E/X-Mag Microcontroller Programming (Atmel AT90S2313)

**LorneCash** · 02-02-2006, 06:49 PM

Originally Posted by 11 Bravo

Do you have an address where we can see the programmer that we are waiting on?

Never mind. Is this it?http://www.atmel.com/dyn/products/to...+&tool_id=3808

Yea that's what we're all waiting for.

**LorneCash** · 02-06-2006, 11:52 PM

I just got a set of break beam eyes from Destructive Customs (thanks guys) and will be working on making my software compatible with break beam eyes now too. I'll keep updating this thread with my progress.

I have to find time to work on the software which i haven't since early Jan. I've been busy with other stuff. I have to rewrite one of the subroutines to free up some space before anything else and that will take me several hours to do at least. It's one of those things i did the easy way and now I have to go back and tweak it to make it take less space.

The latest news on the programmers is the first 10 programmers should ship from the distributer on 2-28-06. The rest should come a little over a month later when they are readily available.

I just got the ATSTK500 today and have not had the chance to test it yet. I should be able to do this tomorrow after work.

**VK700** · 02-07-2006, 10:54 PM

Lorne I've been following this thread since it started. Man my hat is off to you. Decided to try my hand at writing my own code for the Pandora board in my Excalibur. It's taken me 2 weeks to finally get all the pin outs done.

I've reread the first 4 pages of this thread numerous times and Miscue's leaky bucket finally makes sense. Sorry Miscue but until you try to actually program one of these processors the logic seems bizarre. I believe Miscue's discussion of EEPROM bits being corrupted explains an issue many have had with the Pandora board mysteriously reloading factory defaults.

Had to go back to my old electronic notes to figure out why I/O pins were always high at boot up. I spent 2 or 3 days trying to figure that one out.

Lorne how did you ever figure out the ROF timing issue? I'm trying to write my main firing routines ( no I haven't even ventured into the EEPROM but that looks fairly easy). Oh for added pleasure my chip is an AT89S8252. Which is an 8502 derivative and I believe yours is the newer AVR type. I'm just looking for a logic path. My best guess is a cyclic buffer?

One thing I have found on the break beam eyes. You need to pulse them or they will get hot. I've just been slowly bumping up the duty cycle until I get a firm 1 to 0 transistion on the eye receive pin.

Thanks for any help or comments you can provide.

**Miscue** · 02-08-2006, 01:46 AM

Originally Posted by VK700

Lorne I've been following this thread since it started. Man my hat is off to you. Decided to try my hand at writing my own code for the Pandora board in my Excalibur. It's taken me 2 weeks to finally get all the pin outs done.

I've reread the first 4 pages of this thread numerous times and Miscue's leaky bucket finally makes sense. Sorry Miscue but until you try to actually program one of these processors the logic seems bizarre. I believe Miscue's discussion of EEPROM bits being corrupted explains an issue many have had with the Pandora board mysteriously reloading factory defaults.

Yeah, it isn't very intuitive. Check the trigger, check the ACE, pull the solenoid, wait for the timer... release the solenoid... wait for the timer.... etc... in a chronological order, makes a lot more sense, but the code gets unnecessarily big and clumsy.

The way I suggested, you can have "fire the gun" part of the code appear before you even checked the trigger... because it doesn't matter what order it is in chronologically in the code. The logic of how it fires is not obvious because it's built into a sneaky way to set your timers.

You don't need your code to determine: A can't happen until B, which can't happen until C... and build in the chronology. If you set timer A to 30, B to 20, C to 10... A can't happen until B which can't happen until C... all by itself... without having to code explicit logic for this. Expressing this with ASM works out pretty well... a simple, small loop handles all the timers... the rest of the code is very small and simple. It just isn't straight forward how it works when you look at the code.

I'm not sure if it's the best way... I'm sure there is something better... but I think it works out well.

I spent a great deal of time working on getting the EEPROM stuff to be bullet-proof... putting the gun in low-battery conditions... and just doing stuff to try to induce EEPROM corruption so I could get rid of those rare/intermittant problems. I was kind of irritated that the microcontroller wasn't more resilient with this stuff... more went into my EEPROM code than what I wrote to shoot the gun. I probably disconnected the battery and stuff over 10,000 times working on this...

The suggestions that I mentioned already I think sum up how to approach it. The solutions seem simple and obvious, but only after discovery.

**LorneCash** · 02-08-2006, 11:10 PM

Originally Posted by LorneCash

Ok I've ordered an Atmel ATSTK500 from Digikey to test to make sure they will work. As soon as I verify that it does work I will make a post so that anyone who does not want to wait for the AVRISP mkII can get this for an additional $50.

OK guys I got the ATSTK500 to work, so anyone who wants to can order one from me for $150 (or just $50 additional if you already paid for an AVRISP mkII ). Unless a bunch of people do this i'll just email a seperate addendum to the manual only the people who order one. I think this is a better option than permanantly adding it since once the AVRISP's are available I don't really see many people spending the extra $50 on this. Click on the links to see a picture of what each device looks like.

I'm sorry It took me an extra day, I had to call tech support to get this thing to work but all I was doing wrong was that I had to hook the ATSTK500 up to a 9V battery to get it to work rather than using the gun's battery like the AVRISP.

**LorneCash** · 02-09-2006, 08:00 AM

AVRISP mkII's have shipped from DigiKey

that's right the company that wasn't supposed to ship them until April (DigiKey) has just shipped me 10 units. I have a tracking number . I still have 10 units backordered with Digikey and am also still expecting 10 units from AllAmerican by the end of the month but if digikey has parts now maybe the rest will come sooner too!!!

They first 10 are scheduled for delivery to my house tomorrow If that is so I will make sure and have them all shipped by monday. I am working all day friday and just won't have time.

**BigEvil** · 02-09-2006, 10:16 AM

Good news!!. Thanks for the update. Im sure I can speak for everyone interested in this when I say that it is greatly appreciated.

Also, any new updates to the software or manual? You are probably going to get a bunch of people asking questions once they get their programers. (Me included)

**LorneCash** · 02-09-2006, 11:01 AM

I've made a minor update to the manual but I'm waiting until I get A programmer to release it because then I can include pictures and screen shots of how to do the installation.

I've been working on a new version of the software this past week and i think I've almost got it. The change will be transparent to the user but it will save 10% of the memory space. It has to do with how i display numbers in the menu. Once this is done it will open up a few more possibilities for things I can add. I probably won't release it until I make a change that will affect users since there is no point otherwise.

I really don't know of any bugs left to fix. I really need some people to use this and let me know if they find anything.

**11 Bravo** · 02-09-2006, 11:16 AM

OH YEAH! Its time for the dancing bananas

**Rudz** · 02-09-2006, 12:35 PM

yeah this does deserve dancing bananas!!!

uhhh yeah get down ... so the first ten are coming..i hope the next 10 are soon to follow..i think i got number 11..awww thats what i get for being a procrastinator...

**Canuckmag** · 02-09-2006, 12:50 PM

I gotta tell ya... I'm pretty pumped.
I'm not sure where I am on the list, maybe 13. It would be great if you could compile a list to show what order the programmers will be shipped (first 10units, second 10 units, ect). Mostly for my own curiosity, I have the patience of a 6 year old.

Jeremy

**Naash** · 02-09-2006, 01:40 PM

That's good news !

Now, let emags rule the paintball word !

**Rudz** · 02-09-2006, 01:45 PM

i got guys at sc village waiting to see this rip faces off.. they dont believe me..heheeh..ooh yeah?? first you question my tunamax emag..asking if its a " real" mag?? lol then you call her old??? lol.. ok..ill let her do the talkin fer mehh!!!!

**hitech** · 02-09-2006, 02:26 PM

Originally Posted by BigEvil

Im sure I can speak for everyone interested in this when I say that it is greatly appreciated.

I'm sure you do. I certainly do.

**Jotsy** · 02-09-2006, 03:01 PM

umm, what if we actually want the atk500? how long will that take to get to us?

**LorneCash** · 02-09-2006, 04:16 PM

Originally Posted by Jotsy

umm, what if we actually want the atk500? how long will that take to get to us?

Only a few extra days, DigiKey has those in stock. If anyone else would rather have the ATSTK500 send me an email at pniedfeldt@gmail.com

**White_Noise** · 02-09-2006, 09:32 PM

Originally Posted by LorneCash

I really don't know of any bugs left to fix. I really need some people to use this and let me know if they find anything.

hey,

im playing in a tourney at Atomix in Michigan on saturday, i should get plenty of time to play with the mag and let you know how she shoots. ive done dry fire testing and my roommate cant believe how fast it is.

**LorneCash** · 02-10-2006, 12:32 AM

Here's an Interesting fact that I never thought of including before...

My code checks for a trigger pull 1,333,333.33 times per second. Does anyone know how that compares to other boards/software?

I read on some website (that I can't seem to find again) where some board claimed to be the fastest at 1 million checks per second and it got me thinking... so I calculated what mine was and there it is, I win!!! I check It once every 3 cycles on a 4MHz chip so all I did was take 4,000,000/3 and got 1.3 million.

**BigEvil** · 02-10-2006, 08:11 AM

Originally Posted by LorneCash

Here's an Interesting fact that I never thought of including before...

My code checks for a trigger pull 1,333,333.33 times per second. Does anyone know how that compares to other boards/software?

I read on some website (that I can't seem to find again) where some board claimed to be the fastest at 1 million checks per second and it got me thinking... so I calculated what mine was and there it is, I win!!! I check It once every 3 cycles on a 4MHz chip so all I did was take 4,000,000/3 and got 1.3 million.

Lorne gets to go to the head of the class.

You know, you could make a bundle if someone made a custom board with lots of room for you to put code on.

**sean33** · 02-10-2006, 09:05 AM

i PM you....

**11 Bravo** · 02-10-2006, 12:11 PM

Originally Posted by BigEvil

Lorne gets to go to the head of the class.

You know, you could make a bundle if someone made a custom board with lots of room for you to put code on.

When firing it in semi can you tell a difference between it and 3.2?

**BigEvil** · 02-10-2006, 12:52 PM

Originally Posted by 11 Bravo

When firing it in semi can you tell a difference between it and 3.2?

I'll let you know for sure once I get it on the gun. Im waiting on the programer.

There is definately a difference between 3.2 and the 4.xx software, and its pretty noticable. So Im betting there is a big differnce between the Xmod and 3.2

**11 Bravo** · 02-10-2006, 12:58 PM

I am waiting to. I'll race you to see who gets it loaded faster.

**LorneCash** · 02-10-2006, 01:57 PM

Originally Posted by 11 Bravo

When firing it in semi can you tell a difference between it and 3.2?

In general you probably won't notice a difference in SEMI between XMOD, 3.2, and 4.x assuming that the dwell in 4.x is set to 10.

Here's an another interesting bit of information... On my guns (3 X-Mags), with FIX at 0, depending on how the solenoid is wired, either SEMI or RELS has a full auto bug but never both. This means that you could use no FIX (Shot buffering window at max) and not have any full auto bug problems. Assuming this is the same for everyone it would seem faster than 3.2 but really it depends on what the FIX (shot buffering) is set to. I can pull 16bps walking the trigger and 24bps raking the trigger with my software in SEMI or RELS.

**11 Bravo** · 02-10-2006, 02:41 PM

OOOoooohhhwww! They have been delivered... now I am doing the Martin Short Saturday night live dance.......

**LorneCash** · 02-10-2006, 02:53 PM

Originally Posted by 11 Bravo

OOOoooohhhwww! They have been delivered... now I am doing the Martin Short Saturday night live dance.......

You are SICK! You made this post before I even knew that they were delivered and I told my brother to send me a text message as soon as they got to the house.

**Beemer** · 02-10-2006, 02:59 PM

He must have set automatic refresh on that tracking number

**Rudz** · 02-10-2006, 05:18 PM

he camped out in front of lornecashes house...he cheated..

**11 Bravo** · 02-10-2006, 06:17 PM

I am watching you all...

**Rudz** · 02-11-2006, 12:07 AM

omg omg!!!! mommy..theres man watching me..again!!!!!

Thread: E/X-Mag Microcontroller Programming (Atmel AT90S2313)

Thread Tools

Display

yipppeee

Whoo hooo

ooh yeahh

nope..

ooooh

Posting Permissions