CHDKPTP - PC Remote Control Performance Analysis - page 31 - RAW Shooting and Processing - CHDK Forum

CHDKPTP - PC Remote Control Performance Analysis

  • 465 Replies
  • 94286 Views
*

Offline reyalp

  • ******
  • 13624
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #300 on: 29 / September / 2012, 00:04:37 »
Advertisements
@reyalp
The function call did work =set_config_value(67,3) // it's not elegant but still running fine for 6 hours. 
Where do I find these defines ... http://chdk.setepontos.com/index.php?topic=8613.msg91341#msg91341  ??
http://trac.assembla.com/chdk/browser/trunk/include/conf.h
Don't forget what the H stands for.

Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #301 on: 29 / September / 2012, 09:08:35 »
I did it on an S50 many years ago.  The job is not for the faint-of-heart.

So, the S70 was not modified ?

Either way, you said these "technical" issues had previously been covered in your discussions with reyalp.

In what thread, I cannot find any reference here ?

With the S90, you can at least bend the sensor FPC back on itself and attach to the connector with a piece of VHB tape.

It is then facing in the opposite direction.

Don't know if that hinders TEC.

*

Offline SticK

  • *****
  • 779
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #302 on: 29 / September / 2012, 10:54:36 »
Quote "So, the S70 was not modified ?"

Nope, never was // there was very little to be gained over the S50.  In contrast, the S90 provides a massive gain in low-light sensitivity, live image definition, ease of manipulation, and data speed (if srsa can pull it off).

Quote "Either way, you said these "technical" issues had previously been covered in your discussions with reyalp"

I don't know what exactly you are referring to ... perhaps you're thinking about this ? ...  I did S50 a very long time ago and have nothing on the web anywhere // I did not even take photos then - a stack of pencil drawings!  For the S90, if successful, which I really doubt, I will post something here.

Quote "With the S90, you can at least bend the sensor FPC back on itself and attach to the connector with a piece of VHB tape."  "It is then facing in the opposite direction."

Correct.  In my explanation yesterday, I referred to this installation option as "parallel in front of the large black chip to the right of its (the CCD's) connector."  I also explained there this mounting option gives anyone wishing to do it a maximum about 1 cm of clearance between back of the CCD and large black chip.  Please refer to Fig 2 in that post.  For my application, if successful at ridding the lens, the CCD will go perpendicular in which case I will have a machine shop churn out the thermal and mounting hardware // that will happen in Phase IV much later on.


Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #303 on: 29 / September / 2012, 12:09:22 »
I don't know what exactly you are referring to

http://chdk.setepontos.com/index.php?topic=8613.msg90638#msg90638

though it simply states the technical issues were resolved long ago without being specific.



*

Offline SticK

  • *****
  • 779
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #304 on: 29 / September / 2012, 13:36:37 »
There's a fella from Argentina that modified an early PowerShot in around the same era that I did the S50 and he documented his efforts on a web site.  I loosely recall his name might be Miguel.  If you can't google it, read through my early discussions, likely in one of my SX110 threads, where someone (possibly reyalp) pointed me there.  Follow the links to his site.  If you find it, then that's the idea.  The S90 lens will be significantly more complicated // we'll see when the item arrives here // hopefully in a week or two.

Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #305 on: 29 / September / 2012, 13:45:08 »
Very tricky.

Some signals are expected to change state within a certain time period.

Another signal is expected to be continual pulses for a certain length of time.
Another is expected to lead to an increase or decrease of light on the sensor.

Obviously,manual mode with "zoom at wide angle signal" (after slight delay) would seem a good start and obviously no zooming.

You could start by disconnecting the lens-extended sensor and simulating it with a simple switch that you close (or open) within one second of booting.


Check this out (but you cannot see any images)  :-

http://www.cloudynights.com/ubbthreads/showflat.php/Cat/0/Number/2742127
« Last Edit: 29 / September / 2012, 14:03:16 by Microfunguy »

*

Offline SticK

  • *****
  • 779
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #306 on: 29 / September / 2012, 14:38:13 »
Quote "Tricky."

Indeed.  The connector is very dense ... I do my microelectronics work under a mineralogical stereo microscope ... so yes ... we *are* on the same page.

Quote "Some signals are expected to change state within a certain time period. Another signal is expected to be continual pulses for a certain length of time. Another is expected to lead to an increase or decrease of light on the sensor."

Sure no question about that.  The hope I have is that most of those will be ignored by firmware when the camera is put (or better PUPed) into Manual mode, and ignored before the camera is in its defined PUP state ready-to-shoot.  I question I will seek to answer are which ones .. basically on the same line of thinking for what I did with the camera ... strip the lens down as far as possible before failure occurs ... and then how to get around it, just gets too complicated to speculate about!

Quote "You could start by disconnecting the lens-extended sensor and simulating it with a simple switch that you close (or open) within one second of booting."

Certainly ... that is one of the obvious possibilities.  I think the best approach is first to study the lens with a step-wise non-destructive disassembly as I mention above without intervention, testing functionality at each step.  We have to understand what does what first in terms of identifiable subassemblies before even attempting to isolate groups of signal traces responsible for each.  Just that alone could take me several days.

Quote "Check this out (but you cannot see any images)"

Thanks .. yes I can see images ... click on "attachment" in the post header, when there is one.  The Miguel site I think is more informative if you can find it.

Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #307 on: 29 / September / 2012, 14:43:54 »
I do my microelectronics work under a mineralogical stereo microscope
So do I and with an Antex 0.2mm soldering iron tip.

Time for dinner  .........


Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #308 on: 29 / September / 2012, 14:52:32 »
There's a fella from Argentina that modified an early PowerShot in around the same era that I did the S50 and he documented his efforts on a web site. 
http://chdk.setepontos.com/index.php?topic=3409.0
Ported :   A1200    SD940   G10    Powershot N    G16

*

Offline SticK

  • *****
  • 779
Re: CHDKPTP - PC Remote Control Performance Analysis
« Reply #309 on: 29 / September / 2012, 15:03:11 »
@Microfunguy
... with the hands and skills of a surgeon.  Buen provecho, in case case you're reading the Leonardo site.

@waterwingz
Thanks .. the 1st link is the good one.

 

Related Topics