3.0 Software question: Shot buffering = bps increase?!

[DiRTyBuNNy]

and no..I'm not going to ask when...but what I did want to ask you TK is if there is going to be any noticeable speed differential between those E(and X)mags with the 3.0 software and those that don't because of the shot buffering...is it going to be easier to shoot 1bps faster(for example if someone can normally shoot 12bps with previous software will they know be able to hit 13, 14, or 15 with the 3.0?) I know that shot buffering has done a lot for other guns. I know that I'm in the current Xmag batch at ano but I've been torn because I've tried another gun and the plus with this other marker is that fact that not only can it shoot fast but because of the marker's software it seems to be easier than other markers to shoot faster (it's easy to demonstrate but hard to explain). I was just hoping to get your input on this question..thanks..

[Muzikman]

It should because the system will count the pulls that are otherwise missed now. If you can pull the trigger 3 times during one cycle, currently it will shoot 1 ball and miss the other two pulls. So with shot buffereing, if you pull the trigger 3 times during one cycle, it will buffer the other two shots and wait for the gun to finish the cycle and then fire the next ball.

Think of it as the keyboad buffer on a PC. If you have a slow PC you can type faster than the computer can put the characters on the screen, but it bufferes each key stoke so that keys you press are not missed. People who used BBS's on the old 1200 baud modems will remember this.

So I guess to answer your question, some people will have increased ROF while others wouldn't notice a difference. It's all in how fast you pull the trigger.

[pito189]

I wonder if 3.0 will be ready by World Cup? Doesn't that start like October 15th or somewhere in there.

[paintballbeaver]

close to the 8-9th of oct. pretty sure

[RRfireblade]

WAS "claims " that his board only buffers +one shot and you have to pull,I believe,faster than 22bps between 2 shots in order for it to buffer them/it.However,there is no way to verify that with out looking at the code (he ain't gonna show anyone)and he will still argue that his boards will increase "your" ROF.If that was true you'd have to pull 23bps and then you wouldn't need a WAS board to do it.(Other than a cap above that).So, I think he's full of it(big surprise),'cause every WAS boarded gun I've shot/owned has been crazy fast,and I don't normally pull 23bps,20-21 tops.:D

;) JAY.

[AGD]

3.0 is in field testing now. It is faster when you walk the trigger but seems to be limited by the pull time of the solenoid. We are trying to see if we can look for trigger pulls inside the solenoid window.

Shot buffering sees its best performance when walking triggers. As the software is right now, if you set the BPS to 11 and use your finger to rattle the trigger like a "roller trigger", the shots come out in a steady stream at 11 bps. At 14-15 you start missing intermitent shots and thats what we are working on.

AGD

[pito189]

Originally posted by AGD
At 14-15 you start missing intermitent shots and thats what we are working on.

AGD

I personally hate missing shots when I'm shooting 15bps. You definately need to work on that. :p ;)

[Jack & Coke]

Originally posted by AGD


3.0 is in field testing now. It is faster when you walk the trigger...

...Shot buffering sees its best performance when walking triggers...

AGD

Woo hoo!!! :)http://www.automags.org/forums/images/icons/icon14.gif


Originally posted by AGD


As the software is right now, if you set the BPS to 11 and use your finger to rattle the trigger like a "roller trigger", the shots come out in a steady stream at 11 bps. At 14-15 you start missing intermitent shots and thats what we are working on.

AGD

Interesting... I have 2.4 and I have noticed a few "skipped shots" as well.

I attributed this to my inexperience with shooting and setting up the emag trigger system (fine tunning). I thought, maybe I was just used to my GZ Timmy and had to "get used to" the emag style trigger.

Are "skipped" shots a normal "bug" with emags?

If I walk the trigger slow, it's fine. But if I try to go balls out, I get skipped shots every now and then and my rhythm gets messed up. Unlike my GZ where I can keep a consistant walking rhythm at high speed.

So far, from my limited experience with the emag trigger system:

Walking the trigger as fast as I can on the GZ (with WAS Equalizer) is like singing.

Walking the trigger as fast as I can on the emag trigger system is like singing while you have the hicups. (Still FAST, but not continuous)

I hope 3.0 can address this issue...

[Muzikman]

Jack, Yep...that's what 3.0 should fix...

[RRfireblade]

Originally posted by AGD
3.0 is in field testing now. It is faster when you walk the trigger but seems to be limited by the pull time of the solenoid. We are trying to see if we can look for trigger pulls inside the solenoid window.

Shot buffering sees its best performance when walking triggers. As the software is right now, if you set the BPS to 11 and use your finger to rattle the trigger like a "roller trigger", the shots come out in a steady stream at 11 bps. At 14-15 you start missing intermitent shots and thats what we are working on.

AGD

At 14-15pulls/per/second it skips if set at 11bps or if set to 14-15bps,it skips at 14-15 pulls/per/second.? (if that makes sense)

I hope it's ready by World Cup.

Jay.

[yagrmiestr]

wouldn't it be more likely to skip at a lower bps because the cycle time is longer (talking non-buffered code)??

[paintballbeaver]

thats what the genral thought was but it varies from gun to gun shooter to shooter.

[DiRTyBuNNy]

Originally posted by Jack & Coke

If I walk the trigger slow, it's fine. But if I try to go balls out, I get skipped shots every now and then and my rhythm gets messed up. Unlike my GZ where I can keep a consistant walking rhythm at high speed.


you got it J&C...that's exactly what I was talking about...exactly..

[DiRTyBuNNy]

nobody else cares to comment? well..I'm wondering about this walking 14/15bps comment..is this with the marker set to it's highest bps with the ace on (with a halo you shouldn't be getting skipped shots due to feeding so it's either your finger or the gun)..

[nippinout]

I'm actually trying to come up with some shot buffering code, but I am having trouble.

The only solution my friend and I could think of doesn't work. Run a loop, but this screws up the solenoid on time.

yagrmiester, any suggestions? I know you've got some tricky code. ;)

[yagrmiestr]

I've got something I've been playing with that uses counters and interrupts to control the firing sequence while still being able to continuously poll the trigger. It's part of something larger I am experimenting with. ;)

Edit: A little more info....
Basically how it works in "psuedo code"

If trigger pull then
energize solenoid
start Son time counter
go back to main loop (watch trigger/input loop)
End if

when the counter reaches what ever Son time you want it will set an interrupt, then you shut off the solenoid and reset the counter for the valve recharge time. when that counter triggers the interrupt flag you can fire another shot. The two counter times are based on what bps you have set. this way you can keep polling the trigger during the complete cycle and just store number of pulls in a variable.

FYI I don't have this running on a stock board, I'm using a little larger micro that I'm trying some different stuff on.

[DiRTyBuNNy]

SWEET!

[nippinout]

Originally posted by yagrmiestr
I've got something I've been playing with that uses counters and interrupts to control the firing sequence while still being able to continuously poll the trigger. It's part of something larger I am experimenting with. ;)

Anything that can help drive me in the right direction?

[yagrmiestr]

oh, you don't really have to use interrupts you can just read the counter value and when it exceeds your desired time you jump out of you main loop and set/reset the solenoid, etc. There are a few ways you could do this. Also I would not poll the menu buttons while in a fire or trigger poll loop, writing the display is a long enough event to effect the firing timing enough (at least in my code it is).

[yagrmiestr]

Originally posted by nippinout


Anything that can help drive me in the right direction?

I posted more info above... but the actual code is still in progress (when I find time... works been busy of late :( )

[nippinout]

Originally posted by yagrmiestr
oh, you don't really have to use interrupts you can just read the counter value and when it exceeds your desired time you jump out of you main loop and set/reset the solenoid, etc. There are a few ways you could do this. Also I would not poll the menu buttons while in a fire or trigger poll loop, writing the display is a long enough event to effect the firing timing enough (at least in my code it is).

Let's just say that went way over my head.

I don't know how to deal with trigger events during the solenoid on time.

Do I set the pin out to the solenoid to high and run a counter and include a check on the trigger within the count?

[yagrmiestr]

Originally posted by nippinout


Let's just say that went way over my head.

I don't know how to deal with trigger events during the solenoid on time.

Do I set the pin out to the solenoid to high and run a counter and include a check on the trigger within the count?

solenoid high and start the counter then go back to your main loop, or one that polls the trigger and watches the counter. Make sense? Sounds like you got the general idea. having a large enough counter or setting it up for the right resolution to get the right times and accuracy is what is important. I haven't played with stock micro to see how many bytes deep the counter is.

Edit: Looking at the datasheet it only has a 16 bit counter. With 1024 prescale you can time about 65ms in 256 steps of ~250us. It might not quite be long enough at 11bps and under this might not quite be long enough for the Soff time, but that could be worked around.

[Steelrat]

Yagrmeistr-Hey, he even plays paintball pretty well, too.

His skills with the emag board are phenomenal. This guy could be the next Jim Drew ;)

[nippinout]

I don't have an oscilliscope. :/

[yagrmiestr]

Originally posted by nippinout
I don't have an oscilliscope. :/

You should have seen the looks I got when I had my emag board hooked up to the scope in the lab, lol. And no I didn't try and get the whole marker past security ;)

[speedyfox666]

I'm still a little new on the software thing

but how can I get the new software on my board ? (when it 's out)

you need a data cable, new board,...

[pito189]

Originally posted by speedyfox666
I'm still a little new on the software thing

but how can I get the new software on my board ? (when it 's out)

you need a data cable, new board,...

AGD flashes your board. You can either go to an event where they are at. Our send it to them, and they will do it for you.

[Bolter]

We have it on our X-mags (it arrived at John Sosta's last Thursday) and my god it is a lot quicker!!! I mean I am pretty fast on the trigger probably 15 ish bps, but this thing is cool as you like. Its like whatever you put the gun on, say 20bps, it will shoot 20bps constant (as long as you walk it) until you let go of the trigger.

I have never shot a gun like it. Get it. Get it NOW!!!!!

:D :D :D :D

[Skoad]

i notice mine skipping shots but i just thought it was because my fingers were being retarded.

hope it comes out soon.

[Jack & Coke]

Originally posted by Bolter
We have it on our X-mags (it arrived at John Sosta's last Thursday) and my god it is a lot quicker!!! I mean I am pretty fast on the trigger probably 15 ish bps, but this thing is cool as you like. Its like whatever you put the gun on, say 20bps, it will shoot 20bps constant (as long as you walk it) until you let go of the trigger.

I have never shot a gun like it. Get it. Get it NOW!!!!!

:D :D :D :D

Spoooooooon!:D

[Bolter]

I'm sure it'll be available for all soon enough!!

Man its quick......

[joker4hire]

Originally posted by Bolter
We have it on our X-mags (it arrived at John Sosta's last Thursday) and my god it is a lot quicker!!! I mean I am pretty fast on the trigger probably 15 ish bps, but this thing is cool as you like. Its like whatever you put the gun on, say 20bps, it will shoot 20bps constant (as long as you walk it) until you let go of the trigger.

I have never shot a gun like it. Get it. Get it NOW!!!!!

:D :D :D :D

I think this is why AGD has such a "cult" following. They continue to improve their equipment WITHOUT having to buy a whole new marker.... well.. if what Bolter says is true.

I think I am going to try and find an emag for sale :)

[Miscue]

/me wonders if it's the higher bps cap more so than the shot buffering that is getting these results...

Anyone compare an XMag at same bps cap to the new 3.0?

[yagrmiestr]

Originally posted by Miscue
/me wonders if it's the higher bps cap more so than the shot buffering that is getting these results...

Anyone compare an XMag at same bps cap to the new 3.0?

I have my emag set at 20bps and I think it is part of it. Doesn't seem to exhibit as many (actually if any) skipped shots as everyone is talking about. At 20bps the marker cycles in 50ms, at 16bps it's about 63ms. So if you can release and pull the trigger in less time than that you will miss a shot. I'd like to get a part number on the HES sensor so I can see how much hysterisis it has. I wonder if going to a linear HES sensor and doing debounce in software would make the trigger more sensitive.

[Jack & Coke]

Of all the guns considered "fastest" and "easiest" to walk, the Matrix and The Timmy use the same type of mechanical micro switch.

Are micros switches better than HE sensors for this particular type of application (i.e. activating paintball guns at HIGH ROF) ?

Although HE sensors do not wear out over time due to physical friction, are they as fast as physical micro switches with regard to on/off cycling times?

I have no idea... just throwing the question out there... :)

[Evil Bob]

I have to agree with Miscue on this, in both the 1.37 (emag) and the 2.4 (xmag) firmware, setting the top BPS seems to affect this.

On my emag, I can hit the software limit of 16 walking, but I need to concentrate on keeping a steady rhythm at a specific rate or I get skips in the beat. Fortunately, I've been shooting it long enough for it to not really be an issue. On my son's xmag when set to 24 bps, I can walk the crap out of the trigger easily without skipping a beat, it doesn't start skipping until I set the bps below 16 and fire at my normal rhythm, at which point the skipping appears. That sounds like it's pretty clear what is happening, when setting it to a lower shot rate we're losing the odd shot by catching the circuit mid cycle.

I believe the shot buffering will cure this problem.

-Evil Bob

[Miscue]

I have a question...

What problem does shot-buffering fix when your bps cap is set really high?

I know that the idea is... you can pull the trigger twice within one shot window. Now... how small is this window? 25ms? Because, if it's a small value like 25ms... that would mean you would need to shoot two shots at 40bps.

How often can you shoot at 40bps, or whatever x bps it happens to be, in one second? Sustained across several seconds? You'd have to do it quite frequently to produce a noticeable ROF increase. I have a hard time understanding the feasibility of this.

I'm just trying to have a clearer understanding of what problem shot-buffering is supposed to fix... and have to wonder if the best thing about 3.0 is the higher bps cap. I have a feeling that the higher you set the max bps, the less effect buffering has... until it eventually becomes moot once you pass some threshold.

The chip is 10MIPS/MHz afterall...

[Jack & Coke]

Shot buffering = no more skipped shots

While walking the trigger, it is EASY have a situation where you've released the trigger with one finger, and have the second finger strike the trigger at just the time in order to produce a 40 bps rate (back-to-back shots).

WITH A HIGH ROF + NO BUFFERING

The gun WILL NOT shoot the second shot and will result in a SKIPPED shot. Since I've tested and shown the max bolt cycle speed of the X-Valve to be around 32-35 cps, I think the gun will not be ready to fire and the window will be missed.

WITH ANY ROF + BUFFERING

The gun WILL shoot the second shot and will NOT result in a skipped shot.

[Miscue]

I've heard this explanation several times... and was aware of it before I put up my last post, but I do not know whether to throw in on top of the stack of paintball myths, or if it can be verified.

Has anyone actually recorded in some way, the physical act of pulling a trigger, letting it go, and pulling it again at ~32 bps? Now, how often can they do this per second? How often can they do this for a particular duration? How many people can do this?

Also, does anyone know how long it takes for a trigger to return on it's own once released?

[Bolter]

I don't know about all this technical stuff, but believe me it is very quick. I was very dubious before installation of the software (the day b4 a tournie!!) but it seemed like after the software was added, it was only then that I realised how many shots I must have been missing before.

[Jack & Coke]

Originally posted by Bolter


...it seemed like after the software was added, it was only then that I realised how many shots I must have been missing before...



WOW!

Great feed back!

This is certainly encouraging for those of us puzzled at why it's difficult to shoot the EMAG as fast consistantly as a Timmy or Matrix.

[FreshmanBob]

The old nelson boards for the viking had the same problem. They were capped at 13 but before the equalizer boards I remember barely being able to hit 6 or 7 consistantly. Just for old times sakes I'll throw in my explanation.

In the 2.4 e mag software and nelson boarded vikings there is a certain time after you pull the trigger in which new pulls will be lost.

For example, say your gun is set to 10 bps. You pull the trigger once, then once more but there is a gap between the pulls less than 1/10th of a second (100 ms). Lets say you pull in 1/11th of a second (90 ms) The board doesn't let the pull fire another shot because that would be a higher ROF than 10 bps. That trigger pull is ignored so you have to pull the trigger again. Lets say there is another gap of 90 ms between the second and third pulls. That means the total gap between actual shots was 180 ms instead of 100 ms. That makes your total ROF much lower (5.55~ bps to be exact).

With shot buffering that trigger pull would be stored instead of ignored. That way, after 100 ms was up the gun would immediately cycle. If your pulling 15 a second then the gun will shot 10 a second, saving trigger pulls instead of ignoring them all.

So yes, shot buffering will make a big difference.