SpudFiles

Pity you have to launch such a large projectile, you could have had a solar panel going into a deep-cycle truck battery and fill using an automotive compressor.
Or link a tandem bicycle to a compressor pump!

got to thinking about Poland-spuds idea about a pneumatic trigger.
several problems with polands design = the QDV piston is acting against the spring pressure thus slowing the piston movement down.
then the problem of getting a good seal on the rod.
Came up with solution to both problems
THINK of a reverse cannon-the trigger that moves the QDV backwards would be started in motion by a smaller miniature air gun mechanism.
think of the trigger piston as your projectile with the end of the cylinder open thus the QDV doesn't have to act against any air pressure caused by the trigger piston moving or any spring pressure.
to arm the cannon, just push the trigger piston in until it stops. just like the rod with a golf ball on the end.
just have the rear cylinder long enough so the rod doesn't protrude out when triggering the cannon (could be dangerous.)
this trigger may be faster than the rod/QDV/golf ball setup?
[/img]

forgot to mention, I won the Ebay auction last night for a pressure transducer.
This way we can get a digital reading of chamber air pressure with out using a tire gauge. This will be much more accurate.
its a DSY-150. Talked to company and find this sensor sells retail for $200+
snagged it for $8.50 + $10 shipping. description says NEW, never used.
http://www.proportionair.com/index.php/ ... lypage.tpl

your idea is more less the same as mine... the only difference is that I would rather use an off the shelf cylinder to do the job and that mine is slightly simpler (but I am aware that yours offers consistent opening times)


I don't know how much experience you have with air cylinders but they are pretty fast and they can act with a lot of force... (some have enough to crush bones or break fingers)

at 15 bar even a 12mm bore air cylinder would have quite a lot of force to overcome the spring and open a large valve relatively fast... so I don't think that you should worry about the air cylinder


however, the biggest problem here is the valve between the chamber and the air cylinder... if it won't open in a consistent and fast way it might affect performance and consistency to some extent... since accuracy is quite important for you it might not the the best idea really


of course it's not something that you can't solve with an indirect acting manual spool/poppet valve... that would be much simpler than building a small air cannon to pressurise the air cylinder


but yeah now that I written it I realize that's pretty much what you wanted to achieve but I refined the idea for you

The idea is basically two air cannons back to back.
the rear trigger cylinder perhaps just a 1inch dia pvc long enough so nothing protrudes out the rear so it is a safer unit that respect.
the rear cylinder is triggered by small sprinkler valve so consistency should be no concern, I hope.
as far as experience=very little
the trigger valve should work just like a typical spud gun.
No O rings needed as it would add to weight and drag, thus slowing down the QDV action, I think??
Just thinking of ways to cut down weight and drag for fast actuation of the QDV.

well it's not that air cylinders require a lot experience to be used properly or something... the point was that normally people don't realize how powerful they can be
at 15 bar a 20mm bore air cylinder can lift almost 50 kg


my idea and iknowmy3tables' design was aiming at improving the QDV so that it would act both as the main valve and the spool valve at the same time...

so the gun would work just like a QEV/piston gun with a 3 way valve as a pilot valve but you wouldn't have to buy a 3 way valve for it because the QDV itself would partly act as a 3 way valve...
you would only need a simple valve that when triggered would retract the aircylinder and open the QDV

as I mentioned in my previous post the air cylinder probably won't be the problem here... however, if the trigger valve can be a problem as it has to allow full flow to the cylinder in order for it to retract fast enough...

I don't know if this will be a big problem or not... I know you are concerned with accuracy so you need a design that is consistent


I know it can be solved by using an indirect acting manual valve...
indirect acting valves have a small in-built valves in them that use air pressure to switch the main spool/poppet in them - so they switch fully and in a more consistent way

the whole point of this design, at least form my point of view, was the fact that it didn't need a 3 way valve

but with extra valves it won't be any better than a normal piston/QEV valve piloted by an indirect acting 3 way valve but probably it will be more difficult to build

so, while it would be cool if someone build this I think it you'd be better of with a more traditional design employing something as simple as a solenoid sprinkler valve as the pilot valve

EDIT
oh god dammit... I was so concentrated on writing this post to show off with my knowledge that I didn't notice that you want to use a sprinkler valve as the 'trigger' valve

yeah that would probably work all right... I think there is no point in adding a separate chamber for it.. you could hook it up straight to the main chamber of the gun

but still, as I mentioned earlier, it might be easier for you to build a simple piston valve piloted by a sprinkler valve

your right that I don't need a separate air chamber BUT then I rob some air from the main cannon thus decreasing my consistency.
If I use two separate chambers then the amount of air in the second chamber has no influence on consistency of the main cannon. As long as the air chamber in the main cannon is identical each time.
Using two separate chambers will assure me of this repeatability
using a sprinkler valve to fire the trigger piston to move the QDV rod eliminates any eratic movements due to jerking back on the QDV rod.
Is a spool valve the same as a QDV?
If the QDV opens at the same speed each time adds to accuracy.
at least thats my opnion
weather its right??

MrDEB wrote:forgot to mention, I won the Ebay auction last night for a pressure transducer.
This way we can get a digital reading of chamber air pressure with out using a tire gauge. This will be much more accurate.
its a DSY-150. Talked to company and find this sensor sells retail for $200+
snagged it for $8.50 + $10 shipping. description says NEW, never used.
http://www.proportionair.com/index.php/ ... lypage.tpl

Nice purchase!

Just need a power supply and a $10 digital panel meter and you've got a pressure gauge.

A bit more to calibrate it but it really probably doesn't need to be calibrated since reproducibility is more important than absolute accuracy.

The documentation says the response time is 0.1mS so it would even work to monitor the peak pressure in the chamber of a combustion gun.

well yeah... in theory it might seem it would be more consistent with two chambers but that increases complexity of the design...
you would have to make sure that the pressure is exactly the same each and every time in two chambers not just one.. and that increases the chance that something will go wrong

moreover, the amount of air used to power the air cylinder will be the same every time... so I don't think that can ruin consistency

spool valves are one kind of directional control valves.... valves which can switch the flow between two or more ports
have a lookhere

you might also readthis text about pneumatics

If the QDV opens at the same speed each time adds to accuracy

yes, consistency is what you should be aiming at

but i still think that it would be easier if you build a QEV/piston gun and pilot it with a sprinkler valve

you might as well use a sprinkler valve as your main valve... I am not an expert here as I've never used one myself (they are rare in europe) but I think they are quite consistent

Found the list of projectiles. When will you know what will you be launching?

Air cannons get real crazy at low pressure where the expansion volume is less than the volume of the barrel. This will cause problems in the near shots with accuracy. The large chamber can improve this but then the new volume adds more power to the short distance shots, so pressure control at lower pressure becomes critical if the projectile seals in the barrel and projectile friction becomes an issue with low pressure.

On that note I was playing more with my 2.5 inch barrel and the sabot shown on page 1. In testing with the small launcher (Marshmallow cannon) I finally got to find out the highest pressure I can use without ejecting the sabot. In the 3 foot barrel, I can go up to 40 PSI and still not launch the sabot out of the barrel. This sabot then becomes a variable length barrel for launching projectiles to match my supply pressure for throttlable power. I spent part of yesterday launching marshmallows (too lazy to go pick up golf balls) and found it was easy to drop a close grouping of marshmallows at 10 yards one right after another. I got a grouping of a half dozen marshmallows mostly within 5 feet of each other. (marshmallow aerodynamics still sux)

If I am reading Techs info right, if we use a sabot in the barrel for the short distance shots we will be more accurate?
on the subject of sabots, what material are you recommending?
could one be made of homemade marshmallows?
Fr the pressure transducer, going to connect to a micro-controller chip which is to be connected to an LCD display.
the display will indicate wind speed, wind direction (using the barrel pointed towards the target as reference), muzzle velocity and chamber pressure.
a lot going on but want to eliminate most of the variables..
using the data one can calculate how much air pressure and barrel to target deviation to achieve accuracy with every shot.
at least that's what were hopeing.
plan right now is to design circuit, design and code the PIC so all desired functions work. Start procurring parts for the cannon and start building.
come spring we hope to be able to start test firing (too cold right now (17degrees last night but 51 today.) boy was it cold out fishing today on the lake. just glade I got my deer already.
come spring we can start drawing up some data charts.

MrDEB wrote:If I am reading Techs info right, if we use a sabot in the barrel for the short distance shots we will be more accurate?

It is much more accurate. Tested it with golfballs, marshmallows, hard candy, lightsticks, and eggs.

on the subject of sabots, what material are you recommending?
could one be made of homemade marshmallows?

Marshmallows is highly not recommended. They smear in the barrel and make the inside behave like it was lined with duct tape.

Use a rigid sabot that will not change shape while being shot. Use one that is low friction and has little blow by space. Mine gently sinks in the barrel under gravity between shots. This provides good low pressure performance.

Fr the pressure transducer, going to connect to a micro-controller chip which is to be connected to an LCD display.
the display will indicate wind speed, wind direction (using the barrel pointed towards the target as reference), muzzle velocity and chamber pressure.
a lot going on but want to eliminate most of the variables..

A good start is to get the launcher to a high degree of repeatibility, then work on secondary variables.

using the data one can calculate how much air pressure and barrel to target deviation to achieve accuracy with every shot.
at least that's what were hopeing.

Graph your results. You will find most any launcher has enough varibilities, that simple math models have a deviation from your real world graph.

plan right now is to design circuit, design and code the PIC so all desired functions work. Start procurring parts for the cannon and start building.
come spring we hope to be able to start test firing (too cold right now (17degrees last night but 51 today.) boy was it cold out fishing today on the lake. just glade I got my deer already.
come spring we can start drawing up some data charts.

Leave room in your PIC for correction factors for each general range. A table with the mean distance with no wind is a good start for a baseline look up table. Your table will be non-linear.

the PIC will just display data such as wind speed, muzzle velocity etc.
no real cals as to what pressure to hit target at X distance.
this will all be drawen up using actual test firings/data retrieval then drawing up graphs using data that was aquired.
need to test on windy days as well as calm days etc.
plannng on using poland spuds idea of one tank for both the QDV and the trigger.
just don't like the idea of pulling back on a rod each time you want to fire the cannon.
too many added variables as well as possable injury due to the rod comming back too fast and hitting ??
have a sprinkler valve between the main air chamber and the trigger chamber.
just need to locate some material for the QDV piston.

MrDEB wrote: too many added variables as well as possable injury due to the rod comming back too fast and hitting ??

Need to study my video.. The rod only comes back as fast as you pull it. The piston moves faster, but does not make the rod move when it does.

The QDV can be triggered by a blast of pressure instead of the rod by using a sprinkler valve to apply pressure between the projectile and piston. The rod can be eliminated and it can be full electric triggering.

What video are you talking about
Haven'yt seen a QDV w/o a rod??