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 3 4 5 6 next last

Move Thread

LAN_403

klugesmith

Fri Aug 13 2010, 03:03AM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

Sorry, that was just a hint, a starting point, and my ASCII picture was not a fully thought-out circuit.
As you guessed and somebody else posted here, there -are- standard circuits for the job.

Thanks for posting a schematic in response, including the coil clamp diode and resistor
(which let you trade turn-off time against peak "kick" voltage).

On the right side of your turn-off capacitor, try adding a resistor from +350V.
Start with both SCR and IGBT (or BJT or SCR #2 or MOSFET) in the off state, so C is uncharged (both sides at +350V).
When the SCR fires, current starts to build in main load while left side of C is pulled down to the SCR "on" voltage.
Right side of C promptly drops to that same voltage, then rises toward +350V as it's charged through the R.
Suppose it reaches +50V during the intended SCR On time (a couple ms?).
Now turn on the secondary semiconductor, pulling right side of C to a small positive voltage
while left side of C would go to about -50V and turn off the SCR. Full load current is diverted to the secondary semiconductor, dropping off as the C charges in the opposite direction and as the main clamp diode takes the current.

If I were to simulate this, would include some nanohenries of parasitic inductance in the places where large currents suddenly switch from one loop to another loop.

You might look up circuits for a closely related application: "thyristor" automatic electronic flash units. These manage to stop the current in an already-ignited flashlamp and leave the unused energy in the capacitor.

To add a picture to your own existing post, you first upload the picture to a new "dummy" post in the Attachments forum. Then cut and paste the link and thumbnail image from your Attachments post to your original post.

DerAlbi

Sat Aug 14 2010, 02:14AM

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

I would really love tosimulate this. I understand now what your approach is.
I am not able to find SCR and IGBT models for LTSpice

DerAlbi

Wed Aug 25 2010, 04:22AM

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

I am designing a trigger-circuit now. I got some big SCRs for now so I want to use these instead of IGBTs. The plan is:

There will be used one IGBT to avoid a current through the 2nd SCR so that it will stop conducting.
The problem is: the C needs to be bipolar so far i can see

-first both siddes at 350V.
-then fireing, both sides at 0V, cap is charging through the IGBT and R to 20..50(?) Volts.
-the 2nd (right) SCR is tuned on and the IGBT is turned off. => 1st SCRs top side is at -20..-50V the SCR turns of after 80Âµs (as far as the cap delivers the current for so long)
-when SCR is turned off the cap is still charging on its left site to the 350V.

So three cases are present at the capacitor:
1) Same Voltage at both sided at C. 350----||----350
2) Capacitor is charged positivly: 0-----||----20..50V
3) 2nd SCR is tuned on, 1st SCR is conductiuve: -20..-50V----||----0V => 350V----||----0V

Bipolar capacitors are not verry big. I do not have any Idea how big the cap must be.. will 20ÂµF be enough?
Or do i need bigger capacitors? If yes, the solution with the 2 SCR will not work.. right?
Please notice: 5kA are the goal.

I am really sorry that i change my concept every post

It is really hard to find out the best solution.
If there is the possibility to save some IGBTs it would reduce costs extremely

GhostNull

Wed Aug 25 2010, 10:15AM

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

How about considering the V-switch?

DerAlbi

Thu Aug 26 2010, 08:35AM

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

OK, i am working on a trigger-circuit. V-Switch seems the easyest way. Its just a matter of capacity of the second cap.
But i have a question concerning dI/dt for the second SCR:
The first SCR discharges the big Cap. There the current rise time/rise rate is limited by the inductance of the coil. When the second SCR is turned on, then _immediately_ at least the current of the first SCR flows through SCR2, further rising if then slowed down though the inductance again..
Is this correct? Is this really a source for some problems?

Sat Sept 04 2010, 09:45AM

Registered Member #85 Joined: Thu Feb 09 2006, 01:22PM
Location:
Posts: 21

For multistage coilguns i recommend using one SCR per coil and 2 IGBTs and 2 diodes in an "asymmetric halfbridge" configuration. This configuration has several advantages: -it recovers energy from the coils into the capacitors. -the current in the coil decays quickly after turnoff. -you only need 2igbts for as many stages as you want.

if i am not mistaken, 5 years ago a member of this board achieved 21% efficiency with this configuration. (

GhostNull

Sun Sept 05 2010, 01:17AM

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

EN:the link is a bit screwy

Here's the schematic
UT25

DerAlbi

Sun Sept 05 2010, 07:27AM

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

This seems the cleverest concept at all. The biggest problem will be to switch the upper IGBT because i will need a voltage 20V abovwethe CapVoltage. But this is managable.
However this concept has a big disadvantage: the IGBTs will not longer be used pulsed. Since the IGBTs provide the current for every Coil the die has no time to cool down.
I think i will build this concept. I already have all the parts.. it shouldnt be too hard.

Maybe 10 of my 320A (continous) /700A(pulse 1ms) IGBT in parallel will do the job.

Thanks for that!

Edit:
A general question to the Light beam triggering: since photodiodes do not have a very high sensitivity the inputimpedance of the input-circuit needs to be relativley high.
In additon we have long cabled from the diode to the circuit... right next to the Coils. I really would like to hear some experiences if this can cause some problems...

Mon Sept 06 2010, 08:44AM

Registered Member #85 Joined: Thu Feb 09 2006, 01:22PM
Location:
Posts: 21

have you considered phototransistors? they should be fast enough and will work with a much simpler circuit. i used phototransistors on all my CGs with good success. twist the connection wire pair and you will have less noise problems.

DerAlbi

Mon Sept 06 2010, 09:33AM

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

I also thought about shielded audio cables. Last night i have designed a 10x12cm PCB cabable of 2x evaluating 16 inputs and firing 16 SCRs. Additonally a driver circuit for many parallel IGBT-Gates for high and low voltage (upper und lower IGBT)
To use a Phototransistor would reduce my input impedance to several Kiloohms. Better than 220k for photodiodes. But in principal both types use the same ciruit. So changing the Pullupvalue is not that hard.
Additionally i have decided to use CPLDs for inputevaluation. This gives me the chance of some logic implementation.. e.g. only considering the next input that should be triggered. This also will give a much better noise imunity than ORing all signals together.

first previous 3 4 5 6 next last

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