Trigger Circuit / Electromagnetic Projectile Accelerators / Forums

Forums

4hv.org :: Forums :: Electromagnetic Projectile Accelerators

« Previous topic | Next topic »

Trigger Circuit

1 2 next

Move Thread

LAN_403

NiG

Sun Mar 09 2008, 10:08PM

Registered Member #1372 Joined: Tue Mar 04 2008, 12:16AM
Location:
Posts: 6

Hi all

I'm an italian guy, and I'm currently working on a portable multi-stage coilgun.
My objective is to build a portable semi-automatic shotgun, with at least 5J of output cinetic energy with a 12grams projectile weight.

Obviusly I have many questions, but I wanna begin with a question about the stage sinchronization circuit.

I saw some guys (like

) use advanced solution, like PIC. Anyway, except from PIC, I didn't found any scheme on the net... well... frankly I didn't found much material on multi stage triggering.

The question is this: wich type of circuit do you use to correctly trigger the stages ?

My scheme is now composed by SCR that completely discharge the capacitors, and BJT, controlled by RC line, that power the SCR. I discarded the use of a 'delicate' astable multivibrator made of a 4001 and two RC lines because I noted that the EMI make it don't work.

With patience, and perfect adjusting of the R, it's possible to obtain good result thanks to good synchronization.

Actually I have an energy output of 2.75J with a source/output ratio of about 2.5%.

Thanks a lot,
bye

Shaun

Sun Mar 09 2008, 10:45PM

Registered Member #690 Joined: Tue May 08 2007, 03:47AM
Location: New Jersey, USA
Posts: 616

Welcome, NiG!

You need a way for the circuit to detect when the projectile is at the beginning of each coil; easiest way to do this is wire or foil brushes in series with the SCR gate and gate resistor.

You manually trigger the first SCR, then as the conductive projectile connects the two brushes at the beginning of the second coil, it fires the second SCR.

An alternative is using photodiodes/transistors and an LED so as the projectile enters each coil it blocks the light path for a moment, then a logic circuit detects this pulse and triggers the SCR.

Best of luck, and I have to say thats a very good efficiency for a manually timed circuit.

NiG

Mon Mar 10 2008, 12:37PM

Registered Member #1372 Joined: Tue Mar 04 2008, 12:16AM
Location:
Posts: 6

I know already all this methods, but I personally decided to choose a fixed timing triggering sequence.

I don't wanna use non-solid state switching, because I think they are not absolutely precise, and I think with meccanic switch you can fall into unwanted effects, like sparks, jumping contacts, and similar.

I also dislike the classic optical switch, because there is only one point where you have to put the receiver diode (and the transmitter) to obtain the best performance. At least, you must have the possibility to freely move the diode also for minimal distance (step), like one millimeter.

For all this reasons, I finally choosed the fixed timing triggering, because with one trimmer (variable resistor) per stage you can determine the best moment to trigger the coil.

Do someone also choose this method ?

If yes, what type of circuit do you used ?

---

I have also thought about a perfectly precise non-fixed triggering, where the trigger for the stage N is done in respect to the velocity of the projectile out of the stage N-1.
I don't wanna use a PIC, so the scheme I found involve three operational amplifier plus, obviously, other passive components... but this sound a little hard to make

Maybe I can choose this for my next project :D

BYE

teslacoolguy

Mon Mar 10 2008, 06:17PM

Registered Member #1107 Joined: Thu Nov 08 2007, 10:09PM
Location:
Posts: 792

you can wind a small coil right before of the work coil and as the projectile passes over it a small voltage will be produced triggering a transistor or something and have that trigger the scr.

Slid

Tue Mar 11 2008, 10:20AM

Registered Member #853 Joined: Thu Jun 21 2007, 03:08PM
Location:
Posts: 14

If you want to make a semiautomatic, all your projectiles must be exactly alike for the fixed timing sequence. When you get some velocity, small things as dust in your barrel or a slightly variable voltage in capacitor due to the natural voltage drop while waiting to fire, cap-age and cap-temperature can distrupt the fixed timing.
And if one of your components gets killed, you have to adjust the sequence again, especially since most capacitors have a huge % tolerance.

Why don't you use a stage optical trigger setup where you can delay each triggerpulse individual?

Bjørn

Tue Mar 11 2008, 11:05AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

Fixed timing and multiple stages is usually not a good idea. I don't recall anyone that tried it and ended up pleased with it in the end.

Tiny errors will be amplified in every stage until they are significant, change in temperature and other things you don't have control over can make a big difference. Even aging of the components will make it drift into an area of low efficiency.

The owner of the website you linked to has blocked linking from this website so your link does not work.

Do you have a special reason for not using a microcontroller?

NiG

Wed Mar 12 2008, 05:58PM

Registered Member #1372 Joined: Tue Mar 04 2008, 12:16AM
Location:
Posts: 6

Those are the main problems I noted with fixet timing:

1- The speed error introduced by the natural discharging of the caps when waiting to fire. But I resolve this thanks to a feedback with a comparator with a low hysteresis, that keep very stable the cap's voltage.

2- Dependency of components tollerance. Well, this is true, but in a good scheme, no components should break. So the first calibration should last.

3- Dependency of external variables, like temperature. This is also true, but the temperature vary very low in the main circuit. I think the difference with a jump of 20Â° is lower than 1%.

I don't use optical triggering because, like I said, there is only a single point where you should put the sensor to obtain the best performance.
Moreover, if something goes wrong in a previuous stage, the error (probably in speed) will propagate anyway in the stages... just because also che sensor is fixed

and it does't look ad the speed

Infact, I think we should say "open loop" (maybe "partial open loop") also with sensor in each stage.

---

Actually, I think that the only optimal solution is to use a mechanism that takes into account the speed in every stage.
This could be done with:
* PIC
* dedicated circuit (with "standard components")

I don't use the PIC because I don't know them, and I don't have the proper equipment

I designed a scheme suitable for this (without PIC), but for my first coilgun experience I think a fixed timing with a good main-circuit should suffice.

Bye !;-)

pbfree

Thu Mar 13 2008, 03:31AM

Registered Member #1386 Joined: Tue Mar 11 2008, 08:16AM
Location:
Posts: 13

Yes, the cap back should be regulated always. Optics will not overcame change in cap voltage.

Optics is open loop because it in no way is reactive. Unless by some form of magic, when a projectile enters a certain stage too slow the optics cannot speed up the projectile anyway hence calibration does propagate to remaining stages. You must keep track of projectile speed at each stage, compare the difference against a look-up table and then compensate. Compensation of course would be to increase the delay between projectile optical detection coil firing proportional to decreased projection entry velocity. Compensation would not included a decrease in coil firing delay after detecting an unexpected increase in projectile entry velocity because this would mean the coils simply were not calibrated in the first place.

Sirartanis

Thu Mar 13 2008, 04:54AM

Registered Member #1390 Joined: Thu Mar 13 2008, 04:46AM
Location:
Posts: 1

Hello, Iam currently building building a multi-stage coil gun using a PIC16F877 to trigger the multiple SCR that controle all the stage. The Idea here is to use some analogues or digital inputs to set the waiting time between the stages. This way u can experiment different timing and ajust it with dipswitch or sum potentiometer. If u want some code exemple i can show u if u want. its preaty easy using a picstar programer probably available at any college that teach electronic.

So for you now the best thing u can do is to do it with a pic and its not so hard trust me i can show you

Bjørn

Thu Mar 13 2008, 11:14AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

I don't use the PIC because I don't know them, and I don't have the proper equipment

That is not really a valid argument since there are people on this forum that managed to build their programmer and make their first useful assembly program in a single day.

The cost is insignificant too, even if you can't get chips for free as samplea they cost down to $0.5 each.

It is is mostly a matter of will.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob