CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD - page 20 - Creative Uses of CHDK - CHDK Forum supplierdeeply
« previous next »
Pages: 1 ... 15 16 17 18 19 [20] 21 22 23 24 25 ... 71   Go Down

CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD

  • 704 Replies
  • 170955 Views
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #190 on: 15 / November / 2012, 17:15:06 »
Advertisements
TEST RESULTS

Test #1:  IRIS connected: PUP OK into PLAY.  PLAY->REC Ok.  REC->PLAY Ok.  PDN. // code validates fine.

Test #2: Disconnect IRIS:  PUP OK into PLAY.  PLAY-> REC: immediate PTP ERROR.  Camera continues to run, ie no auto-shutdown // PDN with POWER SWITCH.  Good step forward?

Test #3: Reconnect IRIS: PUP OK into PLAY.  ROMLOG_01: IrisController.c, attached.
Logged
*

Offline srsa_4c

  • ******
  • 4451
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #191 on: 15 / November / 2012, 18:15:06 »
Quote from: SticK on 15 / November / 2012, 17:15:06
Test #2: Disconnect IRIS:  PUP OK into PLAY.  PLAY-> REC: immediate PTP ERROR.  Camera continues to run, ie no auto-shutdown // PDN with POWER SWITCH.  Good step forward?

Test #3: Reconnect IRIS: PUP OK into PLAY.  ROMLOG_01: IrisController.c, attached.
I think that the lack of error shutdown is probably caused by the non-default value in 0x7938.
The next diskboot should suppress the assert seen in your romlog.
https://subversion.assembla.com/svn/chdk-s1.bin/files/DISKBOOT_s90_101a_lensctrl_try2.zip
Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #192 on: 15 / November / 2012, 19:53:39 »
Yes it suppresses the IrisContoller.c assert but not the effect.  That is, PTP gets disabled on REC.  I can still PDN with camera POWER switch.  That's the only difference that I can see from before.

As usual, the IRIS has to be reconnected to revive the camera and the USB.

edit:
ROMLOG_02
« Last Edit: 15 / November / 2012, 20:04:17 by SticK »
Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #193 on: 15 / November / 2012, 20:49:52 »
ROMLOG_02 as edit in my previous post.
Logged
*

Offline srsa_4c

  • ******
  • 4451
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #194 on: 16 / November / 2012, 18:59:05 »
@SticK
We could continue like this (one assert at a time), but I think that would take a long time.
I could also disable all asserts, but I don't know what possible serious side effects that could cause. (disable = all asserts ignored, even unrelated ones, like "out of memory")
The iris control related code seems pretty complicated (apparently it has many checks while the camera operates), wonder how complicated the other mecha parts are...
Faking hw signals may be easy, but they certainly are not easy to find/follow in the code. We can also forget about faking them from software (the used GPIO addresses are not known, they are buried deep in the code).
Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #195 on: 16 / November / 2012, 19:33:19 »
@srsa_4c
Quote "I could also disable all asserts, but I don't know what possible serious side effects that could cause. (disable = all asserts ignored, even unrelated ones, like "out of memory")"
What we know so far is the only real serious threat is the malloc fail assert, and we know that is predictable by observing the decline of the available pool over shots taken and rebooting.  So if possible to do, this could be a good gross test just to see fundamentally if the camera would continue to take shots.  What do you think?  Possible to try?

Quote "The iris control related code seems pretty complicated (apparently it has many checks while the camera operates)"
One thing I know for certain is that once the IRIS has been moved out of stow on a PLAY->REC, the PIs do not get powered so it does not check switch states during normal shooting.  It would only check the IRIS switch state after it stows from REC->PLAY.   Hence the operations you a referring to are probably all normal shooting operations and likely have no relation to error checking ... unless of course it tests motor loads occasionally which is possible, but that is not something I can find out directly anyway without building an IRIS electronic dummy.

Quote "wonder how complicated the other mecha parts are."
FOCUS in general resembles IRIS.  ZOOM has two parts: a) the stow/unstow switch which is like FOCUS, and b) servo-control feedback control of the motor (servo motor unlike stepper in FOCUS and IRIS).  To handle that one electronically might be very tough.  In the worst case, if the ZOOM unstow/unstow can be handled (FOCUS has to go first), then motor can be cut out of the assembly, removing the major bulk ... would mean butchering the hardware but can be done.

Quote "Faking hw signals may be easy, but they certainly are not easy to find/follow in the code. We can also forget about faking them from software (the used GPIO addresses are not known, they are buried deep in the code"
Understood. edit: We had originally toyed with the idea of removing the PIs are driving the LEDs to fake the switch signals.  However, we now know that the PI power of all three is tied so toggling it would affect all three at once ... a recipe that goes nowhere anyway.  Thus even knowing the pins is of not much use I think given all the new info on the subject.  As far as the phototransistors go, they're AINs anyway and my guess is they can't be toggled either.

« Last Edit: 16 / November / 2012, 19:55:53 by SticK »
Logged
*

Offline philmoz

  • *****
  • 3450
    • Photos
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #196 on: 16 / November / 2012, 19:45:27 »
Just a suggestion - the forum tools for quoting previous messages make the posts easier to read (IMO).

Phil.
Logged
CHDK ports:
  sx30is (1.00c, 1.00h, 1.00l, 1.00n & 1.00p)
  g12 (1.00c, 1.00e, 1.00f & 1.00g)
  sx130is (1.01d & 1.01f)
  ixus310hs (1.00a & 1.01a)
  sx40hs (1.00d, 1.00g & 1.00i)
  g1x (1.00e, 1.00f & 1.00g)
  g5x (1.00c, 1.01a, 1.01b)
  g7x2 (1.01a, 1.01b, 1.10b)
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #197 on: 16 / November / 2012, 20:03:30 »
@philmoz
I haven't figured out how to quote a part of a message: select and quote??  When I press Insert Quote it puts in the whole message.
Logged
*

Offline srsa_4c

  • ******
  • 4451
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #198 on: 16 / November / 2012, 20:17:51 »
Quote from: SticK on 16 / November / 2012, 19:33:19
What we know so far is the only real serious threat is the malloc fail assert, and we know that is predictable by observing the decline of the available pool over shots taken and rebooting.  So if possible to do, this could be a good gross test just to see fundamentally if the camera would continue to take shots.  What do you think?  Possible to try?
Possible, but - as always - on your own risk.

Quote
One thing I know for certain is that once the IRIS has been moved out of stow on a PLAY->REC, the PIs do not get powered so it does not check switch states during normal shooting.  It would only check the IRIS switch state after it stows from REC->PLAY.
That's useful to know.
Quote
   Hence the operations you a referring to are probably all normal shooting operations and likely have no relation to error checking
Good point. Since the current "solution" is to ignore the errors, the state of the hw still looks bad from the view of the firmware. The problem part is that I don't know how many spots would have to be fixed to make this go away.

Quote
FOCUS in general resembles IRIS.  ZOOM has two parts: a) the stow/unstow switch which is like FOCUS, and b) servo-control feedback control of the motor (servo motor unlike stepper in FOCUS and IRIS).  To handle that one electronically might be very tough.  In the worst case, if the ZOOM unstow/unstow can be handled (FOCUS has to go first), then motor can be cut out of the assembly, removing the major bulk ... would mean butchering the hardware but can be done.
Thanks for the description. Instead of "butchering" your hardware, you could get a broken (Canon) camera to extract similar (motor + opto) parts from. Not much difference though, you won't use S90 lens assembly you are working on anyway (for its original purpose).

Quoting: there's a quote button above the edit area
Select some text, then press the button. I usually type the quote tags myself  :haha
« Last Edit: 16 / November / 2012, 20:25:26 by srsa_4c »
Logged
*

Offline philmoz

  • *****
  • 3450
    • Photos
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #199 on: 16 / November / 2012, 20:19:52 »
Quote from: SticK on 16 / November / 2012, 20:03:30
@philmoz
I haven't figured out how to quote a part of a message: select and quote??  When I press Insert Quote it puts in the whole message.

Click the 'Preview' button will take you to the full editor.

Phil.
Logged
CHDK ports:
  sx30is (1.00c, 1.00h, 1.00l, 1.00n & 1.00p)
  g12 (1.00c, 1.00e, 1.00f & 1.00g)
  sx130is (1.01d & 1.01f)
  ixus310hs (1.00a & 1.01a)
  sx40hs (1.00d, 1.00g & 1.00i)
  g1x (1.00e, 1.00f & 1.00g)
  g5x (1.00c, 1.01a, 1.01b)
  g7x2 (1.01a, 1.01b, 1.10b)
Pages: 1 ... 15 16 17 18 19 [20] 21 22 23 24 25 ... 71   Go Up
« previous next »
 

Related Topics