New to Coilguns - Coilgun Buildlog (Charger apparently fixed) / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

New to Coilguns - Coilgun Buildlog (Charger apparently fixed)

first previous 9 10 11 12 next last

Move Thread

LAN_403

Artikbot

Fri Oct 08 2010, 10:11PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Yeah, now that you say it, I plugged the wrong value! 1pF is not true, I'm actually using 1nF LOL I plugged that in Yenka dunno why or how, but I've ordered and calculated 1nF. Damn me for my idiotness!

I've ordered no fixed resistors, everything are pots, so i can adjust and finetune anything that's not working as it should.

Security resistor, sounds good for redundant protection... I don't want the mosfet to blow, it's right after my solid switch, the most expensive component on my board! I'll add that, sounds great!

The pieces haven't arrived yet... But I've found an error on my order

I've ordered a triac instead of a thyristor... But browsing the nets it can be plugged as a triac since I have a diode right before it... It just will not protect from reverse currents (just like a normal switch would do). I guess I haven't failed as hard as I thought.

Second main issue is.. Thyristors/triacs can't be turned off! Does that mean I'll need a physical switch between the capacitor and the triac to stop it from draining current and let them charge again? This is a drawback I hadn't thought about. Maybe just "charging > switching off charger > firing >switching on again" will do the trick o.O

Well... I've started to work on the body of the gun and I've started to do some testing with my new rounds... No more 1.5mm nails, I've moved to low carbon 5mm iron rod, I'll be making 40x5mm rounds with very sharp point to increase penetration

I've left a 1mm gap between barrel and round to allow it slide thru "frictionless".

Oh well I'll start to think about how to plug all that stuff... Parts will be here on monday!! I hope to have the armature build by then and just stick on the solenoid+barrel, solder the board, stick it and go BANG BANG all over the place xD

just joking, I'll test the firing power, do some measurements and start to build a light trap to measure the speed of the rounds

It's a never ending circuit I guess...

Thanks again!!

GhostNull

Fri Oct 08 2010, 11:00PM

Registered Member #2648 Joined: Sun Jan 24 2010, 12:45PM
Location: Australia
Posts: 291

Phew, that's good.

A Triac is just two identical SCRs in anti parallel so it should be fine. It will protect for reverse current as long as the voltage change, time does not exceed the Triac's limits. I don't think it will have any reverse voltage on the Triac any way through, the only thing that could cause reverse voltage on the triac would be the inductance of the solenoid and you got a flyback diode to take care of that any way.
Thyristors turn off at zero volts (in theory) so, as you say, you will have to turn off or disconnect (disconnecting is better) the charging circuit before firing, then turn it on again after.

It pretty much any electronics project is an endless circuit. You always learn, improve, add, learn, improve, add and so on, continually gaining knowledge, experience and under stand. That is, until you can't be stuff to do any more and quit >.<

You've done really great work here. Looking back from the first post you made, It's amazing to see the progress you've made in both your knowledge and understand as well as the project itself. Great work, keep it!
-Ken

Artikbot

Sat Oct 09 2010, 07:45PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Well well... I've been doing some optimisation in both the round and the cannon. I'm leaving the aluminium one till I have the whole thing finished and working, then I'll possibly wrap another coil (8 layers, but this time 16AWG wire), or just swap barrels to a fiberglass one.

I'm quite impressed with the close-range power of this thing.

I'm just using 7 120uF caps charged to 300V and it has enough power to break a crystal jar into pieces, go through 30mm of Styrofoam, perforate a few layers of cardboard, substantially mark meacrylate and almost perforate a soda can! I don't wanna know what's gonna happen when I power this with 20 caps... Or even with 40!

Artikbot

Thu Oct 14 2010, 10:45AM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Alright. This days since the ICs havent arrived yet (omfg this is taking ages!), I've been doing some testing with barrels, cap banks, and I've extracted some conclusions.

As someone stated (can't remember who was, maybe Barry?) Eddy currents represent a 25% energy loss on the coilgun. I've done the following test:

1: Fired the coilgun against a thick (60mm) styrofoam block using 960uF at 300V. Marked the round position to repeat it later. Held the barrel so recoil didn't interfere on my tests.
The round went 26mm into the styrofoam as an average

2: Repeated the test, but with 1760uF this time, at the same 300V. Round stuck 29mm inside the styrofoam as an average.

Observations: The coil was MUCH hotter at 1760uF than 960uF. This implies that the power from the additional capacitors was actually there.

Conclusions: Or either Eddy currents produce LARGE inefficiencies (in this case, up to 60% energy lost), or either my coil is too resistive and causes suckback to appear.

Both cases are pretty easy to solve. In the eddy currents case, I've already ordered a carbon fiber tube, I'll test with it next week. In the case of resistive coil, I can either rewind using thicker gauge or simply place the projectile a bit further from the coil so it it sucked a bit longer and suckback disappears (or minimizes).

And as a side result... I discovered that the mechanical switch I was using wasn't very situable, it burnt (literally melted) at the twentieth shot... Opened it and it was half melted half welded xD

Do you guys think it's suckback or eddy currents? Or maybe just an inefficient switch, but I think that's not the biggest issue imho. It's one of the others.

Thanks!

Edit: I've cancelled the carbon fibre tubing order, as I've read its still conductive. Way less than aluminium, but still conductive, and I'm looking forward to achieve as much efficiency as possible, keeping speed above an usable value. Instead, I've bought a fiberglass tube from Decathlon, that also costs 1/4th of the carbon fiber's cost.

ScotchTapeLord

Thu Oct 14 2010, 12:33PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

If you add more capacitance for the same given coil, then you increase the duration of the pulse, which will require you to sit the projectile further back to prevent suckback. Unfortunately, placing the projectile further back weakens the initial effect of the magnetic field (which is when it is strongest), so you need to find a compromise between pulse width and projectile position.

The fact that your coil got hot means that resistive losses in your coil have dominated.

Artikbot

Thu Oct 14 2010, 12:40PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Pulse lenght increase while increasing capacitance... Hadn't thought about that, I had in mind that parallel=same ESR, series=ESR summatory.

If it is the other way around... Guess what. I'll charge them in parallel and discharge them in series, less current>less temperature... but less field strength as well, and this is undesirable.

A thicker gauge coil would mean less resistance, therefore less pulse duration, alright? Ima get a thicker gauge wire asap, maybe 16 AWG. Also, the wire thread is 32m long in contraposition to the 20AWG thread, which has 22m. That might allow me to do a second stage somewhere in the future.

I'll start with the fiberglass barrel, then try to find out what's delaying so much my discharge. There's a chance that it was my mechanical switch right?

Thanks :)

Edit: I'm thinking about getting a computer based 'scope and an LRC multimeter for this christmas... I might need them quite a lot as I continue to advance in my projects. What do you think?

Artikbot

Sat Oct 16 2010, 03:02PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Alright, PCB is built, components are soldered and everything is ready to go.

I've tested it, and here it goes what's happening:

Voltage starts to build up at a quite fast rate. Reaches 24V in less than two seconds, but then it stays there. The mosfet heats up quite a lot, it has a small heatsink attached to it (coming from an old motherboard chipset heatsink, it's rated at 15W power dissipation I think). If in two seconds it can heat up to the point it feels hot, maybe something isn't working exactly well.
Apart from the inductor, nothing on the circuit builds up heat. And the inductor is warm to touch. This is the wiring diagram for the mosfet:

NE555 output - 47Ohm resistor - Gate
Inductor output - Drain
Negative (ground) terminal - Source.

It's an IRFP450 rated at 400V 16A sustained. The circuit should never build up currents higher than 0.1Amp at the circuitry behind the inductor.

Possible causes:

1- The MOSFET is connected backwards/improperly wired? (I followed the Yenka wiring diagram I used to simulate the circuit)
2- The diode can't switch fast enough? (don't think this is the cause)
3- The heatsink is insufficient? (don't think this is the cause either, the mosfet is rated at 16 Amps and the current flowing through it is not even 1/10)
4- Improperly calculated voltage dividers for the inverting input of the 741 op amp? (this would be the cause of voltage stopping at 24V, but for the mosfet to build up heat?)

Thanks!

Edit: I've found something that can be the cause, but I need confirmation.

I'm using a reverse biased zener (its what my show had in stock, and they are fine aswell) as a voltage limiter for the non-inv input of the op amp, but I calculated my resistor wrong, and it might be outputting the wrong voltage. Could this cause the Op amp to make the 555 stop outputting the signal and therefore cutting voltage at 24V?

GhostNull

Sat Oct 16 2010, 09:29PM

Registered Member #2648 Joined: Sun Jan 24 2010, 12:45PM
Location: Australia
Posts: 291

I haven't had any experience troubleshooting a boost converter so the only advice I can give you is just general trouble shooting.

Take a DMM and check the voltage on the non-inverting input and other parts of the circuit. Double check the polarities. Check that all the resistors and components are the right values. Ensure that the ICs are alive and well. If you have an Oscilloscope available, you might want to try that but that might be more trouble than it is worth.

Be sure to tell us what happens!
-Ken

Artikbot

Sat Oct 16 2010, 11:22PM

Registered Member #3247 Joined: Mon Sept 27 2010, 09:42AM
Location: Spain
Posts: 137

Issues found so far:

1: Zener diode was connected in the wrong way. Solved.
2: The resistor for the Zener diode was incorrect. Solved.

So far, I don't see any more problems on the circuitry.

Tomorrow I'm perforating the heatsink to attach it properly to the MOSFET, until that is done, I won't power up again the circuit.

Will update post tomorrow morning.
Thanks again!

Update: Seems like that was not causing the issue. The voltage on the capacitors stays at 24V, but the worst of it all, when current is cut, they discharge almost instantaneously.

Also, the status LED doesn't light up. And it is wired to the main power source. What the hell is going on in there? Batterly supplies 7.56V to be precise.

Polarities have been double checked and resistor values checked again. Everything seems to be in order.
my probl
Can a slow diode be causing all this trouble? Would a Schottky diode instead of a regular silicon one solve the problem? Can the inductor be the problem? I haven't checked its inductance, but according to some estimations it must be somewhere around 300uH, and my circuit does fine with 75uH.

I don't understand what's happening here, and without a 'scope... it's hard to tel where is the problem. I'll keep double-checking stuff, rewiring and recalculating resistors.

Oh wait, another potential issue:

I'm feeding the IRFP450's gate thru the NE555 timer (fed at 7.4V)... That makes me wonder if the voltage isn't enough to turn on "hard" the MOSFET (by hard I mean fast > less switching losses > lower heat > higher switching speeds). Can that cause the MOSFET to switch slow, therefore heat up and don't let voltage rise fast?

GhostNull

Thu Oct 21 2010, 11:10AM

Registered Member #2648 Joined: Sun Jan 24 2010, 12:45PM
Location: Australia
Posts: 291

Check the comparator and the reset pin. I was doing some simulations on LTspice, I was wokring on a comparator with feed back and I wired it up wrong. It caused to capacitor to charge to about 20v then stoped. Check that, it might be the problem.

You should use a sufficiently fast diode too, the voltage drop doesn't matter that much, just means lost energy. What diode are you currently using?

first previous 9 10 11 12 next last

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