Ring Launcher Controller - Charging/Monitoring Challenges

[ryanmills]

Relevant Datasheets
PBHV3160ZX: https://assets.nexperia.com/documents/data-sheet/PBHV3160Z.pdf
G2RL-1-E DC5: https://components.omron.com/us-en/datasheet_pdf/J117-E1.pdf
INA260: https://www.ti.com/lit/ds/symlink/ina260.pdf


Given the higher voltages and significant energy stored, I'm seeking some guidance. This is a controller for a ring launcher (similar to a rail gun), we previously bought something similar, but it was discontinued and prone to failure. I have attached what I thought in my head was a simple concept, but I see a few issues and I'm unsure the "most correct" way to solve them. It uses a dedicated 12VDC power supply for the MCU and a second external "capacitor charger" with a max 400VDC output to charge the bank. An MCU (not shown) will be added to display of the bank voltage, launch control and charge control. I'm well aware of the dangers this circuit poses and posting here seeking knowledge on how do it correctly and safely. If anyone has worked on something similar, I'm open to any suggestions on a better way to do any of this.


1. The bank voltage monitor needs to constantly measure, and the voltage divider is going to be parasitic on the capacitor bank this is both good and bad, I do want to have some sort of bleed but need to keep it very slow. In this case since the only load is the voltage divider and the INA260 I'm unsure how to determine R1's watt requirements in this scenario. I believe changing the voltage divider from 36V to a say 5V limit would further reduce the parasitics but is there a better way to monitor the voltage that would limit drain on the capacitor bank? I looked thru TI's entire INA line, and they max out at 120V, all the example higher voltage demo implantations use a voltage divider.


2. The goal is to have feedback in the charging, they will turn a handle to charge to the bank and control how much charge before launch so I need on/off control. The PNP is rated for 600V but only 100ma. The max current output from the external charger is unknown with no datasheet. I believe as drawn I would smoke the PNP unless the charger output was below 100ma. I plan to test the draw but in this scenario I really need the full 400V and i'm not sure the best way to limit the current going into the PNP. Or if a PNP is even the best way, a relay is safe but its contacts would get abused and noisy and not many can handle 400VDC, SSR's are too expensive, didn't really see any ideal optocoupler that could handle 400V volts without the same 100ma limit.


Schematic:

[TurboTom]

Ryan -

your schematic is problematic in several places:

The charging switch won't work the way suggested; when enabled, R5 will conduct a current up to 0.4A, thus dissipating 160W. Q2 needs to be capable of blocking the full charging voltage. If it's got to be an external charge voltage supply, better use an NPN power transistor (or even still better a darlington) in common-collector circuit, place a resistor between base and collector and use another (smaller) high voltage transistor to pull the base low when charging should be disabled.

Yet, my prefered approach would be to directly make a flyback converter (depending on the time you've got to charge the capacitors, a simple inverter with a single inductor and a fast MOSFET switch will do the trick). You could power the whole thing via a 24V DC supply and use a feedback line to disable the converter when the required voltage has been reached.

You may want to place bleeder resistors across each of the charge storage capacitors to even out the voltage distribution.

Your relay won't survive the torture of dumping the charge into the coil for long. The contacts of the 16A relay will weld closed within a few shots. Also contact bounce is a very bad thing here. That's a typical job for a "hockey puck" SCR (Thyristor) which can be found second hand relatively inexpensively and comes rated for impulse currents of several tens of thousands of amps...

I've still got a similar (unfinished) project in a drawer that's been designed to magnetize neodymium magnets after installing the "blanks" inside a hollow shaft, yet the power is much higher, using a 3200µF 1kV charge storage cap... Even though the project that this contraption was supposed to be used for eventually came to a stop (high speed turbo generator), I may still build it, just for fun and since the compopnens are already "collecting dust" in the basement.   

Most important when "playing" with such voltages / energies, always be safe and don't electrocute yourself!