Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

npn driving mosfet, low current draw

Status
Not open for further replies.
F

fjb2069

Newbie level 6
Joined
Jul 25, 2007
Messages
13
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,379
Using a npn to drive N channels mosfet switched at 18 volts.

I need to get npn current consumption in the micro amps range (~100uA). Currently a ~1mA.

Increasing R3 will lower current but also reduces switching speed. I need to keep switch speed as fast as possible.

Any suggestion on a method to do this?

npn_circuit.png

<a title="npn_circuit.png" href="http://obrazki.elektroda.pl/2208834200_1479829655.png"><img src="http://obrazki.elektroda.pl/2208834200_1479829655_thumb.jpg" alt="npn_circuit.png" /></a>
 

Hi,

Why the NPN at all?

Use a logic level mosfet and control the gate directely from microcontroller.

***
You talk about speed. Give values.
Now you have a 10k pullup. This surely slows down switch ON time.

The Mosfet isn´t well chosen. There are much better ones. You don´t need 200V.

Klaus
 

Actually using a IPB80N03S4L-03 mosfet that is 30 V 80 AMP capacity. It is an automotive fet with low RDS. I can't find logic level type at similar price range with automotive specs and low RDS.

I guess I need to create logic level circuit that would be embedded in logic level mosfet.
 

A single-transistor logic level converter is always going to require high quiescent current in order to have high switching speed. You should instead use a complementary driver, as can be obtained in special ICs. IXDN604 comes to mind.
 

mtwieg said:
A single-transistor logic level converter is always going to require high quiescent current in order to have high switching speed. You should instead use a complementary driver, as can be obtained in special ICs. IXDN604 comes to mind.
Click to expand...


Based on additional circuitry costs, cost of logic level fet may be way to go. Didn't know if anyone had a low cost solution... < $0.50 per unit.
 

Hi,

still missing information about switching speed.

Klaus
 


KlausST said:
Hi,

still missing information about switching speed.

Klaus
Click to expand...

Klaus,

I am really not sure how fast I need. Maybe you can shed some light on how to determine. I am using mosfets as high power switch 18volts 50 amps, so switching over a 50us might cause it to not saturate fast enough where it can damage fet.

What would be a reasonable switching speed required?
 

Hi,

I need to keep switch speed as fast as possible.
Click to expand...
I really don't know what you are talking about.

I guess nobody here can tell you how fast switching you need. I hope you know.

Some designers like very high switching (ton, toff in less than 20ns) for low power loss, some designers say they want slow switching to reduce EMI.

Klaus
 

KlausST said:
Hi,


I really don't know what you are talking about.

I guess nobody here can tell you how fast switching you need. I hope you know.

Some designers like very high switching (ton, toff in less than 20ns) for low power loss, some designers say they want slow switching to reduce EMI.

Klaus
Click to expand...

I am not concerned about EMI, more about not damaging mosfet from too slow switching 50A load. So I guess the 10-20ns range would be better.
 

Hi,

So I guess the 10-20ns range would be better.
Click to expand...
Guessing won't help. Why don't you calculate?
Mosfet case? Switching frequency? Max allowed switching loss?...

Klaus
 

IPB80N03S4L-03 I2PAK case. Switching Frequency - N/A, will be periodically turned on and off (typically hours or days between).

I can understand how switching loss might be a concern if this device was switching at a constant frequency, but I am using it in a steady state application (as a relay). I know that switching it on too slow can cause it to transfer power during a non saturated period, when the resistance is high, therefore generating a lot of heat and possible damaging the fet. In my application I want to saturate the fet as quickly as possible so that it reaches RDS ON.

So I do not believe switching losses are a concern? If a am understanding switching losses correctly!
 

Hi,

switching time:
Worst case (most power dissipation) is when the MOSFET is at half supply voltage. 18V / 2 = 9V
Then at a ohmic load you also see half of the current. 50A / 2 = 25A.

Now with these two values go to the datasheet "Safe operating area" chart.
To make it easy use 10V VDS. There is a vertical line. Now look for 25A (between 20A line and 30A line) and find the crossing point with the 10V line.
You are below the 1ms SOA line. You are safe with switching times of 1ms.

This is worst case. 1ms continously at 9V/25A. In your case 5ms may be realistic.
--> you are far away with your estimated 20ns. (factor of 50000)

So relax.
This does not mean you should try to get close to 1ms.

With the "gate charge" diagarm you see you need less than 60nC to fully switch the MOSFET ON. (charged to 10V, power_voltage = 24V)
This means with an average gate current of 6mA you are able to switch the MOSFET ON in less than 10us. Now you are a factor of 100 within the SOA.

Conclusion:
* you don´t need extremely fast switching time (10us is very relaxed)
* you don´t need high gate current (a couple of 10mA is well sufficent)

Klaus
 

KlausST said:
Hi,

switching time:
Worst case (most power dissipation) is when the MOSFET is at half supply voltage. 18V / 2 = 9V
Then at a ohmic load you also see half of the current. 50A / 2 = 25A.

Now with these two values go to the datasheet "Safe operating area" chart.
To make it easy use 10V VDS. There is a vertical line. Now look for 25A (between 20A line and 30A line) and find the crossing point with the 10V line.
You are below the 1ms SOA line. You are safe with switching times of 1ms.

This is worst case. 1ms continously at 9V/25A. In your case 5ms may be realistic.
--> you are far away with your estimated 20ns. (factor of 50000)

So relax.
This does not mean you should try to get close to 1ms.

With the "gate charge" diagarm you see you need less than 60nC to fully switch the MOSFET ON. (charged to 10V, power_voltage = 24V)
This means with an average gate current of 6mA you are able to switch the MOSFET ON in less than 10us. Now you are a factor of 100 within the SOA.

Conclusion:
* you don´t need extremely fast switching time (10us is very relaxed)
* you don´t need high gate current (a couple of 10mA is well sufficent)

Klaus
Click to expand...

Klaus,

Thank you for explaining this to me. Very helpful!!!!!!
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top