Linear transformer driver explanation

[artis]

Ok, so to be honest I am not really sure to which subcategory this belongs, but I think it is high energy physics.

So my question will be rather brief and vague , can you explain me the working principle of a linear transformer driver?
Google wasn't of that much help with this one as it seems like a rather new technology and also very field specific.

Now my own very rough guess is that it is a special type of fast high power impulse switch that enables ion/neutral beam? acceleration for certain fusion approaches like the pinch etc types? Any insight is welcome, thanks

 

[Astronuc]

https://en.wikipedia.org/wiki/Linear_transformer_driver - "The LTD is capable of producing high current pulses, up to 1 mega amps (10⁶ ampere), with a risetime of less than 100 ns."

https://www.sandia.gov/Pulsed-Power/res-areas/advanced-pulsed-power-concepts/linear_transformer_driver.html

https://prod-ng.sandia.gov/techlib-noauth/access-control.cgi/2006/067148.pdf

High-Current Linear Transformer Driver Development at Sandia National Laboratories
Michael G. Mazarakis; William E. Fowler; K. L. LeChien; Finis W. Long; M. Keith Matzen; D. H. McDaniel; others
https://ieeexplore.ieee.org/document/5373875 (Purchase required)
Abstract:
Most of the modern high-current high-voltage pulsed power generators require several stages of pulse conditioning (pulse forming) to convert the multimicrosecond pulses of the Marx generator output to the 40-300-ns pulses required by a number of applications including X-ray radiography, pulsed high-current linear accelerators, Z -pinch, isentropic compression, and inertial fusion energy drivers. This makes the devices large, cumbersome to operate, and expensive. Sandia, in collaboration with a number of other institutions, is developing a new paradigm in pulsed power technology: the linear transformer driver (LTD) technology. This technological approach can provide very compact devices that can deliver very fast high-current and high-voltage pulses. The output pulse rise time and width can be easily tailored to the specific application needs. Trains of a large number of high-current pulses can be produced with variable interpulse separation from nanoseconds to milliseconds. Most importantly, these devices can be rep-rated to frequencies only limited by the capacitor specifications (usually 10 Hz). Their footprint, as compared with current day pulsed power accelerators, is considerably smaller since LTD do not require large oil and deionized water tanks. This makes them ideally suited for applications that require portability. In this paper, we present Sandia National Laboratories' broad spectrum of developmental effort to design construct and extensively validate the LTD pulsed power technology.

High-voltage, high-current, fast pulsed and repetitive

 

[artis]

thanks @Astronuc I already went through most of these links before but I can't find a good explanation on how the very LTD itself works and it's parts etc. they mostly mention it together with some other project it is part of.

 

[Astronuc]

Basically an LTD is formed by groups of capacitors and switches (in series) and the groups are all parallel feeding into a set of transformers.

See if these descriptions help:
https://www.sciencedirect.com/science/article/pii/S2468080X17301012

https://plasma-gate.weizmann.ac.il/uploads/bibman/mcbride:2018a.pdf

 

[artis]

oh thanks @Astronuc , these sources brought more understanding to me with regards to the technology.

see this drawing I made in paint, I take that the basic LTD scheme resembles this one, so can I say it is essentially a specific type of a pulsed transformer because the basic transformer action here seems like in that of any other transformer just that the primary winding and it's driving elements are incorporated into the transformer for what I suppose are shorter wire distances and lower inductance for faster and stronger pulse forming.

 

[whyamianswering]

oh thanks @Astronuc , these sources brought more understanding to me with regards to the technology.

View attachment 263564
see this drawing I made in paint, I take that the basic LTD scheme resembles this one, so can I say it is essentially a specific type of a pulsed transformer because the basic transformer action here seems like in that of any other transformer just that the primary winding and it's driving elements are incorporated into the transformer for what I suppose are shorter wire distances and lower inductance for faster and stronger pulse forming.

It's not quite like an actual transformer. The caps are charged and have a direct electrical connection to the load - no magnetic shenanigans there. The reason the magnetic core exists is to increase inductance in the case of the ltd to ensure driving all current through the load. We drive a high current in a loop around the cores (~200A) to induce a magnetic field increasing induction . The russians actually used to just use a big ring of iron iirc, whereas we use thin magnetic tape wound a few thousand times between epoxy. You're correct about NEEDING lower inductance for LTD performance, and that's generally considered heavily in the design of the switch and capacitors and transmission lines.

Also, the "bricks" (2 capacitors and a switch) are all wired parallel. The LTD modules themselves are in series.

Pulse shaping is interesting though. For stronger pulses you obviously need all the bricks to fire at once into the load, but there have been some instances where doing weird pulse shaping has been important for experiments.

 

[artis]

The reason the magnetic core exists is to increase inductance in the case of the ltd to ensure driving all current through the load.

I looked at some schematics again , it seems to be the case , well I guess a better comparison would be that of an autotransformer , it seems the core is used so that there would be voltage gain somewhat like in an autotransformer?

 

[whyamianswering]

Voltage for an LTD is a means to an end - the goal of an LTD is to achieve high current and a fast rise time. The energy stored in a capacitor is ~cv^2, so if we can get a high voltage on the caps, we can obviously deliver a lot of energy.

Also, the magnetic cores are usually just premagnetized before a shot occurs.

Again, the sole reason for the magnetic core is to stop current from going where we don't want it to, aka the case.

I should also clarify by "case", I mean the casing that envelopes the capacitors/switches/etc..

https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8526364

Page 24 should explain it more clearly.

 

[artis]

Ok @whyamianswering I think I get it, so each LTD toroid has a fixed voltage dictated by it;'s capacitors, and it's current is related to the number of "bricks" in parallel

within it, then stacking up many such toroid LTD's in series forming a barrel adds the total maximum voltage between the upper and lower end outputs , so the current is related to the current through an individual LTD toroid, but adding them in series increases the voltage so the total current through the load also then increases.

Also, the magnetic cores are usually just premagnetized before a shot occurs.

Is this done in such a way that the current from the caps is magnetizing the cores in the opposite direction so to increase their inductance per given size?

Anyway I now see the reason for the cores, yes indeed they have nothing to do with a transformer, the name transformer in the LTD seems misplaced.

By the way @whyamianswering are you working with these devices ?