350d boot from cf card

peabody

Re: 350d boot from cf card

« Reply #340 on: 23 / November / 2010, 23:35:14 »

Advertisements

I re-wrote the bootflag.fir file to fix the problems and to check for v1.0.3, and my version seems to work fine. An autoexec.bin file loads and executes.

But the wiki says that when I insert a bootable CF card but with no autoexec.bin file present, I will get an error message saying there's no autoexec.bin file. Well, I don't get any such message. In fact, the camera is completely dead, with no display and no lights. And I have to cycle the battery to get the camera back to normal. It's as though the camera is in an infinite loop as a result of there being no autoexec.bin file present.

What am I missing here? Is the wiki wrong about that error message, or is it just me?

Logged

peabody

Re: 350d boot from cf card

« Reply #341 on: 24 / November / 2010, 09:11:18 »

About the shutter release count - mine shows 21780, which is a good bit lower than I expected. I just got the camera, but I'm the third owner that I know of, and just would have expected a higher number. Anybody else get a suspiciously low number?

One reason it looks suspicious is that I'm on folder 643, which if they started at 100 would indicate over 50k shutters. But I guess that can happen if someone moved a CF from another camera which had a higher count. Well if my release count is right, is there any way to change the folder number back to what would be right for the release count?

I'll try to post my revised boot flag .fir stuff later today, assuming you can still attach files here.

Logged

peabody

Re: 350d boot from cf card

« Reply #342 on: 24 / November / 2010, 18:04:39 »

In case anyone is interested, attached is bootflg2.zip, which contains bootflg2.fir - my reworking of the bootflag.fir file. This version checks for the presence of firmware v1.0.3, and also checks the initial status of the boot flag area, before proceeding with the process of modifying the boot flag. There's a ReadMe.txt included with more detail. Oh, and it does actually zero out those two buffers instead of pretending to. Commented C code is included.

This does work on my camera, but I make no other promises. The new file still depends on the two subroutines called by the original, and we still don't know anything about them. So I think some risk remains.

I would appreciate any comments or suggestions from anyone who wants to review the C code. I don't really know much about C, but did confirm that the object code does what I wanted - see bootflg2.asm.

« Reply #347 on: 30 / November / 2010, 12:11:16 »

when you do the hack, you change only some bits in the firmware, allowing the camera to boot code from the cf memory card. the rest of the hack is contained in the file autoexec.bin. it's loaded and executed when the camera boots. it only contains the additional code, as you said.
for what you want to do, you need to change the actual firmware (where we only change the boot flag). that's much more dangerous because you can easily destroy your camera.

I have seen instructions on the internet to remove the IR filter that account for the difference in auto focusing and try to preserve the af function. I don't know if this helps you, (I don't know the specifics) but maybe you can find some info that tells you what you can do on the hardware part.

Logged

a1ex

671
ML dev

Re: 350d boot from cf card

« Reply #348 on: 30 / November / 2010, 18:00:15 »

You may try to find the amount of misfocus correction; after the camera focused, you call a routine which rotates manually the focus motor. I don't know if this can be done on the 350D; only know that Magic Lantern implemented rack focus, which also rotates the focus motor.

Since you don't have any feedback, don't expect very good results, but I think it will be better than it's now.

For a (much easier) hardware solution, see here: http://www.astrosurf.org/buil/autofocus/adjust.htm . Don't know if the amount of correction is enough for IR, though.

Logged

curunir

Re: 350d boot from cf card

« Reply #349 on: 01 / December / 2010, 16:02:07 »

Okay, so changing the actual firmware is a dangerous job, I understand that. Is what a1ex says do-able? Is there a way to create my own routine that actually drives the focus motor? Can this routine then be bound to one single button, like let's say the star button which I hardly use? So no menu interventions?

Also, I've seen the link you gave me. But I read that this method only rotates it, where I actually want to bridge the gap that removing the hot mirror has created, which is physically impossible (I mounted the sensor in the camera myself ;-)). Do you think this will actually fix it? Then maybe I should just try this!

Logged