NPN transistor to switch on and off a MOSFET

[yash101]

Here's the schematic I followed (attached).

I can't seem to find what's wrong! I know it works because I have gotten it to work beautifully. It before got the motor to full speed and was cool to the touch.
I'm sure that I won't need a heat sink for such a small load! It's gotten an internal resistance of like 0.0025 ohms! That's less than the jumper wires' resistance.

The problem I have is wiring somewhere. Currently, I could care less of inductance because it's not even running high frequency! It's merely, press one button, forward, other -- backwards. Or if you want, both buttons -> self destruct.

[T3sl4co1l]

The high side gates have to be driven to (V+) + Vgs(on).

Tim

[yash101]

The high side gates have to be driven to (V+) + Vgs(on).

Tim

I know. The push buttons in my breadboarded circuit are connected to +12v (VCC) or the drain on the upper transistors. The gates of all the MOSFETs are pulled low using 100k resistors (as suggested by my DE teacher).

[Monkeh]

The high side gates have to be driven to (V+) + Vgs(on).

Tim

I know. The push buttons in my breadboarded circuit are connected to +12v (VCC) or the drain on the upper transistors. The gates of all the MOSFETs are pulled low using 100k resistors (as suggested by my DE teacher).

Your supply is 12V, and your FETs require Vgs 10V to switch on fully. That is the gate being 10V above the source.

How can your Vgs be 10V on the high side FET if you're only applying 12V?

[Monkeh]

The high side gates have to be driven to (V+) + Vgs(on).
Tim

I know. The push buttons in my breadboarded circuit are connected to +12v (VCC) or the drain on the upper transistors. The gates of all the MOSFETs are pulled low using 100k resistors (as suggested by my DE teacher).

Your supply is 12V, and your FETs require Vgs 10V to switch on fully. That is the gate being 10V above the source.
How can your Vgs be 10V on the high side FET if you're only applying 12V?

i think you need to say it plainly.... that he needs +22V voltage on the gate...

I was hoping he might have some neurons fire.

It's already been said pretty plainly on the last page.

[yash101]

I'm sorry, I'm unable to understand what issue you are seeing. I am feeding in between 12 and 14v to the gate, so I'm sure that it is saturating.

I think I may have possibly worked out a solution. Basically, I'm now using an NPN transistor set up in the following way:
The collector is connected to VCC. The base is connected to the source via a 100k resistor (which can be omitted). The emitter is connected to ground via a 100k resistor. Before the resistor, connected to the emitter, the Vout pin exists.

I've attached the schematic below.
Before I build this, please, someone, let me know if this will work! .

[yash101]

I was hoping he might have some neurons fire.

Yeah. I understand what you are saying, but I can't change my power source. I need to do something else!

[Monkeh]

I'm sorry, I'm unable to understand what issue you are seeing. I am feeding in between 12 and 14v to the gate, so I'm sure that it is saturating.

The gate must be 10V above the source. If the FET is fully on, the drain is at the same potential as the source, 12V. You must feed a voltage greater than present at the drain pin into the gate of an N-channel FET on the high side.

I was hoping he might have some neurons fire.

Yeah. I understand what you are saying, but I can't change my power source. I need to do something else!

Use a FET driver capable of taking the gate above the drain, or use a P-channel FET.

[yash101]

My project is due in this week so all I have to work with is what I have currently. After the due date, I can pursue it further and improve it.
Currently, my goal is to get it to work with the least resources possible.

I would preferably be able to build the H-Bridge with my MOSFETs and NPN transistors.

I don't have an easy source for PNP transistors and I'm already exceeding my budget.

I'm sorry if you think I'm just ignoring what you are saying. I'm not. It's just the problem that it's too late for me to do that. I designed this H-Bridge driver from my electronics experience from when I was young and had time. I CADed everything and then ordered all the parts together. Something might look perfect on paper but not work in real practice.

[yash101]

The battery would take another week to arrive.
We would need to change out our motors for 24 volt ones.
We would need to change our control logic voltage regulators because they are currently tiny LDO chips; they would die at that voltage.

Adding another battery is simply not a solution. Not to say, these high quality, high-discharge (65C) LiPos are quite expensive!

This is only practical after my project is over, if I choose to spend time and improve it. I would need to start a new fundraiser, which would be successful due to an actual physical prototype as well as experience from what we are currently facing.

[Monkeh]

We would need to change out our motors for 24 volt ones.

If you change your motors for 24V, and supply 24V to the drain pin, where does that leave your gate voltage?

Adding another battery is simply not a solution. Not to say, these high quality, high-discharge (65C) LiPos are quite expensive!

Use a 9V battery over the existing one. Hey presto, you're probably enough above the source to get away with it.

Got a 555? Build a charge pump. Probably take similar area with rough construction.

[yash101]

Monkeh, that charge pump idea is actually one I'm considering . It's quite complicated for something so simple, but I have a bunch of 555 chips and the current requirement is quite low!

We are using a pretty high quality LiPo from HobbyKing. It would be difficult to source something that can output almost 150 amps continuous that small!

Also, why do I need that? I've gotten my circuit to work. I'm just wondering why it isn't working now! When I got it working perfectly, the MOSFETs were fully saturating and running cold. The motor was running at close to full speed at the voltage being provided.

UPDATE: Maybe I should read a bit slower   "GATES"
A charge pump seems pretty simple to make. I attached a schematic to make sure it's right!

[Monkeh]

We are using a pretty high quality LiPo from HobbyKing. It would be difficult to source something that can output almost 150 amps continuous that small!

You don't need to put out 150A, you're only charging the gate of a FET! A 9V battery would do the trick.

Also, why do I need that? I've gotten my circuit to work. I'm just wondering why it isn't working now!

Because the gate must be raised above the source.. If the FET were fully turned on, the source pin would be at 12V. An unloaded motor might well get up to speed despite a partially turned on high side FET. Load it down and you'll have a lot of smoke in short order.

[yash101]

"You don't need to put out 150A, you're only charging the gate of a FET!"

I got that. Now I see what you are trying to say. It seems like quite a simple change! Thanks, guys!

Now, my new question is how to wire that up? I've never done that.

[T3sl4co1l]

Also, be extremely careful that the gates of a common side aren't turned on at the same time (i.e., both left transistors, or both right transistors).  Add some delay, 100ns minimum.  You might have very low resistance transistors, but if they touch any teensy little bit, RIP your power supply (and probably the transistors with it).

The pulses overlapping is kind of like this, so don't do it. http://roflposters.com/wp-content/uploads/2013/02/1212811803369.jpg.rofl.posters.jpg

Tim

[Monkeh]

"You don't need to put out 150A, you're only charging the gate of a FET!"

I got that. Now I see what you are trying to say. It seems like quite a simple change! Thanks, guys!

Now, my new question is how to wire that up? I've never done that.

In series. Negative terminal of the 9V to positive of your main battery, positive of the 9V to your gate drive. This will net you 8-9V above the main battery, which should adequately turn the FET on.

And as above, make damned sure you never end up with both transistors on a side turned on with a battery like that. Small wires become explosive fuses.

[yash101]

Yeah! I remember when I was playing with my initial iteration of this H-Bridge. It worked really well. It became really hot when I pressed both buttons simultaneously to see what happens. I did have some current limiting so nothing died, but it was fungerous

[yash101]

I came across this article:
http://www.neufeld.newton.ks.us/electronics/?p=133
I think that transistor is acting like an invertor. That means that when the circuit is turned on, everthing will be shorted out until the microcontroller can bring the input high.
To fix that issue, I could use a typical hex invertor. The fix will be really simplistic .

I guess that I should place either the charge pump or 9v battery between +12v and that 10k resistor.

Do you guys think I'm heading in the right direction now?

[Zero999]

I'm sorry, I'm unable to understand what issue you are seeing. I am feeding in between 12 and 14v to the gate, so I'm sure that it is saturating.

I think I may have possibly worked out a solution. Basically, I'm now using an NPN transistor set up in the following way:
The collector is connected to VCC. The base is connected to the source via a 100k resistor (which can be omitted). The emitter is connected to ground via a 100k resistor. Before the resistor, connected to the emitter, the Vout pin exists.

I've attached the schematic below.
Before I build this, please, someone, let me know if this will work! .

All voltages are relative.

The MOSFET will not conduct until its threshold voltage is exceeded.

A BJT will not turn on until the voltage between the base and emitter exceed 0.6V

If the voltage on the MOSFET's source needs to be 12V relative to 0V, then the gate voltage needs to be 12V plus the gate-source voltage required for it to pass the desired current.

Attached is an estimate of the voltages in your circuit.

[yash101]

Thanks! Do you use DipTrace? It looks like you completely redid that schematic! Thanks for the time!

I am going to use the suggestion which includes the 9V battery.

Do you by chance have a schematic for an all-N-Channel H-Bridge which possible won't need that?

Also, for my future reference, should I not use a MOSFET? I don't know if a BJT can handle high current without getting hot.

[T3sl4co1l]

You need a bootstrap gate driver, or a higher supply like the charge pump circuits shown earlier.

See how this uses +18V supply on the logic side, to drive only +12V in the all-N-channel bridge:
http://seventransistorlabs.com/PWM_Generator.pdf
Note that no time delay is provided, so the supply will be shorted out briefly during switching.  This was OK because the application had a current limited or constant current supply (simulated by the 1 ohm resistor).

Tim

[yash101]

Thank You. I just rebuilt my H-Bridge driver. It works cool to the touch with a 6 amp inductive load. As you guys have mentioned, it's not a very good idea. This is just for a display model at the moment. I'll be fixing the parts which you all are saying may not work properly.

Again, Thanks for all the time, guys!

[yash101]

I got some good P-Channel MOSFETs to use for the high side. Now, the circuit is working fantastically. It's taking 40A without much of an issue. There is no heat sink on the MOSFETs and they are connected to a loaded motor. The MOSFETs are barely warm after 30 seconds of use! I guess I could afford to use a small heat sink, just to be on the safe side!