SDrive-MAX ATX support

[_The Doctor__]

Sure, not all cracks are complete or properly dons.. CSS is working to correct this, it may be some years till it's all done...

 

By using an ATX, extended ATX, or PRO image, you get everything for the most part working and timed exactly as written and executed properly without hidden pitfalls that prevent a game or software from working perfectly. Even now people are surprised how diabolical the protections could be and the little changes are the difference between the proper look feel or experience a this title or that...

 

It's also great if one were to want to make real disks, (yeah they are working on that)

 

It's great to see all the loading screens and all that as well, cracked by the modem crew is not as nice as the normal splash screens and loaders.. more reasons but those are the main ones

 

working off of originals to make improvements or hacks is far better than working on someones else hacked mess. Different approaches to fixing glitches and broken parts of the originals also get done. We end up with the very best!

Edited September 25, 2018 by _The Doctor__

[_The Doctor__]

After explaining what the screen type is and why, and explaining screen size refers to the screen itself and not the PCB it is attached to. and all sorts of stuff like that... they finally sent me the correct screens!

 

So next time I build for someone I must remember to keep my own unit instead of giving it to the other person. Figuring this is so easy, I can order it again and have it all done in about 15 minutes. Life doesn't work that way when ordering from china!

 

So now I have one put together and waiting on another ardy for the second one!

 

Long story short I now have an S-Drive Max ATX CAS!

 

but flashed at 8b..... and now I need another micro sd card (well 2 of them)

[Level42]

So, it is possible to re-calibrate the screen without re-flashing. That was always my problem. One mis-touch on the screen would render it useless.

 

This thing is a pain in the ass to re-calibrate. It's all very good for the most part, but the screen calibration sucks.

 

 

 

You can always start the calibrate process without re-flashing !

 

steps:

1) power off the Sdrive-Max

2) put finger (or better, the little plastic pen that comes with the display) on the screen and hold it there (doesn't matter where on the screen)

3) power on the Sdrive-Max

4) do the calibration

 

Using the little pen is the best way to accurately calibrate.

Indeed....a pause between the calibration points is very necessary....it would be even better if there was some kind of feedback....since there is no sound, a short green flash when each cross is pressed would be great.

[_The Doctor__]

I just select rotate screen and ta-DA! re-calibrate ... yes it's re-calibration time... and that's how I do it... and then I save after it's calibrated and the screen and everything is set the way I want.. works great! No need to power up and down, insert a cord or not... just go in there and set it up.. almost like it was meant that way... I hope it still works that way on the newer versions...

Edited September 26, 2018 by _The Doctor__

[+DrVenkman]

So here's a question - is the screen calibration a function of the SDrive-Max firmware, or something baked into the screen firmware and checked/run on every power cycle?

[+remowilliams]

So here's a question - is the screen calibration a function of the SDrive-Max firmware, or something baked into the screen firmware and checked/run on every power cycle?

 

It's in the SDrive-Max code - action_cal ()

[_The Doctor__]

and I just upgraded to 9 it works the same way, so no worries. Rotate to re-cal or do it at start up by tapping holding the screen as it powers up. Set all your things the way you like and save...cfg

 

It's good there is more than one way to do it. Just gotta make sure nothing is touching or pressuring the screen during power on. Your box, case, floppy disk, cat, or fat fingers better stay off it

[Level42]

I didn’t think off/use the rotate way to calibrate because rotating made my screen totally messed up some time ago....but I think that was because of another reason...(flashing).

 

Will try this method tonight.

[E474]

Hi,

 

Thanks very much for the feedback on the step-by-step.

 

- I found that if calibration didn't work OK, the screen wouldn't react properly, so I couldn't select config/re-calibrate (I genuinely thought I had been shipped a defective screen). Powering up while pressing the screen worked fine.

- diode has now been trimmed, but I haven't had the time to apply any heat-shrink.

- copper core was wrapped around diode pin, but only once (iirc).

- I think it's better to make sure flashing the firmware is successful, then plug in the screen as you are testing a simpler configuration.

 

 

[_The Doctor__]

Items to request of developer

 

1 ability to save both calibrations sets from rotate or have the code just flip the calibrations in the code, whichever leaves the most free mem in the ardy... might depend on the screen though

 

2 a load config button, this is useful if you make a mistake or cat etc touches screen and slots show he emptied them or other stuff changed, allows to get it all back without power cycling etc simply loads the last saved config. w00t!

 

3 be able hold more than 2 atx disks (if memory would allow-not sure on this)

 

4. a filter to find all files with or all files without whatever extender...more popular ones on buttons? (don't know/remember if file is cas, atx, atr, or what folder it belonged in) more powerful one in the D0:sdrive.atr if that is done - use a config file saved on the sd card from the atari via the sdrive atr or other utility image.

 

that's about all I see in the thread or from me.... just ideas to consider. someone could pass the list along if it looks reasonable and you think it's complete enough

Edited September 27, 2018 by _The Doctor__

[Level42]

The rotating screen method of re-calibration only works if the calibration is already OK enough to actually be able to hit the CFG, ROTATE and SAVE buttons....so the calibration must be at least close to OK.

 

When you fuck up by accidentally touching the same cross twice (or more) like I have done a couple of times, there is no way to go this route.

 

This is why I mentioned the push-screen-during-power-up way, because this will always work.

[_The Doctor__]

yes the fail safe hasn't been requested to be removed... but if we get good calibrations and can save and load them maybe we can avoid the need to use the fail safe. but it's nice to know it's there.

[Level42]

 

- I think it's better to make sure flashing the firmware is successful, then plug in the screen as you are testing a simpler configuration.

 

 

 

Mmmm..... I think it's helpful to have the screen on the UNO while flashing because it provides feedback in the shape of statusbars. However, it works fine either way. You did a great write-up, I'm sure a lot of people will find it helpful

[Level42]

 

It's in the SDrive-Max code - action_cal ()

Where in the code did you find this ?

 

[EDIT] Never mind, found it

Edited September 28, 2018 by Level42

[Level42]

OK the calibration routine is in the file TFT.C

 

Here's my proposed changes in BOLD

After each calibration action the screen will turn green (hopefully) and there is a pause for 1 second (I think 2 is irritating long).

At least... I think this should work. Now how do I compile this all into a working thing so I can try it on my Sdrive Max ?

 

unsigned int action_cal () {

int px1,px2,py1,py2,diff;

const unsigned int b = 20; //offset from border

unsigned int x = b;

unsigned int y = b;

unsigned char i;

TFT_fill(Black);

for(i = 0; i < 4; i++) {

Draw_H_Line(x-10,x+10,y,White);

Draw_V_Line(x,y-10,y+10,White);

while(isTouching()); //wait for release

while(!isTouching());

p = getRawPoint();

TFT_fill(Green);

delay(1000);

TFT_fill(Black);

switch (i) {

case 0:

px1 = p.x;

py1 = p.y;

x = tft.width-b;

break;

;;

case 1:

px2 = p.x;

py1 = (py1 + p.y) / 2; //middle

y = tft.heigth-b;

break;

;;

case 2:

px2 = (px2 + p.x) / 2; //middle

py2 = p.y;

x = b;

break;

;;

case 3:

px1 = (px1 + p.x) / 2; //middle

py2 = (py2 + p.y) / 2; //middle

}

//print_I(10,50,1,White,Black,p.x);

//print_I(40,50,1,White,Black,p.y);

}

[_The Doctor__]

Nice idea, shorten the delay further maybe. I'm impatient

 

Give it another look if nothing wrong then nothing prevents to modify and see what you get.

 

If you are happy with result put the code / suggestion up on the repository with note and make the request.

Edited September 28, 2018 by _The Doctor__

[Level42]

I have no idea how this whole Github works and I'm not too sure KBR wants any interference in his work.

 

I'd first need to know how to get this modded version in my Sdrive Max.

[Level42]

By the way, I had a look at the code that displays the filenames on the screen. Did you know it has an If/then to display the texts in green if it is a known working file extension instead of blue ?
This gives me the feeling that adding a similar if/then for files that start with ._ to NOT display a line won't be such a resources hog as some thought here.....

 

Adding this is a bit more complicated than the calibration change though.

[Level42]

Nice idea, shorten the delay further maybe. I'm impatient

 

If you are really impatient, the current calibration algorithm is the fastest possible.........

[_The Doctor__]

If you are really impatient, the current calibration algorithm is the fastest possible.........

lol that's for sure.

I decided to slow myself down and it caused the calibration issue some have wrestled with, normally I just tap tap, tap tap and it's done. If I carefully deliberately press and release it's liable to to get a double read and then oh no, I have to do the point to elsewhere on the screen to get a button problem.

 

either way it all works out in the end..

 

um the idea behind the open platforms is to share code and ideas, he can look at it, accept, reject, delete, ignore whatever you send.. but if it is a worry let someone already doing it submit for you or email it to him ,

Edited September 28, 2018 by _The Doctor__

[Level42]

I answered to the GitHub issue about this with my proposal. I think it would be best to have one guy behind the firmware because if Kbr doesn’t want to include my suggestion for the calibration process, I could make my version but will have to update it every time again when Kbr updates his firmware.

Besides that, it will clutter things.best to have one source.

[E474]

One problem I have just discovered is that it's quite easy to crack the screen when fitting it into the 3D printed case. Just thought I would post as a "heads up" for anyone else trying to do the same.

[+remowilliams]

One problem I have just discovered is that it's quite easy to crack the screen when fitting it into the 3D printed case. Just thought I would post as a "heads up" for anyone else trying to do the same.

 

That definitely should not be an issue...

[E474]

You would hope so, but I think there is enough minor variation between the sizes of the components, in this case the screen, and the 3D printed parts, that I always end up having to do a bit of fine tuning, usually with a Dremel. Unfortunately, getting one part to fit correctly means a lot of taking apart and putting back together of the case, and it was sometime during this that the screen got cracked. I still have to get the wiring into the SIO connector done, so I think I will leave the last bit until tomorrow.

 

Perhaps getting the screen seated snugly would have been a better start.

[+remowilliams]

You would hope so, but I think there is enough minor variation between the sizes of the components, in this case the screen, and the 3D printed parts, that I always end up having to do a bit of fine tuning, usually with a Dremel.

 

Having made quite a lot of cases and full units at this point I can tell you without any doubt that you should not have do to any fitting beyond a simple drop in and sliding the screen.

 

If you are having to tune/modify anything then you have an issue with your printer/output being dimensionally correct.