Mosfet as Switch....

Status
Not open for further replies.
A

abuhafss

Full Member level 2
Joined
Aug 8, 2010
Messages
142
Helped
1
Reputation
2
Reaction score
0
Trophy points
1,296
Visit site
Activity points
2,383
Hi

I want to switch an inductive load of 0.6 Ohms (250W) using a MOSFET.
Here is my idea. Simulators does not show 20A passing.
Kindly correct me if I am wrong and suggest me improvements, if any.

Thanks.
 

Attachments

  • Screenshot 2014-08-26 20.24.40.png
    15.1 KB · Views: 177

So if I am reading the circuit right, you are trying to use a NMOS transistor to turn on an inductive load. You are using a npn transistor to drive the MOSFET gate on and off. So what happens is when the switch is off the npn is off and the NMOS is power on. When the ignition switch is on, the gate is grounded and the switch is off. So are you measuring the simulation on the correct switch condition where the NMOS is on? I also did notice that it you have the nmos hooked up backwards that what is usually done. Usually you have the NMOS as a low side switch where your load is connected to power and the NMOS completes the circuit to ground. So you may have a polarity issue here. The way you have the circuit hooked up you would need a charge pump to turn on the NMOS.
 


I made the schematic with FALSTAD simulator and I had used PMOS not NMOS which I selected within the simulator. Anyway, it is a PMOS.
 

Ah you're right, sorry not the greatest eyes. So PMOS makes a lot more sense, are you sure that it is hooked up correctly still to get it to turn on? I am not sure on Falstad if it matter or not.
 

cl10greg said:
Ah you're right, sorry not the greatest eyes. So PMOS makes a lot more sense, are you sure that it is hooked up correctly still to get it to turn on? I am not sure on Falstad if it matter or not.
Click to expand...

So far I have just checked on simulators, none of them is showing 20A which made me doubtful if my configuration is correct or not.
 

By choosing unsuitable MOSFET types, you can easily make the circuit fail. But which did you actually select?

Are you aware of the inverting circuit logic, opening the ignition switch will power the load and vice versa? Is it a bug or a feature?
 

Making a quick test using some commercial components resulted on simulated measures bellow for current flowing at motor:



Note: Not checked if devices were either properly selected or well biased.
 
Okay, got it. The Rds(on) of MOSFET should be lowest.
But why the transistor need to be such big 2N3055 (10A)? Can't 2N3904 do the job?
 

abuhafss said:
why the transistor need to be such big 2N3055 (10A)? Can't 2N3904 do the job?
Click to expand...

As said before, I did´t selected optimum devices sized accordingly to the need, but just placed the first found in the library. Assuming will just perform ON/OFF operation, I presume you could even employ a signal NPN transistor, such as someone from BC54x family.
 

Apart from component selection details and questionable input signal polarity, it should be also mentioned that the circuit isn't suited for operation in a car. In this case it most be able to withstand expectable voltage transients without component damaged. E.g. the MOSFET gate-source voltage must be limited, respectively the driver must be overcurrent protected etc.
 

I tried the configuration with IRF4905 (74A, R-ds 0.02Ohms), 2N3904, 100R/2W and 271K varistor (in place of diode) on a three-wheeler. It worked but, I think the 271K varistor could not protect the mosfet. The actual load is about 0.4 Ohms. The gate resistor draws about 130mA and the base resistor draws about 120mA so, BC54* family cannot be used which can handle only 100mA.

Crutshow>>> I want to use it in place of starter relay to power-up starter solenoid.

Any further improvement suggestions shall be highly appreciated.

Edit: Varistor is 271K 500V
 
Last edited:

For operating a solenoid there's no need for high speed so you can make the collector (gate) resistor 1kΩ and the base resistor 10kΩ for the NPN transistor. The collector current is thus only 12mA and just about any NPN can be used.

Stay with the diode across the solenoid for transient protection but you may want to add about a 30V 1W zener in inverse series with the diode to reduce the turn-off time of the solenoid and minimize contact arcing.
 


Is this what you are purposing? Is 1N4007 OK or some high rated ampere?
 

Attachments

  • Screenshot 2014-08-26 20.24.41.png
    10.7 KB · Views: 190

I think you burnt out the base-emitter junction of your 2N3904 transistor with a base current that is much too high. For the load of 120mA, the base current should be 12mA for it to turn on well and the resistor value should be (12V - 0.7V)/12mA= about 1k ohms, not 47 ohms which produced a base current as high as (12V - 1.0V)/47 ohms= 234mA!! (If the input signal source can supply that much current at 12V.)
 
Last edited:

The best Answer usually follows the best defined Question.

Input:
12V battery
Solenoid with 0.6 Ohm coil resistance 250W load.

Need 12V@1mA high side switch closed to control high side 12V switched output to solenoid with <2.5W loss max and suitable protection for personal use only.


This now defines some scope and implied switch resistance of 6.25 mOhm max. The cost of this low RdsOn may be considered with the cost of the required heatsink for such devices.

Since it is non-inverting and high side, we know the best choice could be either a High-Side ASIC or at least a PMOSFET or consider using a low side if possible.

Thus you search for these parameters; Switch, High-side, non-inverting, 20A, 24V, <6 mOhm

Then sort by price and look at other requirement you forgot to specify; SMT or THT? Etc

One example; https://www.digikey.com/product-detail/en/AUIR3313/AUIR3313-ND/2605404
 


The transistor is Ok.
The MOSFET's gate & drain are short.
 

Hello.

To drive charges from VCC you can use a PNP, PMOS, or PJFET, you have lots of choices.

Try using the P channel version of IRFZ44N (don't remember its name), it's just 18mohm RDs(on) resistance and can drive upto 49A.


Check your VGS voltage, if VGG doesn't reach at least 4/5 VSS (design aproximation) your FET will work at field effect or saturation region (as amplifier), and will disipate power. Your load won't turn off also.

For high current loads BJT are not recommended even more as you're not switching fastly.

Ask if you disn't understand at all.

Bruno.-
 

abuhafss said:
Is this what you are purposing? Is 1N4007 OK or some high rated ampere?
Click to expand...
Yes, that is what I was suggesting. The 1N4007 should be fine. But upon further though I realize that a 1W zener will not tolerate a 20A pulse. You will need a 10W zener or a 30V transorb for that. Alternately you could put a 1.5 ohm resistor in series with the diode, in place of the zener, to reduce the solenoid turn-off time.
 


If using a TVS diode, do I need 1N4007 in series?
Is P6KE (6.8-400V, 600W peak, 5W steady) okay here?

What wattage of 1.5 ohm resistor would be fine?
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…