Current Server Time: 07:08 (Central Europe)

#apertus IRC Channel Logs

2013/12/03

Timezone: UTC


00:18
Wescotte
joined the channel
00:45
intracube
joined the channel
00:52
dmj_nova
left the channel
02:00
Sasha_C
joined the channel
02:41
theverant
left the channel
04:24
jucar
left the channel
04:35
intracube
changed nick to: intracube_afk
06:08
dmj_nova
joined the channel
07:16
dmj_nova
left the channel
07:21
dmj_nova
joined the channel
07:52
Bertl
phew, just returned from the fab lab ... almost another catastrophy ... but the parts for the case should be done now, unless I got something wrong with the design ...
08:10
se6astian
joined the channel
08:10
se6astian
good morning
08:13
Bertl
morning!
08:13
se6astian
Bertl, good to hear some progress, was the lasercutter again not calibrated?
08:14
Bertl
well that is the 'expected state' but this time the dead man's handle got crazy
08:14
Bertl
(i.e. the watchdog circuit)
08:15
Bertl
it looked fine when you started, i.e. like 3minutes till the first key request, but after the first confirmation, the delay dropped to about 20 seconds, then after that about 5 seconds and then it simply turned off the power :/
08:17
Bertl
after playing that three times, I sent an useless and till now unanswered email to the only problem contact they have
08:17
se6astian
oh my
08:17
Bertl
fifteen minutes later, I disabled the watchdog :)
08:17
se6astian
haha
08:19
Bertl
an hour or two later, I found myself in a solid cloud of sawdust
08:19
se6astian
you have cut wood as well?
08:19
Bertl
somebody had turned on the - as we now know - broken vacuum cleaner, which didn't have a filter anymore
08:20
Bertl
so instead of sucking the dust in, it blew it out in the entire lab
08:20
se6astian
ah before I forget, please try if the nikon ring really fits the plastic and the nikon lens, if it does I will acquire the second identical camera from the same seller for Gabe
08:20
se6astian
lol
08:25
Bertl
well, it doesn't work with the aluminum lens mount, but only because the screws are different (i.e. the screw thread, not the position)
08:25
Bertl
as we do not have a thread in the other POM lens mounts (yet) it should be fine
08:26
Bertl
shall I test the bayonet part as well? or can we assume that this part works
08:27
se6astian
please test it through once entirely
08:27
Bertl
says the person who told me that the lens systems work on all cameras :)
08:29
Bertl
yep, bayonet mount works as expected, so I'd say, we can make it work for the POM versions
08:30
Bertl
would be interesting to ask konstantin if he used the original screws or maybe replaced them with more common ones ...
08:37
PhilippeJ
joined the channel
08:38
Bertl
morning PhilippeJ!
08:40
se6astian
great thanks, will acquire the second camera with the nikon ring then
08:40
se6astian
hi PhilippeJ
08:41
PhilippeJ
hello!
08:42
PhilippeJ
se6astian much clearer with the video
08:42
PhilippeJ
altough, I'm wondering if providing some sticks to remove the boards would not be enough
08:42
PhilippeJ
and thus allow a simplier casing
08:43
PhilippeJ
pc-like :-)
08:44
PhilippeJ
also, I'm sending you a link on a doc I'm curently working on :-)
08:45
Bertl
sticks to remove the boards ... do I want to know? :)
08:46
PhilippeJ
Well, the sodimm locking mechanism cannot be accessed if it's at the bottom of the casing, wo I though some "sticks" to push them from the top might help, when you cant to remove a board. Think (a bit) like the car radio metal key antitheft system
08:46
se6astian
sticks?
08:47
PhilippeJ
long piece of nylon
08:47
PhilippeJ
or plastic
08:47
se6astian
hmm, ok I see, well requiring special tools is prone to problems
08:47
se6astian
I would rather suggest we remove the locking parts from the sodimm
08:48
PhilippeJ
it could be done with a screw driver or a piece of wood, but we don't want people to ruin the connector board
08:48
se6astian
Bertl, I created this concept animation
08:48
se6astian
https://www.youtube.com/watch?v=Tav9aVykqgA
08:48
Bertl
have we already found a proper and cheap so-dimm socket?
08:48
Bertl
btw, removing the locking mechanisms is not the best idea :)
08:48
se6astian
I thought we did
08:49
PhilippeJ
digikeys has a lot of them
08:49
PhilippeJ
I don't know if they are "proper"
08:50
Bertl
the key word is vertical, the right angle ones are the bad ones
08:51
Bertl
that is why I original suggested to use the mini-dimm sockets
08:54
Bertl
20% more pins and they usually come in vertical by default
08:55
Bertl
(while the sodimm is usually built at near right angle)
09:02
se6astian
mini dimm is 82mm wide for the module so we would need to increase the size of the enclosure (not a big deal)
09:02
Sasha_C
left the channel
09:33
Bertl
off to bed now ... have a good one everyone!
09:37
se6astian
good night
10:00
Sasha_C
joined the channel
10:48
PhilippeJ
left the channel
11:18
PhilippeJ
joined the channel
11:52
tyrone_
joined the channel
11:54
tyrone_
hi all
11:56
tyrone_
i have seen on the archive from yesterday that se6bastian have made a module concept... have you look at the "vita 74" standard? it's nearly the same concept.
11:57
tyrone_
more information about "vita 74": http://www.vita.com/home/NewsandEvents/Events/Embedded%20Tech%20Trends/2012/Presentations/VITA%2074.pdf
12:02
se6astian
thanks, taking a look now
12:04
se6astian
very interesting!
12:09
tyrone_
i was searching for a solution for a modul concept that supports the vita 57 standard as a add on. that most of the vita 57 module will be supported to reduce the hardware development. but at the moment i couldn't find the right solution.
12:13
se6astian
very nice
12:14
se6astian
the question remains though if SODIMM sockets would not be cheaper though also less robust
12:14
se6astian
VITA57 is around 25$ per piece
12:24
mars_
robust in terms of what? many plug-in cycles or shaking?
12:31
tyrone_
price is one thing, getting it to work (production proven) and developing the hardware the other... backplane, cases... are on the market...
12:36
se6astian
mars_, yes plug-in cycles and mechanical durability of the contacts, etc.
12:37
tyrone_
the "vita 74" was just a information.... i know the connectors a costly... :-/
12:39
mars_
se6astian: your typical so-dimm socket has about 25 cycles
12:41
tyrone_
how many connectors are the target per modul?
12:44
PhilippeJ
two connectors I'd say
12:44
PhilippeJ
btw, very interesting stuff!
12:44
se6astian
mars_, 25 cycles phew thats not a lot
12:45
mars_
it was never meant to be cycled a lot
12:45
se6astian
tyrone_, SODIMM would have 204 IO lines in the DDR3 variant
12:45
se6astian
thats a number that is pretty suitable for our purpose
12:46
se6astian
the image sensor has around 145 IO lines for example
12:46
tyrone_
thanks.... that i wanted :-)
12:46
mars_
pci-e has 50 cycles...
12:46
mars_
for standard connectors
12:49
se6astian
interesting
12:49
mars_
external connectors are a different story, some connectors reach up to 1500 mating cycles
12:53
tyrone_
i was just looking at industrial solutions micro tca is at 200 mating cycles.. price: <20$
12:54
PhilippeJ
Vita 74 looks lovely (http://www.vita.com/home/NewsandEvents/Events/Embedded%20Tech%20Trends/2013/PDF_Presentations/M11%20-%20VITA%2074%20Small%20Form%20Factor%20Systems.pdf )
13:00
PhilippeJ
we should not be too cheap on connectors, if they will be attached to expensive boards
13:01
mars_
true that
13:01
se6astian
the Samtec SEAF-RA connector as suggested by vita74 is $52
13:01
PhilippeJ
not cheap...
13:02
PhilippeJ
but on the whole product prive, if it's more reliable
13:02
se6astian
plus the mating part on the other side again 52$
13:03
tyrone_
okey the x-mas gifts are canceled ;-)
13:03
se6astian
3 modules = 300$
13:03
se6astian
still cheaper than 4 connectors per module in a daisy chain sytem which would require 12 connectors = 600$
13:03
se6astian
for 3 modules
13:04
mars_
and at that (camera) price point you don't wan't failing connectors
13:04
se6astian
agreed
13:04
mars_
-'
13:04
PhilippeJ
I like the idea that the board itself can be plugged in a connector directly, like sodimm and pci, but is there an equivalent with high pin count, and not too expensive connectors ?
13:07
se6astian
searay connector: http://www.samtec.com/technical-specifications/default.aspx?SeriesMaster=SEAF-RA
13:07
se6astian
rated for 100 mate cycles
13:09
PhilippeJ
this as I understand it, is 170 pins : http://www.molex.com/molex/products/datasheet.jsp?part=active/0755940000_EDGE_CARD_CONNECTO.xml
13:13
tyrone_
excactly that molex connector i have looked is <20$ :-)
13:14
se6astian
200 mating cycles, not bad
13:15
mars_
ok, this is pcb to connector, not connector to connector
13:15
PhilippeJ
pcb to connector is nice, because the board are then -as I understand it- cheaper and easier to make
13:16
se6astian
tyrone_, where did you find the matching price?
13:16
mars_
problem is, your pcb is unprotected
13:17
tyrone_
http://www.digikey.com/product-detail/en/0755940000/WM10252-ND/1956169
13:17
se6astian
thanks
13:17
mars_
dirt, swear, dust and everyone can touch your pins
13:17
se6astian
the molex one is also just 68mm wide so would very well fit into the 8x8cm enclosure
13:18
mars_
sweat even
13:18
mars_
i'm not sure how reliable that is
13:20
PhilippeJ
mars_, those are meant to be inside an enclosure
13:23
mars_
and are handled by non-technicians in the field, if they want to reconfigure something
13:24
se6astian
its not meant to be reconfigureable in the field
13:24
se6astian
see latest concept doc philippe just sent to the mailing list
13:24
se6astian
the idea goes more in the direction of building a custom pc with PCI-e slots
13:25
se6astian
and if you want to upgrade/change your configuration you add or swap a PCI card
13:25
se6astian
but not on a daily basis
13:26
Wescotte
left the channel
14:27
tyrone_
left the channel
14:29
tyrone_
joined the channel
14:39
PhilippeJ
left the channel
14:40
Sasha_C
left the channel
14:53
tyrone_
left the channel
15:07
PhilippeJ
joined the channel
15:40
PhilippeJ
left the channel
15:50
PhilippeJ
joined the channel
15:54
guesst
i missed the connector discussion... what is the price point of the system?
15:56
se6astian
hi guesst, which system?
16:01
PhilippeJ
left the channel
16:08
guesst
where the connectors will go
16:09
se6astian
the pricepoint of the system where the connectors will go, I am afraid I dont understand your question
16:09
guesst
as there are some hidden info / i am not part of internal discussion / the concept show on apertus website does not need to use those fmc photoshopped connectors
16:09
guesst
assembly of that kind can use solderless connectors from samtec
16:20
PhilippeJ
joined the channel
16:21
PhilippeJ
guesst do you have a link to those?
16:26
PhilippeJ
so, gotta go, see you later everyone !
16:27
PhilippeJ
left the channel
16:27
guesst
https://www.samtec.com/technical-specifications/default.aspx?SeriesMaster=FSI
16:27
guesst
not that sorry
16:32
[1]se6astian
joined the channel
16:35
se6astian
left the channel
16:35
[1]se6astian
changed nick to: se6astian
17:28
VBenjFR
joined the channel
17:29
VBenjFR
Good evening!
17:32
se6astian
hey there
17:32
se6astian
did the import of the assembly I sent you go well?
17:35
VBenjFR
Yes, thaks o lot!
17:36
VBenjFR
thanks
17:36
VBenjFR
It was quite a big file,
17:36
VBenjFR
there is a lot information in that
17:39
se6astian
indeed :)
17:39
se6astian
any feedback about the assembly so far?
17:39
se6astian
what would you do differently
17:39
se6astian
and do you fully understand what each part is for and why I created it that way?
17:40
se6astian
please take a critical look at everything I created
17:40
se6astian
as I am not a mechanical engineer, I tried to do everything in a way that its most useful and can be manufactured but my actual CAD design experience is limited
17:40
VBenjFR
I did not look so deeply for the moment
17:41
VBenjFR
ok
17:41
se6astian
ok
17:41
se6astian
take your time
18:19
Bertl
morning everyone!
18:19
VBenjFR
Se6astian, how did you produce each part for the prototype?
18:20
se6astian
hi Bertl
18:20
se6astian
how do you mean "produce" ?
18:22
VBenjFR
Did you buy the different part in shop, or did you produced them in specialised industry or in labs?
18:22
VBenjFR
Hi Bertl
18:22
se6astian
the CAD files I sent you have not been manufactured yet
18:23
se6astian
or do you mean the parts of the Alpha prototype (not the files I sent you) ?
18:23
VBenjFR
ok, but you said the second prototype was ready few days ago
18:23
se6astian
ah, yes
18:23
VBenjFR
yes, the part of alpha proto
18:23
se6astian
the lens mount we ordered as CNC cut tool in germany
18:24
se6astian
the nikon bayonet ring we took out of old nikon slr cameras
18:24
se6astian
the PCB is ordered in the US and Bertl assemled them manually
18:25
Bertl
the rest is lasercutted PMMA and a bunch of screws and nuts
18:26
VBenjFR
OK, that was my question, as I did not have any dimensional tolerence on part
18:26
VBenjFR
maybe because the exchange format we used is not for that purpose
18:27
se6astian
there are non set so far
18:27
VBenjFR
Or partially et, as you know the dimension of your lens
18:28
VBenjFR
and used some commercial PCB
18:28
VBenjFR
(which is great as it is standardized)
18:30
VBenjFR
do we know the amount of camera we plan to produce?
18:30
VBenjFR
(approximatively)
18:31
VBenjFR
10, 100 ,1000?
18:31
se6astian
min 100, maybe more
18:31
VBenjFR
If that 10, so we will maybe not be able to continu to produce the parts by lasercutting?
18:32
VBenjFR
so we will maybe not be able to continu to produce the parts by lasercutting? (first part is wrong)
18:32
se6astian
we are mixing two projects now
18:32
se6astian
the Alpha prototype is what we currently have, thats not going to be produced at all beside a few hand assembled pieces
18:33
se6astian
that we already have now, one team member is interested in creating one himself as well
18:33
se6astian
what we plan to produce in higher volume is the Axiom camera
18:33
se6astian
not the Alpha prototype
18:34
VBenjFR
OK, I ask about that because as the quantity increase, we will have to produce more pieces externally
18:34
Bertl
which will have its own prototype (the axiom camera)
18:34
se6astian
yes, to make things even more confusing ;)
18:35
Bertl
i.e. the Axiom Alpha prototype is a proof of concept
18:35
se6astian
Axiom camera produced externally agreed
18:35
se6astian
maybe assembled in house
18:35
se6astian
Ben, do you have experience with aluminum molding?
18:35
VBenjFR
so we will have to give technical drawings for the production with dimensions and tolerence
18:35
VBenjFR
drawings separated for each parts
18:35
se6astian
how many pieces does one need to make that a viable solution
18:36
se6astian
high cost to produce mold forms, low cost for actual pieces, where is the break even point roughly?
18:36
VBenjFR
I have real experience with ceramics molding
18:36
VBenjFR
For aluminium molding, I only studied it
18:36
se6astian
good :)
18:37
se6astian
Am I right expecting to require a 2 piece steel mold form that is likely to be CNC milled?
18:37
VBenjFR
A small series is about 50 - 100 pieces
18:37
VBenjFR
by mould
18:40
VBenjFR
Small series productions are not really intended for low cost pieces
18:41
VBenjFR
Which are the part you want to realised by molding?
18:41
se6astian
at what volume is the break even point where it starts to make sense?
18:42
Bertl
probably 500+
18:42
VBenjFR
It depend of the sell price:
18:43
VBenjFR
If you plan to produce 100 pieces, you will choose a price affected to each part
18:43
VBenjFR
and regarding that you will choose the producing method appropriated
18:44
VBenjFR
More standard pieces used, cheaper is the final object
18:44
VBenjFR
For a aluminium mold, you can use a cheap one in sand
18:45
VBenjFR
the external aspect wil not be good, but price will be OK
18:47
VBenjFR
I will try to get information on price for aluminium molding on 100 pieces...
18:48
VBenjFR
and for 500 if it would be cheaper
18:49
VBenjFR
Sorry, I have to go for my little girl. I will come back with more informations.
18:49
VBenjFR
See you!
18:50
VBenjFR
left the channel
21:01
philippej
joined the channel
21:32
philippej
left the channel
22:31
se6astian
good night
22:31
se6astian
left the channel
23:10
intracube_afk
changed nick to: intracube