From: Jim Yanik
Subject: Re: SMPS Topology Question
Date: Fri, 15 Nov 2002 15:47:23 +0000 (UTC)
Organization: Friends of Lummox
References: <firstname.lastname@example.org> <7iJ7uRBnpA19EwBQ@jmwa.demon.co.uk> <email@example.com>
NNTP-Posting-Date: Fri, 15 Nov 2002 15:47:23 +0000 (UTC)
firstname.lastname@example.org (james) wrote in
> John Woodgate wrote in message
>> I read in sci.electronics.design that james
>> wrote (in <email@example.com>) about
>> 'SMPS Topology Question', on Thu, 14 Nov 2002:
>> >I'm thinking about designing an offline SMPS to supply a voltage of
>> >400V at a peak current of 1A. This is for charging a large, high
>> >voltage capacitor bank for use with linear magnetic accelerators
>> >My question is: what topology to use?
>> >Since we are talking a peak power of 400W, should i be thinking of a
>> >forward converter or even half bridge or push pull?
>> >Anyone have any views on this or know of any resoureces that may
>> >help in this decision.
>> >thanks in advance
>> Forgive me, but I can't see why on earth you'd want to do that.
>> Doesn't a coil gun operate just as a one-shot device? So you have a
>> relatively LONG time to charge the cap. Even if you are on 120 V
>> mains, a voltage- doubler rectifier will give you 340 V just from the
>> filter cap. You can charge your big cap through a resistor at, say
>> 100 mA or even less. It just takes longer to charge. If it's 1000 uF,
>> then at 0.1 A constant current charge, it would take 340 x 0.001 =
>> 0.34 coulomb to charge it, which 100 mA supplies in 3.4 seconds. Of
>> course, the current won't be constant, but after 10 seconds the
>> voltage will be within 5% of 340 V.
> I appreciate what you are saying, but this is the prototype idea for a
> multistage, rapid firing coilgun. This will require rapid charging of
> the capacitor banks. My current cap bank size is 6.8mF with 200V caps,
> but I will be increasing the size anyway soon.
> I am just trying to make something that will be farily future proof
> for my designs.
> Again, thanks for all your comments.
For a prototype or demonstrator,I'd go with a current-limiting direct line
charger.Keeps it simple,fewer things to go wrong. Once you've got the coil
firing circuitry working,the cap charging supply can be made to work from
line V or from a portable supply.(generator?) Besides,you're not going to
get any faster charging than from direct line V,unless you go to great
expense and complexity.
The first step is to get the coil sequencer operation.And IMO,you need a
cap bank for each coil.A single cap bank would need to be fully charged in
milliseconds or less,very difficult to get a single cap bank to recharge
that fast for each successive coil on the gun assy. You need the high coil
currents to drive the projectile.
Think of a photoflash unit,and how they do multiple flashes from a single
cap.They cannot use the entire energy of the cap for the first flash,as the
charging inverter cannot recharge fast enough. They use a thyristor to shut
off current flow after each flash so the cap never discharges
completely(below flash potential).
Jim Yanik,NRA member
remove X to contact me