driving a mosfet with optocoupler and a voltage source between 5V to 60V

Thread Starter

Carlos Delfino

Joined Dec 16, 2016
16
I am studying how to design an optocoupled active load, and I need to control the MOSFET with the signal obtained from a 4n25. The voltage that will be used at the gate will come from the load to be tested, so it will range from 5V (or 10V) to 60V.

What would be the best way to make the driver for these active load mosfets?
 

Thread Starter

Carlos Delfino

Joined Dec 16, 2016
16
The circuit on the LED side is not powered by a PWM source, but by a set of opamps that will provide the current level necessary to activate or not the mosfet through the 4n25.
 

LowQCab

Joined Nov 6, 2012
4,309
If your FET is not going to be used for any type of PWM, or "High-Speed-Switching",
a simple Photo-Voltaic-Gate-Driver will work with no other Components.

As an extra bonus, the outputs can be treated just like a free-floating Battery,
meaning that they can be paralleled for faster Switching-Speeds,
or connected in series for a higher Output-Voltage if needed.
.
.
.
 

Attachments

Thread Starter

Carlos Delfino

Joined Dec 16, 2016
16
@LowQCab
I don't know if I understood correctly from what I read in the datasheet, but no external component is necessary to compose the mosfet driver, could you please show me a circuit that uses this optocoupler, as the datasheet does not exemplify its use graphically.
 

Ian0

Joined Aug 7, 2020
10,277
Presumably, this must be some sort of variable current drive, because otherwise the MOSFET is almost a dead short across the power supply. How is the current regulated? There seems to be no sign of a feedback loop.
 

Thread Starter

Carlos Delfino

Joined Dec 16, 2016
16
Presumably, this must be some sort of variable current drive, because otherwise the MOSFET is almost a dead short across the power supply. How is the current regulated? There seems to be no sign of a feedback loop.
Being an active load, there will be another optocoupler on the shunt resistor represented by R1, the other side of the optocoupler will be connected to a comparator/amplifier circuit to compare the references and generate control for the optocoupler that will control the mosfet.
 

Ian0

Joined Aug 7, 2020
10,277
Being an active load, there will be another optocoupler on the shunt resistor represented by R1, the other side of the optocoupler will be connected to a comparator/amplifier circuit to compare the references and generate control for the optocoupler that will control the mosfet.
Now that’s just asking for trouble! Optoisolators are SLOW and worse than that they are slower to switch off than they are to switch on. Each one will put a pole in the feedback loop which equates to a zero in the transfer function. Each pole gives a 90° phase shift, and and a amplifier with gain 180° phase shift is called an oscillator.
Unless you are happy with an exceedingly slow response time, it will almost impossible to get stable.
 

LowQCab

Joined Nov 6, 2012
4,309
It's so simple that You don't need to simulate it's operation.

The only necessary consideration when using this device is the Input-Capacitance of the FET.
This will determine how fast the FET will be fully turned-on.

The FET does not like being turned-on slowly, this causes heat to be created very fast.

So the only thing to calculate is how big and tough the FET must be for your expected Load-Wattage.
How big and tough the FET is
will directly affect the value of the Gate-Capacitance of the chosen FET,
and consequently,
the Gate-Capacitance will determine the amount of time required to fully "Turn-On".


For Simulations .....................

The FDA217 can be represented by a simple "Current-Source",
which is then connected to a Capacitor,
with the same nF / pF Gate-Input-Capacitance rating as the
FET that You have chosen to use in your project.

The resulting Voltage-Curve should be compared to the
Gate-Voltage vs Rds-On graph or the, Gate-Voltage vs Drain to Source Current
in the FET Specification-Sheet.

Doing this will let You calculate the amount of HEAT that will be generated during the
time that the FET is not fully turned-on.

The FDA217 has a special "Turn-Off" Circuit that turns-off the FET very fast,
so the Turn-Off characteristics are generally not a problem that must be considered,
only the Turn-On-Speed is important to calculate.

Important point ..........
Remember that the FDA217 can be wired in parallel to double the amount of Turn-On-Current.
This will cut the Turn-On-Time in half.

.
.
.
 

LowQCab

Joined Nov 6, 2012
4,309
Being an active load, there will be another optocoupler on the shunt resistor represented by R1, the other side of the optocoupler will be connected to a comparator/amplifier circuit to compare the references and generate control for the optocoupler that will control the mosfet.
.
What does "Active-Load" mean ?????

What is the complete, and detailed, description of the "Load' ?

High-Speed-Switching can not be used with an FDA217 Gate-Driver.
.
.
.
 

Thread Starter

Carlos Delfino

Joined Dec 16, 2016
16
Now that’s just asking for trouble! Optoisolators are SLOW and worse than that they are slower to switch off than they are to switch on. Each one will put a pole in the feedback loop which equates to a zero in the transfer function. Each pole gives a 90° phase shift, and and a amplifier with gain 180° phase shift is called an oscillator.
Unless you are happy with an exceedingly slow response time, it will almost impossible to get stable.
I understand the risk of self-oscillation of the circuit, but as for the slowness of the optocouplers it is not a problem in the first instance, I consider it extremely important to isolate the control part from the load part, this is the proposal of this new project, with its pros and cons.
 

Thread Starter

Carlos Delfino

Joined Dec 16, 2016
16
.
What does "Active-Load" mean ?????

What is the complete, and detailed, description of the "Load' ?

High-Speed-Switching can not be used with an FDA217 Gate-Driver.
.
.
.
An active load is a load that electronically controls the current through a source, for example to test the discharge time or quality of a battery, it could also be used to test a new voltage/current source design by imposing a resistive load controlled.
 

Thread Starter

Carlos Delfino

Joined Dec 16, 2016
16
I was suggested to replace the FDA217 with a VOM1271, it really seems like a good replacement, as the VOM1271 is much better documented and already has a spice model.
 

LowQCab

Joined Nov 6, 2012
4,309
Now I'm totally confused .........

Are You attempting to create an adjustable Load ?,
or are You attempting to switch a device powered by an adjustable-Load ????

Please explain what it is that this project is supposed to do, in detail.

Every description so far is entirely too vague to be able to understand it.
.
.
.
 

Thread Starter

Carlos Delfino

Joined Dec 16, 2016
16
Well, I don't know how to be more detailed on the subject, so for an example, watch this video where an active load proposal is presented. I'm looking to make a model similar to the one in the video but with a new approach, in particular, fully isolating the control parts from the load, that's why the use of optocouplers, despite being a diagmos, didactic project, I'm trying to make it as viable as possible.

Video of the project that inspired my project, see at 32:33 minutes the reference circuit for controlling the active load.:
 

Ian0

Joined Aug 7, 2020
10,277
BB148AD3-05B9-4EA2-A090-68B0A338B5BC.jpegI made one using an almost identical circuit to that. As it was for12V batteries, and I never intended to discharge them below 10V I chose R5 to have 10V across it at maximum current, which reduced the dissipation on the MOSFETs.
 
Top