on your own so far and please do not hesitate to comment or critizise
13:26
Konstantin
So far i only saw some of the digital circuits im not quit sure that i found all what i should
13:27
Konstantin
I have to say im a heavy analog guy not that i know nothing about digital but im not so into programming unfortunately
13:28
Konstantin
Some of the layouts u have done are quite good, i know from my self a layout is never finished so u will always find something, miner details
13:29
Bertl
yes, indeed, that's why I'm always interested in feedback
13:29
Bertl
I remember that you mentioned power management in your email, is that something you would like to work on?
13:31
Konstantin
yup Power management would be something and i thing that is portably easier to integrate
13:31
Konstantin
if u need any help in that area
13:31
Konstantin
im open to every thing ;)
13:32
Bertl
well, we have a very ambitious plan there, so it's probably an interesting challenge
13:33
Konstantin
i think too
13:33
Bertl
first, we want to accomodate a wide range of supply voltages, basically 5-40V, but not necessarily with the same board
13:34
Bertl
we've done some successful test with a switching regulator to cover the 6-35V range so far
13:34
Bertl
but we also plan to have flexible supplies for the sensors
13:34
Konstantin
ok but the rage is not that difficult
13:35
Bertl
good to hear :)
13:36
Konstantin
depends on the specs, of curse
13:36
Bertl
the sensor board is designed to have up to 6 voltages available, which can be limited in range by the sensor board, but can be regulated by the Beta power supply (via processor control)
13:37
Bertl
the interface board (which goes between sensor and Beta board) has another two voltages
13:37
Konstantin
so we have 6 different voltages ones plus the 2
13:37
Bertl
at least two of the sensor voltages need to be capable of up to 1.5A
13:37
Bertl
and the interface board supplies need to be able to do similar
13:38
Konstantin
probablythe biggest issue would proabably be the space for the differnt supplies
13:39
Bertl
yes, space is definitely an issue
13:39
Konstantin
witch powerlevels are involved like over all for the sensor Supply etc
13:40
Bertl
we can assume that the voltages lie between 1.2 and 4.0 V probably more 1.25 and 3.8V
13:40
Bertl
the interface board will most likely require 1.8 to 3.3V
13:40
Konstantin
ok that what i thougt for the core voltages for the sensor
13:41
Konstantin
is there currently a spreadsheet witch the "needed" supplies and there raitings?
13:41
Bertl
each sensor brings their own list
13:41
Bertl
if you check out the CMV12000 datasheet, you will see a list for this particular sensor
13:41
Konstantin
probablyproabably the supply for the sensor should be in some places low nois and stuff
13:42
Bertl
yes, the sensor supply must be low noise (i.e. well conditioned and filtered)
13:42
Konstantin
jup clearly
13:44
Konstantin
ok but the good thing is that the supply powere is not that big so that makes it easy
13:45
Bertl
to some degree, the tricky part is to keep the voltage variable, while limiting it to the "allowed" ranges for each sensor
13:46
Konstantin
ok the limits have to be set by the software and the rest is up to the control loop of the supply
13:47
Bertl
the current plan is to enforce the limits via hardware and from the sensor side, to prevent accidents
13:48
Bertl
so basically the sensor tells the power supply (with two signals) that a provided voltage is over or under the desired range
13:48
Bertl
s/sensor/sensor board/
13:48
Bertl
and if that is the case, no power will be supplied to the sensor
13:49
Konstantin
that is one way of dooing it how is the voltage detectet by the sensor itself ore the circuitry on the board?
13:50
Bertl
the circuitry on the sensor board
13:50
Bertl
sec
13:50
Bertl
http://vserver.13thfloor.at/Stuff/AXIOM/BETA/axiom_beta_sensor_cmv12000_v0.12.pdf
13:51
Bertl
argl, sec
13:51
Konstantin
ok then we have to be carefull not to disturb the control loop
13:52
Konstantin
i have the board plans i will have a short look
13:52
Bertl
please reload, it should now contain all pages
13:53
Bertl
on page 3, you'll see the mechanism which checks the voltages
13:54
Konstantin
ok u will use different refs
13:54
Bertl
it is basically a voltage reference, plus six comparators, which compare the provided voltage "sample" with the configured (via resitor ladder) range
13:55
Bertl
and then generate over/under voltage signals accordingly
13:55
Bertl
the eeprom on page 2 provides information about those ranges as well as a default value for each voltage
13:56
Konstantin
ok over that u can get the conditions
13:58
Konstantin
ok but that still means the voltage has to be comanded
13:59
Bertl
yes, the process is like this:
13:59
Bertl
the sensor is connected and the beta powered up
13:59
Bertl
the beta board checks for the eeprom and reads the identification as well as the ranges and default voltages
14:00
Bertl
it then decides to go for the default voltages (for example) and thus configures the appropriate voltages on the power subsystem
14:00
designbybeck_
joined the channel
14:00
Konstantin
i figured
14:00
Bertl
the sensor board immediately checks the voltages and gives the okay feedback via over/under voltage signals
14:01
Bertl
this is the point where the power sequencing can start
14:01
Bertl
i.e. the beta board can now enable the actual voltages to the sensor
14:01
Konstantin
and if it is out of grage u can powere of ore regulate
14:01
Konstantin
jup ok
14:02
Bertl
if one of the over/under voltage lines triggers, the power subsystem should cut off all power to the sensor
14:02
designbybeck_
left the channel
14:02
Bertl
until manual intervention from the beta board
14:02
Bertl
(i.e. new phase of power sequencing)
14:02
Konstantin
of curse
14:03
Bertl
within the ranges, the power subsystem should do the best to regulate the voltage
14:04
Konstantin
atleast if the voltages are out of the window u can detect it fast and react, if the stettings are not correct and a sensor is not that cheap :)
14:04
Bertl
correct, that's the idea behind it
14:05
Konstantin
i would do it the same way
14:05
Bertl
I had a simpler concept first, which basically used power mosfets to short-circuit in overvoltage conditions (which would be harmless because of the current limited supply) but I decided to choose a more complex but also more flexible design instead
14:06
Bertl
sensors are usually in the 400-1600 EUR range, so they are worth a little extra efford to protect them
14:07
Konstantin
i know both solution if ur power supply is well enough design u can switch fast enough
14:07
Bertl
so does that sound like something you would want to work on?
14:07
Konstantin
the complexer one is often used in expensive gear
14:08
Konstantin
definitely!
14:08
Bertl
great! then let's get started! :)
14:08
designbybeck_
joined the channel
14:08
Konstantin
how many people are actually working on the beta itselve
14:09
Bertl
very much depends on what you consider "working on"
14:10
Konstantin
maybe more precise on the hardware?
14:10
Bertl
as you probably saw from the schematics and board layouts, most of the hardware design is done by myself, with some help and contributions from mars_ and Gabe
14:10
Konstantin
is probably easier to answer
14:10
Konstantin
ok good to know
14:11
Bertl
so for coordination, checking with me is probably more then enough
14:11
Konstantin
^^
14:11
Konstantin
one good thing is that u are using eagle too
14:11
Bertl
if you have questions regarding the existing designs, or if you want to bounce off a few ideas, feel free to do so any time
14:12
Bertl
yeah, Eagle is more a left over here, we still plan to switch to KiCad at some point, but I've been using Eagle for a long time now and thus I'm very comfortable with it
14:13
Konstantin
i used eagle for the last 6 yeras
14:13
Konstantin
is there any like FTP were i can find projects Documents and stuff?
14:14
Bertl
documentation is mostly on the wiki, but you won't find much documentation about the hardware (yet), as this is still a to-do
14:15
Bertl
you can find the designs either on my server (http://vserver.13thfloor.at/Stuff/AXIOM/BETA/) or on github (once I checked them in :)
14:15
Bertl
if you have questions, do not hesitate to ask
14:15
Konstantin
ok maybe it would be a idear to work out a hole powere plan makes designing much easier
14:16
Bertl
sounds good to me, I'll help where I can
14:19
Konstantin
it is probably better if we talk about these things. IRC could be a bit involved,
14:20
Bertl
IRC is fine, and we have a log if we want to revisit something
14:20
Konstantin
ok
14:21
Bertl
and in the occasion that somebody wants to contribute anything, she/he is free to do so
14:23
Bertl
maybe somebody even starts transcribing the ideas to the wiki at some point, which would save us spending more time on documentation
14:23
Konstantin
i will try to look into the supply an work out a sheet where we can list the specs so we can use it an probably a concept too
14:24
Bertl
great!
14:24
Konstantin
i will scream next time in IRC wen i have news
14:25
Konstantin
se ya all
14:25
Bertl
cya
14:29
Konstantin
left the channel
14:31
se6astian
I need someone to review an english article draft again, anyone around (native speakers?) ?
14:32
surami
bye all
14:32
surami
left the channel
15:27
davidak
joined the channel
15:52
lab-bot
derkiki created T213: A entry level low price version of the Axiom. http://lab.apertus.org/T213
17:38
derWalter
joined the channel
17:59
derWalter
left the channel
18:11
danieel
se6astian: are you subscribed to some DNG mailing list or did adobe approach you directly?
18:11
danieel
i have also some extra flags to add (rolling shutter direction and timing)
18:59
se6astian
mailing list
19:00
se6astian
I can add you to the list
19:00
se6astian
which email address should I subscribe?
19:03
lab-bot
rgerst created T215: LED light (1 or 2) either built into front of camera or on small extender arms. http://lab.apertus.org/T215
20:08
lab-bot
christophvarga created T216: phabricator, communication, web, basic understanding of this interface. http://lab.apertus.org/T216
21:24
se6astian
time for bed
21:24
Bertl
good night!
21:25
se6astian
changed nick to: se6astian|away
21:40
designbybeck_
left the channel
21:41
davidak
left the channel
22:24
mooseboobs
left the channel
22:28
mooseboobs
joined the channel