Capacitor discharge switch

Hi, I have a 28800uf capacitor bank charged at 400vdc that I need to discharge fast. I was thinking of using a MOSFET to switch it. The source pin would go to the negative, the drain pin would go to the load, there would be a 100k resistor between the gate and source. 5 volts would be sent to the gate pin to "trigger" it. Would this work?

Hi Tacamo!
i think if you use mosfet as the capacitor, in your case you have to connect the gate of Nmos to load, and all other pins Drain, Source, and Bulk connect to the Gnd, and for Pmos you connect Gate also to load, and all other pins as mentioned before connect to Vdd.
Hope this help!
Regard.

I will suggest you to use a Resistor and a relay instead of MOSFET.

TACAMO said:

Hi, I have a 28800uf capacitor bank charged at 400vdc that I need to discharge fast. I was thinking of using a MOSFET to switch it. The source pin would go to the negative, the drain pin would go to the load, there would be a 100k resistor between the gate and source. 5 volts would be sent to the gate pin to "trigger" it. Would this work?

Click to expand...

I don't understand the purpose of that resistor. Also, you may have a hard time finding a 400+ volt rated logic mosfet.

As an alternative, I would consider an SCR (thyristor). They are easily triggered with 5 volts, available in high-voltage versions, and can control high currents as well.

John

I think throught higher value of resistor....I will take same time to discharge the capacitor.....as it have expoental decay.....But ofcouse you need to have some resistance in series with correspoding power rating with switch ( any thing MOSFET of SCR with equal power rating)

Hi, I have a 28800uf capacitor bank charged at 400vdc that I need to discharge fast.

Click to expand...

If fast discharging is of interest this means in a short time you wish to discharge the entire charge. This results into very large currents. If you look to the size of capacitor, the potential to which it is charged you realise that.
If this is the case, and Precise time of discharge and jitter is not very critical then it is advisable to use a relay for discharging. Also the relay current rating is for continuous current. This rating has to be viewed in terms of the charge transfer per second.
C = 0.0288 F
V = 400 V
Q = C*V = 0.0288*400 = 11.52 Coulomb, this is a huge charge.
Knowing that the discharge is not usually a square pulse, but to understand the issue we consider the following:
Discharge time in µs and Current during discharge in Amps is shown in table attached


If the discharge is slow enough lasting for 1 mS (1000 µs) the magnitude of the discharge current is to the tunes of tens of Kilo Ampere (11.5 KA). In fact the discharge current is not a steady current so the peak current involved is going to be much higher.
THEREFORE SELECT THE DURATION OF DISCHARGE AND SELECT A SUITABLE SWITCHING DEVICE CAPABLE OF HANDLING SUCH A CURRENT.
Other possibilities are thyratron, Krytron.
Thyratron - Wikipedia, the free encyclopedia
Krytron - Wikipedia, the free encyclopedia

In any case of discharging this capacity, you have to use a series resistor. Otherwise as ark5230 gave you the charged capacity may damage a lot of under capacity components and at the same time the shortened discharge is also not recommended for the capacitors life.

WOW thank you. Back to the drawing board. So about this SCR. It can be triggered with 5 volts? What amperage should the SCR be rated for.

TACAMO,
Handling of fast discharges like this and High currents in KA regime requires lot of skill. Size of conductors also determine parameters that we did not bother.
If you decide how fast it should be, the Time constant of the discharge circuit, you can find suitable resistance ( if it is repetitive process the wattage is also going to matter).
Once these factors are figured out then you start planning for the switch. Whether you are going for SCR, Thyratron or something else will depend on the operating conditions at hand.
Lot of EMI will be generated that will have serious effect on sensitive digital and analogue circuitry around.

TACAMO

As just stated, it would help enormously to know what you really are trying to do. I suspect you are trying to make some sort of rail gun. If you are trying to make a capacitor discharge welder, your voltage is way too high. If it is something else, please give some indication of the time and current you want to achieve.

Remember that the capacitors have a built in resistance (esr). In the best case, that will limit the current you can reach. Of course inductance in the capacitor and other components as well as resistance of other components will ultimately determine the current.

Now, on to the question about SCR (thyristor). Here are some useful links with annotation.

Powerex is a common supplier in the US. I used the T500 series for a CD welder mainly because I got some used ones really cheap and wanted the stud mounting. Look at the non-repetitive surge current ratings.

Powerex Result
https://www.pwrx.com/pwrx/docs/t500.pdf

Semikron is another excellent product. Its application notes are a little hard to find, but it has an excellent note on gate drives. Unfortunately, I couldn't find it in the time I had today. I will look more for it later. The first link will take you to a table showing lots of offerings. The second is a specific device I arbitrarily picked so i coud reference a specific illustration. Scroll to Figure 9 (Gate Trigger) of that datasheet to see what the gate drive needs to do. You want to turn it on quickly, but not over drive it.

**broken link removed**
**broken link removed**

John

TACAMO,
As the jpanhalt gave you the links, there are available Thyristors rating current from 500 ampere to 1500 amperes or more with high working voltage, too. If your are planning to make Rail Gun or a Capacitor discharge Welder or you planning to Charge some Big Magnets, you must put here detail for the sake of proper help not for any other purpose, atleast.
In all the cases suggested you in the Earlier posts you must be aware of the High Current handling and the result of any accidental direct discharge of the capacitors.

I found the Semikron note, but it has been revised. Here are some related application notes, including what is apparently the revised note I referenced above. For the Semikron note, search on "di/dt" to get to a section on gate triggering. The Semikron note is also particularly good at describing what each of the specifications in an SCR datasheet means.

https://www.onsemi.com/pub_link/Collateral/HBD855-D.PDF

https://www.littelfuse.com/data/en/Application_Notes/AN1001.pdf
https://www.littelfuse.com/data/en/Application_Notes/AN1003.pdf
https://www.littelfuse.com/data/en/Application_Notes/AN1008.pdf

https://www.semikron.com/skcompub/en/technical_explanation_discretes.pdf

John

Sorry, I guess I should of gave more info. I am making a smaller version of "Thumper". (Thumper - Aluminium Mesh - YouTube). It has no purpose what so ever, and I had a capacitor bank just lying around.

Just be very careful with those voltages and currents. I think the idea is neat, but at anything over 25 volts or so, it can be "very" deadly -- no second chances. Thumper is related to a rail gun.

A couple of years ago I published my capacitor discharge welder. It used 0.5 F at up to 30+ volts. After several months, I was concerned about some of the questions I was getting and removed it. I was a bit amazed the first time I fired it using 4 AWG cables to the electrodes. They quite noticeably jumped.

Good luck and be safe. Maybe a CD welder (useful for battery tabs and other tack welds) is more appropriate and useful? I think we would all be more comfortable helping you build something like that.

John

Just imagining of 400 volts on a huge bank of capacitors (28,800µF), I get like thinking of safety first. As such energy will be real deadly with no chances to survive, as John also said. Better to go for some innovative creation.

Additionally in such fast discharges, intense EMI is generated and line inductance and different capacitive parts give rise to high frequency oscillations at very high voltages. If you just bring a neon lamp (or a line tester with neon lamp) near the discharge circuitry it will show a glow for a short time without any electrical contact because of the EMI. This is also a majour concern if the discharge is really fast involving high currents.