Discharging negative voltage through mosfet

Thread Starter

tmarc

Joined Jul 23, 2017
10
Hi guys,
I want to create a circuit to use a series of positive (0-5V) pulses applied to the gate of a N-channel MOSFET in order to
periodically switch a negative supply voltage (-Vc) through a load resistor. If I understand correctly, because of the
negative voltage on the source, I would need a level-shifting circuit to decouple the gate driver circuit from ground,
so that the MOSFET can be biased and switched correctly. However I am having trouble figuring out more specifically
what kind of circuit I could use. Does anyone have any suggestions? Thanks
 

Attachments

jpanhalt

Joined Jan 18, 2008
11,087
When you say "discharge," I am not sure what you mean. If you simply want a negative voltage switch, have you searched on it? Years ago (6 years) I wanted to turn off a negative voltage before the positive voltage. I came up with this: Draft1.asc

Never tested it, as I later found that sequence really was not necessary.
 

Attachments

crutschow

Joined Mar 14, 2008
34,281
Below is the LTspice simulation of a simple level shifting circuit using a PNP transistor.

What is the value of the -Vc voltage in your application?
How fast does it need to switch?

upload_2019-4-17_15-55-38.png
 

Thread Starter

tmarc

Joined Jul 23, 2017
10
Hi jpanhalt/crutschow,

thanks for helping out. My -Vc is about -25V, switching speed is approx. 150 kHz. I have tested Jpanhalt's circuit in ltspice
and it does what I hoped for, I'm trying to implement it now.
 

crutschow

Joined Mar 14, 2008
34,281
One difference between jpanhalt's and my circuits, is that mine is ON for a positive input and his is ON for zero input (if that's important to you).
 

Thread Starter

tmarc

Joined Jul 23, 2017
10
Hi guys, I'm having some trouble implementing the level shifting circuit. The way it was provided by Jpanhalt it works well, however when
I try to integrate it into a circuit where there is a capacitor at -Vc that needs to be periodically switched then for some reason the switching
fails. The switch for the positive half cycle does what it's supposed to. It is evident that I'm not generating clean pulses to the gate of the -Vc MOSFET switch, but I'm not sure how to solve this. Attached is the schematic of my circuit. Thanks
 

Attachments

Thread Starter

tmarc

Joined Jul 23, 2017
10
Hi Crutschow,
I posted the full schematic to illustrate what I am trying to do and to give a complete picture - the level shifting circuit in this case
is connected to the gate of mosfet M2. The idea is to alternately connect capacitors C1 and C2 to a load (R2) through the MOSFET's.
The mosfet M1 is supposed to discharge C1 (+V) during its inactive (negative) half-cycle, so when it is not being charged. This is supposed to
alternative with mosfet M2, which should switch on capacitor C2 (-V) during the positive half-cycle. If you look at the waveforms at the gate
of M1 and capacitor C1 you can see that it switches the way I want it to, off during active half-cycle, on during inactive half-cycle. However,
I cannot get M2 to switch on the same principle. I hope my explanation is somewhat clear, please let me know if I should add anything.
 

crutschow

Joined Mar 14, 2008
34,281
Sorry, but the function of this circuit is rather confusing. :confused:
Exactly what are you trying to accomplish with this rather complex circuit?
The input is a high-frequency sinewave.
What output do you want?
 

Thread Starter

tmarc

Joined Jul 23, 2017
10
Hi Crutschow,
The voltage source in the schematic actually represents the secondary of a resonant transformer. The capacitor on the secondary that normally makes a LC circuit, is split into two capacitors where C=C1+C2. The capacitors are "isolated" by diodes so that they only feel either the positive or the negative half-cycle of the HF sinewave. The idea is to discharge one capacitor while the other is charged, and then discharge the latter while the former is charged, and so on. The idea behind this is to extract power from the secondary, without "loading" the source, because the high-Q tuned transformer is influenced by loading/insertion of lumped resistance. The triggering is done with comparators, where a positive (0-5V) output pulse of the comparator is meant to switch on the respective mosfet according to the condition that the capacitor being discharged at that moment is not being charged at the same time. The output is meant to go through a full bridge rectifier to a DC smoothing capacitor and finally to a (arbitrary) load. I hope I have made things a bit clearer. Thanks for helping out
 

Thread Starter

tmarc

Joined Jul 23, 2017
10
update: I have managed to get the circuit to work by using a dual-rail comparator and keeping the gate of the MOSFET at -Vc during the
off-time. Thanks for your help everyone!
 

Thread Starter

tmarc

Joined Jul 23, 2017
10
I don't think it is "over-unity" or anything like that, I am just experimenting with some HF switching techniques
for a resonant transformer. The idea is to create a split "buffer" in the system, where each half of the buffer is
switched to the load during the half-cycle where it is not participating in the LC resonance.
 

Thread Starter

tmarc

Joined Jul 23, 2017
10
Edit: maybe I misunderstood, do you mean that in order to have the resonance itself the energy must be able to cycle between L and C (or C1+C2), and so even if you extract it from the capacitor that is not currently participating in a cycle, you are loading the source?
 

crutschow

Joined Mar 14, 2008
34,281
even if you extract it from the capacitor that is not currently participating in a cycle, you are loading the source?
The law of conservation of energy means that any energy you extract from the output must, at some point in the cycle, be provided by the source.
There's no free (energy) lunch. :rolleyes:
 
Top