a 3D printed "fake" battery adapter - Hotwire! Hardware Mods, Accessories and Insights - CHDK Forum  
« previous next »
Pages: [1] 2   Go Down

a 3D printed "fake" battery adapter

  • 16 Replies
  • 5958 Views
*

Offline waterwingz

  • ******
  • 12138
    • Waterwingz
a 3D printed "fake" battery adapter
« on: 22 / January / 2015, 19:46:28 »
Advertisements
This might be of some interest :

3D Printed Camera Battery Emulator from hackaday.com.

Sure beats carving one out of wood.
Logged
Ported :   A1200    SD940   G10    Powershot N    G16
*

Offline andrew.stephens.754365

  • ****
  • 461
Re: a 3D printed "fake" battery adapter
« Reply #1 on: 23 / January / 2015, 08:02:41 »
Had just been thinking about this - nice one.
Logged
*

Offline apaolucci

  • *
  • 6
Re: a 3D printed "fake" battery adapter
« Reply #2 on: 01 / February / 2015, 06:24:40 »
@waterwingz I'm realizing a 64 camera Rig based on the project started and realized by mphx as described in his thread http://chdk.setepontos.com/index.php?topic=11667.0 where you, reyalp and others had supported him to solve a lot of problems.

My project, instead of the Powershot a2500 that is out of production, is based on 64 Canon Ixus 132 (8 stands by 8 cameras) and since I'm a 3d Printing enthusiastic, owner of a Sharebot 3d printer, I've designed and 3dprinted the NB-11L fake battery planning to optimize the power of the camera with a single power supply per stand. I've bought a 4.5V by 2500mA power supply should be enough for 8 cameras.
AS you can see in the fotos, I've used a normal bipolar cable (the one normaly used for audio HiFi) thinking the power was not so big to generate signal interference on usb cable (all shielded).

Instead, when I set the cameras to rec mode using the ckdkptp comand:
Code: [Select]
!return mc:cmdwait('rec')only 3/4 cameras remain connected (I received an I/O error) in a random way. Instead if I connect max 3 cameras everything works correctly.

Looking for some hint in the forum about realizing fake batteries, I've found this thread so I've thought to post my issue here. I hope I have done well.
I've also read in your thread http://chdk.wikia.com/wiki/Shooting_with_Multiple_Cameras_using_CHDK in the Power paragraph: " Be careful about net power draw, grounding and wiring if you go this route.", but I don't know what you mean about wiring and grounding.
I think the problem should be the eddy current, but I'm not aware how to solve it.
Can you help me with some hint on which wire to use and how to connect the fake battery and power supply?

As reference I've attached two fotos: a detail of the camera wired toghether and NB-11L fake batteries.

Any suggestions are wellcome, thanks in advance.
Logged
*

Offline waterwingz

  • ******
  • 12138
    • Waterwingz
Re: a 3D printed "fake" battery adapter
« Reply #3 on: 01 / February / 2015, 10:12:16 »
Quote from: apaolucci on 01 / February / 2015, 06:24:40
@waterwingz I'm realizing a 64 camera Rig based on the project started and realized by mphx as described in his thread http://chdk.setepontos.com/index.php?topic=11667.0 where you, reyalp and others had supported him to solve a lot of problems.
Are you worried about the "sync" quality for your project?  Or hoping that PTP shooting (using reyalp's sync calibration code) will be enough?  If you are going with hardward sync, then in the most recent posts there,  I've started to think we have made things too hard by trying to switch all the +5V signals to the cameras seperately.   Too many hardward mods & switches & MOSFETS and stuff - I believe there is an easier and thus cleaner way.

Quote
I've bought a 4.5V by 2500mA power supply should be enough for 8 cameras.
Hmmmm ... not sure how you did the math,  but in shooting mode you should assume a Powershot will need at least 500mA with peaks to a much as 1A.  For eight cameras,  that means you need between 4000mA and 10000mA if they are all wired in parrallel.   Most of the time the peaks will not occur at the same time so you should only crash occassionally if you use less than a 10A supply.   Also, be aware that a supply rated 2500mA may not be capable of producing that current continuously at room temperature.

If you want to use the 2500mA supplies, I suggest only four cameras per supply and some heavy wire to connect them ( at least 16 gage).

Quote
A you can see in the fotos, I've used a normal bipolar cable (the one normaly used for audio HiFi) thinking the power was not so big to generate signal interference on usb cable (all shielded).
The issue here will be wire gage - the thicker the better!   For power wiring, that's far more important that shielding.  During peak current draw, thin wires can cause enough voltage drop for the camera to shut down thinking its batteries have expired.

I'd also recommend running the power wires from each camera all the way back to the power supply - one dedicated run per camera.  You seem to have them daisy chained and that could cause problems.   

Wire length also matters (longer wires being worse) so the closer you can get the power supplies to the cameras the better.  Mounting them half way up your poles would probably be optimal.

Quote
Instead, when I set the cameras to rec mode using the ckdkptp comand:
Code: [Select]
!return mc:cmdwait('rec')only 3/4 cameras remain connected (I received an I/O error) in a random way. Instead if I connect max 3 cameras everything works correctly.
Not surprising given the comments I have posted above.

Quote
I've also read in your thread http://chdk.wikia.com/wiki/Shooting_with_Multiple_Cameras_using_CHDK in the Power paragraph: " Be careful about net power draw, grounding and wiring if you go this route.", but I don't know what you mean about wiring and grounding.
See my comments above.

Quote
I think the problem should be the eddy current, but I'm not aware how to solve it.
Eddy current is a very specific term in electrical engineering and not likely to be an issue here.  You might be thinking about Ground Loops, another very specific term, and that could be an issue here.  But your more likely issue is just plain old voltage drop as the current increases.

Really nice job on the "fake batteries" by the way!   And thanks for attaching the pictures - always helps when talking through hardware builds!  I'll see if I can find a wiring diagram that I can modify with my current thinking about the +5V switching.   What are you planning on using for USB hubs ?


Edit :  PDF of some camera current measurement attached
« Last Edit: 01 / February / 2015, 10:36:01 by waterwingz »
Logged
Ported :   A1200    SD940   G10    Powershot N    G16
*

Offline reyalp

  • ******
  • 14033
Re: a 3D printed "fake" battery adapter
« Reply #4 on: 01 / February / 2015, 13:59:21 »
Quote from: waterwingz on 01 / February / 2015, 10:12:16
Are you worried about the "sync" quality for your project?  Or hoping that PTP shooting (using reyalp's sync calibration code) will be enough?  If you are going with hardward sync, then in the most recent posts there,  I've started to think we have made things too hard by trying to switch all the +5V signals to the cameras seperately.
Since apaolucci is using an cameras with a lithium ion battery, it might be worth pointing out the option would be to use the battery temp sensor for triggering: http://chdk.setepontos.com/index.php?topic=10385.0

With 3d printed fake batteries it should be possible to integrate this very cleanly.
Logged
Don't forget what the H stands for.
*

Offline apaolucci

  • *
  • 6
Re: a 3D printed "fake" battery adapter
« Reply #5 on: 01 / February / 2015, 18:04:50 »
@waterwingz
Quote
Are you worried about the "sync" quality for your project?  Or hoping that PTP shooting (using reyalp's sync calibration code) will be enough?
I'm using chdkptp multicam lua script and chdk v1.3 stable (build 3947). I do not need extreme syncing since the model stand in the middle of the rig and do not move. I've already realized a prototype with only 5 cameras (I moved around the model to take fotos) that let me scan and 3d print a model dressed as a bride with encouraging results. The only thing I've missed was the correct color balance (and also white balance) for 3dprint, resulting in a gray miniature!!!! :o  I've attached a pic of it as reference. You can imagine my surprise when I received the miniature ... I was totaly astonished!!!  :blink: :blink:

Quote
Hmmmm ... not sure how you did the math,  but in shooting mode you should assume a Powershot will need at least 500mA with peaks to a much as 1A.
I based my math on the NB-11L specification (3.6V by 680mAh), ok I've completely missed the right size of the power supply, but I've also planned to use one supply per 4 cameras. My primary mission was not to have 64 power supplies  8), so 16 should be also ok.

Quote
... and some heavy wire to connect them ( at least 16 gage).
Ok, I've used a 0.75mm^2 wire, 16 Gauge is about 1.3mm^2 the bouble I've used. I'll try with big one.

Quote
I'd also recommend running the power wires from each camera all the way back to the power supply - one dedicated run per camera.  You seem to have them daisy chained and that could cause problems.   
Wire length also matters (longer wires being worse) so the closer you can get the power supplies to the cameras the better.  Mounting them half way up your poles would probably be optimal.
Ok this sound good, I'll try with this configuration.

Quote
What are you planning on using for USB hubs ?
The hub are used only to connect cameras to the PC, controll them and download the pics to PC. Every stand has a 10 port hub, 8 port for the cameras and one port to connect to the nearby stand. Since manage 8 stands in series is not so good, I've planned to use two usb PC's port so to make two chain of 4 stand in series. My primary intention is not to use a star schema since manage the central hub is not simple (in term of cable lenght and obstruction). I've not test it, my is only theory for the moment.

As soon as I've tested this new solutions, I'll post my progess.

@reyalp  I've not read all the thread you post me, only first 2 pages, but I did not understand the content. Since it's about 19 pages, can you, in very few words, tell me how it can help me achieving the fake battery structure? How reading the battery temperature can help solve the voltage drop? Thanks in advance.
Logged
*

Offline waterwingz

  • ******
  • 12138
    • Waterwingz
Re: a 3D printed "fake" battery adapter
« Reply #6 on: 01 / February / 2015, 18:24:04 »
Quote from: apaolucci on 01 / February / 2015, 18:04:50
I'm using chdkptp multicam lua script and chdk v1.3 stable (build 3947). I do not need extreme syncing since the model stand in the middle of the rig and do not move. I've already realized a prototype with only 5 cameras (I moved around the model to take fotos) that let me scan and 3d print a model dressed as a bride with encouraging results.
Based on that,  you can probably get away with just using PTP based shooting - no hardware sync tricks required.

Quote from: apaolucci on 01 / February / 2015, 18:04:50
@reyalp  I've not read all the thread you post me, only first 2 pages, but I did not understand the content. Since it's about 19 pages, can you, in very few words, tell me how it can help me achieving the fake battery structure? How reading the battery temperature can help solve the voltage drop? Thanks in advance.
/reyalp mode on
That thread talks about using the "third terminal" that you find on custom Li batteries (used normally by battery chargers to monitor an internal temperature sensor in the battery to ensure it does not overheat during charging).   Canon P&S cameras can read that temperature sensor but don't seem to do anything with the information.   So CHDK has been modified to allow you to attach an external switch to that input and use it to trigger synchronized shooting rather than use the USB port for that purpose. That frees the USB port for PTP communication without any further hacks required.
/reyalp mode off
« Last Edit: 01 / February / 2015, 18:27:40 by waterwingz »
Logged
Ported :   A1200    SD940   G10    Powershot N    G16
*

Offline reyalp

  • ******
  • 14033
Re: a 3D printed "fake" battery adapter
« Reply #7 on: 01 / February / 2015, 18:30:04 »
Quote from: apaolucci on 01 / February / 2015, 18:04:50
I based my math on the NB-11L specification (3.6V by 680mAh), 
The h is important, mAh is total power capacity, not instantaneous current. You might be tempted to take capacity/run time to get average power draw, but this won't work either because as waterwingz post shows, the power consumption depends a lot on what the camera is doing at a given moment.
Quote
@reyalp  I've not read all the thread you post me, only first 2 pages, but I did not understand the content. Since it's about 19 pages, can you, in very few words, tell me how it can help me achieving the fake battery structure? How reading the battery temperature can help solve the voltage drop? Thanks in advance.
Sorry for the confusion. It would not help with the voltage drop. It might simplify the triggering if you need better sync than you can get with pure PTP. In essence, it lets you keep the sync control separate from the USB power.
Logged
Don't forget what the H stands for.
*

Offline apaolucci

  • *
  • 6
Re: a 3D printed "fake" battery adapter
« Reply #8 on: 02 / February / 2015, 10:11:42 »
@reyalp
Quote
It might simplify the triggering if you need better sync than you can get with pure PTP. In essence, it lets you keep the sync control separate from the USB power.
It sounds interesting, surely I will read all the post so I'll have a more detailed idea about the item. At the moment I have no idea how to integrate this new mode of triggering in my pipeline.

@waterwingz Thanks for the pdf with current measures, it contains a lot of useful informations

This evening, when I come back home I'll try the improvements and let you know. Thanks
Logged
*

Offline reyalp

  • ******
  • 14033
Re: a 3D printed "fake" battery adapter
« Reply #9 on: 02 / February / 2015, 17:22:27 »
Quote from: apaolucci on 02 / February / 2015, 10:11:42
It sounds interesting, surely I will read all the post so I'll have a more detailed idea about the item. At the moment I have no idea how to integrate this new mode of triggering in my pipeline.
If you are shooting static objects, using PTP multicam only should be fine, and much simpler than any sort of hardware trigger.

That said, if you print more fake batteries, it would make sense to allow for the option of wiring up the temperature line in case you ever need better sync in the future.
Logged
Don't forget what the H stands for.
Pages: [1] 2   Go Up
« previous next »
 

Related Topics