3x 18mF @ 350V -> What to do? / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

3x 18mF @ 350V -> What to do?

first previous 4 5 6 7 next

Move Thread

LAN_403

DerAlbi

Fri Sept 10 2010, 12:02PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

A Simulation shows that the current through the coils needs the same time to drop as it needed to establish. So turning on the Current will result in the exact double time that the current flows through the coil.
With this half bridge configuration much more Stages are possible. To limit the barrel length, the coils need to be as close toghether as possible. But it is necessary that the last fired SCR is safely off. Is there any idea how to calculate or predict the minimum distance between 2 coils?

The optimal switching time would be that the current is already gone then the projectiles second half starts to escape from the coil. So theoretically the coil distance should be at least one coil length. Is that correct?

Is it ok to to have a 4cm long coil but only 3.5cm long projectile. I have read that both length should be the same.. But a longer Coil gives will give a longer area where no force is applied to the projectile so that suckback is avoided. Is that a goot tradeoff?

I have made a 3D model of the current configuration. Main detail is the distances of the coils.

GhostNull

Fri Sept 10 2010, 12:58PM

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

I've used photo interuptors (with phototransistors) and the rise was about 300uS at best.

Working Out wrote ...

So your bank is 3.3kJ.
Let's say you get an efficiency of 10%
So projectile will have 330J.

Since KE = (m*v^2)/2
2KE/m = v^ 2
v= sqrt(2KE/m)

Lets say you projectile is 15g

so your velocity will be:
v = sqrt(2*330/0.0015)
= sqrt(44000)
= 209 m/s

wait wtf? oh well

so your projectile is 3.5cm long so
209 m/s = 20 900 cm/s
20 900cm/s = 0.209 cm / uS

3.5 / 0.209 ~= 16.74uS

So your projectile would take aprox. 16.74uS to travel past one point... that can't be right
Let's try again!

So your bank is 3.3kJ.
Let's say you get an efficiency of 5%
So projectile will have 165J.

Let's say you projectile is 1/4' dia and 3.5cm and is 1018 low carbon steel
(1/4) inches = 0.635 centimeters
the projectile will have a volume of
pi*r^2*h = pi * 0.635^2 * 3.5
~= 4.4337 cm^3

1018 has a density of 7.7-8.03 Ã—1000 kg/m3 (according to

)
7.7*1000 kg/m^3 = 7700 kg/m^3
7700 kg/m^3 = 0.0077 kg/cm^3

The projectile will therefore weigh:
0.0077 * 4.4337 ~= 0.034139 kg

Since v= sqrt(2KE/m)
v = sqrt(2*165/0.034139)
= sqrt 9666
~= 98.32m/s

Ah, that looks better.

so your projectile is 3.5cm long so
98.32 m/s = 9832 cm/s
9832 cm/s = 0.009832 cm / uS

3.5 / 0.009832 ~= 355.98 uS

Your projectile would take aproximately 356uS to travel past a single point given your projectile is 1/4" diameter and you acheive 5% efficiency (not that hard unless you get suckback)

The bottom line is, Phototransistors would be too slow for a coil gun this size. You'll have to find another method (or my sensor was REALLY sh**)

DerAlbi

Fri Sept 10 2010, 01:50PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Thanks for that fast response. I do see what is your point. I am sorry that i didnt do this simple calculation by myself.
I am suffering from inexperience verry baldy so that my concept changes every minute. In the meantime the barrel has a inner diameter of 1cm

But your calculation are not in vain. This gives me an idea of how long the pulses will be that i will have to handle.

Phototransistor problem:
It is normally a problem of saturation and the fact that PoTr. are indeed slower compared to photodiodes. Like normal NPN transistors that are saturated they are not capable of swithcing high frequencies. Addionally there could be an influence of parasitic capacitors and slow risetimes an high switiching thesholds of the evaluating circuit.
If this becomes really a problem, i could design a low voltage operationg ciruit. Low voltage change eliminates the influence of parasitic capacitance and also it lowers the Impedance of the used circuit. Then a highspeed comperator converts the voltage level to the other logic level. That is not that hard to handle.

The IGBTs switch on in 700ns and off in 1.5Âµs. Thxrisotr Turnontime is 2Âµs.
Since my Gatedrivers are verry oversized your predicted pulselengths will not be imposible.

So i am thankfull for your calculations, and pleased that the project is doable. (isnt it?)
I do really hope that i do not annoy somebody with my inexperience..

GhostNull

Fri Sept 10 2010, 10:46PM

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

No problem, the point is that you're readily moving forward. Photo diodes would be indeed faster. There's also magnetic based sensors if you want to investigate that, but they are more complex.

DerAlbi

Mon Sept 13 2010, 12:21AM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

I have experimented with photodiodes and phototransitors today. So far i have decided to use photodiodes driving a transitor. The amplified signal then is fed into a comparator with adjusable threshold.
This gives sharp rects with a delay <2Âµs also on capacitive load.
Phototransitors are indeed too slow and also not strong enough. Inputcapacitance would still be to high to withstand parasitic induction (if it happens). Now the biggest problem now is to find a transistor with high amplification and ultra low basis-capacitance. This is really the limiting factor (besides the comperator delay)

GhostNull

Mon Sept 13 2010, 10:18AM

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

How about RF transistors, this one looks okay

DerAlbi

Mon Sept 13 2010, 11:36AM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Uhh this long legs should be attached to a woman

I am using SMD, but thank you for your interest

. I doubt that real HF transistors will have enough gain. 200 like your example is not that much considering that (my) photodiodes will only produce ~15ÂµA reverse current max under perfct conditions. In the end application LED and sensor will be ~1.3cm away from each other and the diode receive only a fraction of the emitted light that provided my test circuit.

Lets see what comes out. I will try some standard high gain transistors first. The delay can also be adjusted by setting the comperator reference volatge verry near to the the working point voltage. This should reduce the capacitive influence since the theshold voltagelevel is crossed verry fast/early..

This week i will try to manage the skeleton so that next week when the PCBs will be ready i have a working construction.. I will keep you informed

klugesmith

Tue Sept 14 2010, 01:03AM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1714

DerAlbi wrote ...

I have experimented with photodiodes and phototransitors today. So far i have decided to use photodiodes driving a transitor. The amplified signal then is fed into a comparator with adjusable threshold.
This gives sharp rects with a delay <2Âµs also on capacitive load.
Phototransitors are indeed too slow and also not strong enough. Inputcapacitance would still be to high to withstand parasitic induction (if it happens). Now the biggest problem now is to find a transistor with high amplification and ultra low basis-capacitance. This is really the limiting factor (besides the comperator delay)

Infrared-sensitive photodiodes are used as detectors in multi-gigabit fiber optic receivers.
To minimize the slowing effect of capacitance, the PD should be reverse-biased (photoconductive mode) and drive an amplifier with very low input impedance. You probably already know that -- what's your circuit configuration?
Could use an op-amp configured as transimpedance amplifier, or a bipolar transistor in common-emitter configuration, with provision to deliver some bias current to the base even when the photodiode is in the dark.

DerAlbi

Tue Sept 14 2010, 09:09AM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

I have tried OpAmp - They are really slow in comparison to an optimized Comparator. at least if you pay the same money for both

I think to attach the circuit tells more than a text can do - keeping my english skills in mind

If you have some Ideas to improve the circuit.. feel free.. (but as i said.. this should give a good result and small delays)

klugesmith

Thu Sept 16 2010, 12:01AM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1714

DerAlbi wrote ...
I think to attach the circuit tells more than a text can do - keeping my english skills in mind

If you have some Ideas to improve the circuit.. feel free.. (but as i said.. this should give a good result and small delays)

You have a very nicely drawn schematic and good looking circuit board layout.

Here is a change that might give you a larger, faster signal from the photodiode amplifier.
Connect emitter to ground without 10K resistor.
Insert 470 ohm load resistor between collector and +5V. (now this is an inverting, common emitter amplifier).
Add a 2M resistor between base and +5V, in parallel with photodiode.
Now with diode dark, you might have 2 uA of base current, 1 mA in collector, 0.5V drop in load resistor.
With diode fully illuminated, say 17 uA of base current, 8.5 mA in collector, 4V drop in load resistor.
The base voltage swing is now much smaller, so capacitance doesn't slow it down much.
The amplifier output impedance is now 470 ohms instead of 10K, so its speed is less sensitive to capacitance.

This primitive circuit is very sensitive to Hfe variations with process and temperature,
and comes close to saturating the transistor at max diode current & Hfe.
Improvements could include low frequency feedback to stabilize the operating point,
and/or AC coupling the pulse from amplifier output to comparator input.

first previous 4 5 6 7 next

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