Hello,

This is likely a very simple and silly question. I have a circuit where I am driving a capacitor with a 120V AC signal.

I would like to have a switch on the bottom side of the cap between it and GND so that I can use a low voltage control signal to float the capacitor. What would be the best circuit element to achieve this given I need to do this to ~1000x capacitors with separate control signals for each?

Thanks,
Aman

 

I have a circuit where I am driving a capacitor with a 120V AC signal.

How/why do you drive a capacitor?

But anyway you could put a MOSFET under each cap and control that MOSFET with a 10V, low current control signal. Putting a voltage on the MOSFET's gate would connect your cap to ground. Removing the voltage will open that connection. I'd put a resistor to ground on the gate, to protect agains the controller failing to "open". You want the gate voltage either 0V or 10V, not floating in between.

 

Hi,

You should probably show the circuit you intend to use this with because the complexity depends on the application.

 

Hi all,

Thanks for the feedback. I was under the impression that a MOSFET wouldn't work because it is unable to handle the AC current from drain to source, and that I would need something like a TRIAC (which I would like to avoid as they are big and bulky costing a lot of board space).

When I tried simulating a FET in LTSpice (picked a random one with low Vds), I saw that there was still current flowing when the transistor was "off" (gate driven low). Could someone shed some light as to why this would happen?

Circuit is essentially the following LTSpice simulation attached.


Thanks,
Aman

 

Most MOSFETs have an internal body diode that conducts in the "wrong" direction (source to drain).

 

Possibly a solid-state relay?
http://www.farnell.com/datasheets/77327.pdf

 

Or just shift where "ground" is. Without more details, we're just guessing.

 

The problem seems straightfoward to me. I see the "intended use" as some sort of test fixture to stress caps with an AC voltage. It's just a guess, and I have no guess why an individual cap needs be turned on and off but...

A TRIAC device is ideal for turning AC on and off with a low voltage control. I can't suggest beyond that as I've never actually used one so I lack the nuances you'd need to know before buying a thousand parts.

 

...
When I tried simulating a FET in LTSpice (picked a random one with low Vds), I saw that there was still current flowing when the transistor was "off" (gate driven low). Could someone shed some light as to why this would happen?
...

It is very bad practice to specify a pulsed voltage source that switches from 0V to 240V in zero time. This causes all sorts of unrealistic currents to flow...


When the NFET is turned off (time<403ms), the current decreases (but doesn't go to zero). The reason it doesn't go to zero is because of the capacitance from drain to source, drain to gate, and the capacitance of the reversed-biased parasitic diode from drain to to source.

 

Hi All,

Thanks for the responses.

1) I tried simulating with FET's without a body diode (which to my understanding are only present in Power FETs since they are a consequence of the solution to the intrinsic latchup problem they may have). The reason for this is because obviously with that diode there I would not be able to switch off the current in both directions.

2) Long story short the "cap" I am putting AC voltage across is actually a material that acts like a capacitor and this creates different properties to the material when a larger electric field is present.

3)Any solution able to reduce the electric field across the capacitor to 0 while still driving the top of the cap with +/-120V AC, it will be viable.

4) SSR suffers similarly with the TRIAC in that the part is bulky and space ineffective.

5) Thanks MikeML for catching that, did not realize it would have adverse affects on the simulator. I re-sim'ed with 100ns rise and fall time and still see a significant current across the cap (+/- 1A), which seems much larger than the current that would be caused by the parasitics of the MOSFET

Any other advise/suggestions would be much appreciated.

Thanks,
Aman

 

Just to add on to 5) and 1):

The large current through the cap (+/- 1A) seems to be mainly because the voltage at the bottom of the cap does not seem able to go below 0V when it should "float" to +/- 120V AC.

This may be due to the NFET fixing Vdrain to not go below Vsource, thus forcing current to flow (I don't know the exact mechanism as to why that would be the case though).

Again, although parasitics play a small role, they do not seem to be the major issue here.

Thanks,
Aman

 

Is a SOT-223 too bulky? (6.5x7.0 mm)

http://www.mouser.com/ProductDetail/ON-Semiconductor/MAC08MT1G/?qs=sGAEpiMZZMuAO0%252bGuNbnQm%2fmyygwiDySCgQb0%2flty84%3d

Bob

 

Hi Bob,

Yeah the SOT-223 package is what we have for our current TRIAC and I would to use something smaller ideally. However, if that is my only option than I will make due.

This was kind of why I was trying to reach out to the forum to see if there is a FET/BJT solution that would be viable which would be smaller than the SOT-223 package.

Thanks,
Aman

 

Bump, anybody else know how you could do this with mosfets/transistors that is a very space efficient solution?

Thanks,
Aman