New Data Pics in Data Thread

[AGD]

For your viewing enjoyment.

AGD

[Redkey]

On the behold the mighty angel plot....
the red line is the "forward end of bolt travel"... how is this measured? and what are the units?
Why is the lower portion of the plot labelled "Mag xtreme peak pressure"

On the plot labelled "The beloved cocker"
why is the top of the plot labelled 1/28/02 Angel?
why include the red line if it's not associated with anything?
Why is this plot also labelled as "Emag xtreme peak pressure" on the lower right?


On the Impulse plot...
in the plot area the dotted line is listed as ball out end of barrel. Below the plot area to the left the dotted line is labelled as Emag Hall. Below the plot to the right the dotted line is labelled as bolt end travel. This is very confusing as to what the dotted red line represents. Is the solid line the pressure? the Halo or the barrel end?
Why is this plot also labelled as "Emag xtreme peak pressure" on the lower right?

On the Matrix plot...
What is the red line measuring?

on the emag plot...
What are the units on the red line?
"bolt .110 open - 40"
".060 open - 60"
"barely open - 77"

what are the units with these numbers?


It looks like you have collected some interesting data. However, it would be much easier to understand if the labelling on the plots made it clear what we were looking at.

Sorry for being nit picky.

[Kweasi]

There's some very strange stuff in there.

Please clarify the meaning of the Red line?

And, tell he the vertical scale is not zero'ed!

-Kweasi

[AGD]

If the red line is flat then its not measuring anything. Most of these red line plots were with an infrared eye (same one used in the ACE) placed at various places to measure the timing of the bolt. In the angel plot you see the ball go by first then the bolt closes. The units are relative position on the time scale.


The software is setup so that it always reads two sensors so you always get a red line.

Disregard the text thats repetitive to each plot, I just dont do a good job of changing all that stuff. Ususally the text right in with the plots is the current info.

The time increments are specified at the top 100 counts = ?ms

Matrix red line is showing the close then open of the bolt.

Emag 77= .060 indicates that when the red line reaches 77 counts on the Y scale the bolt is .060 open. I was measuring the pressure buildup against how big a crack was left open by the bolt.

I think I answered everything.

AGD

[Redkey]

Thanks for the answers.

Don't the eyes just register yes or no? Or have you set it up to read the relative output voltage of the sensor... the more light reflected the higher the output signal? Probably not the most precise way to measure it... but if it were calibrated correctly it might work.

What you really need is some medium throw LVDTs. Not sure how easy it would be to hook one up to a mag or angel... but, it would be simple to hook one up to gun with an exposed bolt.

hmm.... interesting. Something new to play with.

[steveg]

About the Matrix It really should be pointed out that,
as with all real electro's the solenoid off and on times
are user setable.
the stock matrix on time ranges from 35ms to 70ms
and the aardvark bolt kit ranges from 10ms to 35ms

stock off time 25ms to 55ms
aardvark 35ms to 45ms

most aardvark bolt users use 45ms off and 14ms on

even Gen E (former airtech) uses the aardvark bolt now.

[AGD]

Redkey,

The input is analog so it does reflect a changing imput to the sensor. It just shows the stuff going by in a rough way we don't try and calibrate it or anything.

We do have LVDT's but they add weight and complexity to hook up to internal parts. We use them in certain apps.

AGD

[Redkey]

So you start adding additional sensing hardware you run into the problem where the measurement system influences the system that is being measured. Something to be avoided.

[Vegeta]

Tom, one question for you:

How do you do it? Literally. How do you know what everything means and how to interpret what the sensors tell you and how do you figure out what is good and what is bad. it is obvious you are playing with some very technical data in the dyno. Do you just learn as you go or you have some sort of fluid mechanics engineer 800 number you call for help? I mean.. when you get stuck on something or don't know something.. who would you ask? There is not liek a book about this stuff.. and not very many classes that indulge into this.....

Did you take courses on the stuff or just learn through trials?

/me is very interested in how you went about it...

[Redkey]

Were it me... I would make damn sure that my measurement system was accurate and repeatable. Then I would take multiple test runs with each configuration... the multiple test runs tell you how much variablity there is in the system you are measuring. See toms desciption of basic statistics, it's a good starting point for understanding how to work with the numbers. Once you know you have good data then you start making educated guesses as to what the data means. If you think you understand the numbers/system... create a test situation to verify your assumptions. If you get stuck or don't understand something... keep running tests and look for trends in the data until it becomes clear. Sometimes you figure it out... other times you'll just have to accept the fact there is no clear answer. Probably one of the biggest things is to determine the significance of your findings... is your data telling you anything useful?

As for education? The following classes would help

English/Technical Writing... if you cannot tell people about your test process or results in a clear precise manner it does not matter how significant your findings are.

Robotics... Lots of useful information about pneumatics and electronic control.

A class on data collection. You'll learn all about pressure transducers, load cells, LVDTs, optical sensors, strain gauges, extensometers... etc. Plus you'll learn about signal conditioning, filtering and calibration.

A class on electronics. This comes in handy when designing your latest electrical marker or coming up with a new test system.

Classes on statistics... you'll learn how to work with the numbers and design meaningful experiments. You should also learn how to use Minitab, a incredibly powerful statistical analysis software package.

A class or two on thermodynamics and fluid flow. You'll learn about flow paths, velocity, friction. All the nifty things that play a small roll in the operation of a pb marker.

A class on materials science. You'll learn about aluminum, steel, plastics and composites and where they should and shouldn't be used.

Take all these classes plus a few extras and you'll end up with an engineering degree. Not only will you understand how things work... you'll also be able to find a job that pays well. Plus, all the hot chicks love engineers, or so I have been told.

The other way of doing it would be to have years and years of experience of designing markers and trying things that sometimes worked and other times didn't. After a while you would have a pretty good feel for what worked and what didn't. Sometimes this can give you a much more realistic view of what is important when it comes to testing.


Just my thoughts. Sorry for being rather long winded

[AGD]

Vegeta,

Graphs always look confusing to the people who didn't design the experiment. What Redkey said is all true knowledge comes in layers, build up enough of it and you wonder how the stack got so high.

In my case I have been doing this type of thing all my life working with electronics and mechanics. While I do not have a formal education in this stuff I can go out and learn what I need to accomplish things. I have built industrial robots, CNC controlers, inumerable industrial machines and they all function on the same principles and techniques. After a while you just put the pieces together into what looks from the outside to be a nightmare setup.

Have no fear everthing is under control!

AGD

[Vegeta]

Thanks alot. I have been wondering how you learnt the trade, weather through classes or through experience.

Just think, years and years from now, Tom Kaye will be that old techhie guy sitting at those big new fancy smanchie tourneys saying (old man vioce) "Yea I 'member when I designed my first marker... it went on to shoot 20 balls a second! oh my! and we liked it! we LOOOOVED it! and there wasn't a damn thing you could do about it!"

[Rusty]

I can remember the first time I seen Tom. It was I bet 11 years ago at Paintball SAms in Wisconsin. I believe is was our 2nd big game. Tom had brung with a proto type he had made. It was called the Panther. it started out pretty good but he started to have troubles with it going full auto. I knew this guy had a good thing going back then. I was giving him any thing he needed to help him out. If I remember right, the gun had broke paint and just needed to be cleaned out around the bolt. And he was shoving a stick down the barrel to recock the bolt. Talk about a man in panic. He got it working and palyed the rest of the game. The rest is history.
Tom do you still have the Panther?
jimb