Soumis par Sebastian le jeu, 06/23/2016 - 16:05
Currently, most of our local team is assigned to the camera assembly facility. That unfortunately means that this website and new team talks will receive a bit less attention than they deserve and be less frequent as we focus all our time and energy on building new AXIOM Betas. So far, eleven full cameras have been assembled and shipped to the community of early adopters. We build 5 to 6 PCBs for each camera. The simpler boards have around 15 components on them, but our power board is the most complex and hosts over 260 components. Each component needs to be picked up and placed on the PCB by a team member, but there are several steps before and after such placement in the process. Here we want to share some impressions from our assembly lab.
Our assembly facility is located at Amescon - a company in the fifteenth district of Vienna that manufactures special purpose machinery and does software and electronics engineering. With the Amescon equipment, we produce electronic boards of high quality - at this point, our focus is on quality above quantity - it's important for us to pay attention to the details of the entire production process and control every aspect ourselves. We take the time to test every assembled board deeply and thoroughly to make sure that all components are working 100% as expected when we ship a camera to an early adopter/developer. Later on, we plan to switch to more automation for most of the components. We are also currently growing our local assembly staff to increase production speed - new staff members have undergone training last week and started producing actual camera boards this week. If this sounds like an interesting job and you live in Vienna or close by, please contact us.
Want to provide feedback or suggest improvements? Want to try your own approach and need help?
16 Commentaires
No anti-static wrist bands ??
No anti-static wrist bands ??
Very observant :)
Very observant :)
You guys are absolutely,
You guys are absolutely, wonderfully nuts. I am a proud backer and so admire what you're doing and how you're doing it.
Much love to the team.
I love it. It's truly an open
I love it. It's truly an open source AND handmade camera!
Cool, that the production of
Cool, that the production of cameras will accelerate.
Waiting for news of software to control the camera and lens (firmware), and internal recording.
Looks like excellent progress
Looks like excellent progress. Keep up the good work!
Thanks, we try our best :)
Thanks, we try our best :)
Absolutely love it. I'm
Absolutely love it. I'm really stoked for the Axiom coming out - hope you recorded some video in the assembly process for the documentary ;)
Reminds me of the days of
Reminds me of the days of fine watchmaking. Attention to detail and quality brings such a nice karma to the project. I admire and respect the Europeans for their craftsmanship.
Nice but something I could
Nice but something I could afford would be better I guess it will comm intime
Is the camera enclosure going
Is the camera enclosure going to be like the one in the pictures?
I did not spot a camera
I did not spot a camera enclosure in any of these pictures, which picture do you mean exactly?
pic14:
pic14:
A column of vias in the middle of the board, comming out of the qfn package, is not a good practice.
I am pretty sure it creates a slot in the GND layers beneath it.
Of course there are stitching GND vias nearby, but especially for high speed signals this should be avoided.
You can move every second diff pair to the front for example.
(camera hardware/electrical engineer myself - designed multiple cameras around CMV12000 sensor)
Note: Pic 14 shows the HDMI
Note: Pic 14 shows the HDMI Plugin Module with solder paste applied
Hey Mishka!
Hey Mishka!
Let's take a look at the "complete picture" first: http://vserver.13thfloor.at/Stuff/AXIOM/1xHDMI_v0.8_vias.png
I wish I could offset the vias in such a way that there would be no slot in the GND layers, but even if I move every other pair closer to the QFN, it would still create a slot unless I drop the GND stitching, which might be an option but probably makes things worse.
Note that I can't move away from the QFN, as I would end up in the pads of the HDMI connector.
If you have any suggestions how to improve on the design without making the plugin larger, I would very much appreciate the input.
All the best,
Herbert
Hey Herbert,
Hey Herbert,
Yes, I understand that it is pretty dense design - and nicely done by the way.
I don't think this would cause any issues of course - it's just me doing the high speed signals for some time and i always watch out for such thing, just to be on the safe side:
- reference plane discontinuities,
- reference plane cutouts under the large pads (to decrease unnecessary capacitance and impedance change)
- via stitching
etc.
I have prepared some pictures with how you could potentially fix this + example of my design (8GT/s PCIe), but my message gets blocked when i copy paste a Google Drive link to these pictures. If you drop me an email to mishka1985@gmail.com, i can send it to you in the attachment, and you can upload it to this discussion.
I like the whole concept of the open source camera that you guys are doing!
Keep it up!
Mishka
Ajouter un commentaire