Lego NXT controlling 12v fans

Thread Starter

wrcsubers

Joined Oct 16, 2015
4
Hello Everyone,

I appreciate any help in advance!

I'm have a basic understanding of DC Circuits, and this is my first time working with transistors. I'm using a Lego NXT to switch multiple fans On and Off. I have a 12v NiMH Battery Supply that is connected to a 12v 4.0amp fan and the NXT is switching the transistor with a 7.5v supply. Currently I'm using a single MJE3055T Transistor. My issue is this: When using ANY value base resistor the E3055T rapidly generates heat, if I'm not using a resistor it gets hot but MUCH slower than it does with a resistor.

How can I get this transistor in to saturation without overheating? Do I need to be using a Darlington?

I'm guessing that this is happening because the circuit is trying to pull all amperage from the base? When the voltage at the base is low the fan is drawing a lot more current?

Thanks!
 

GS3

Joined Sep 21, 2007
408
Do I need to be using a Darlington?
The MJE3055T specs say min gain at 4 A is 20 so you would need to supply 0.2 A to the base. If you can do that directly then you are fine. At 7.5 V you can try a 27 ohm resistor. If your 7.5 V source cannot supply this base current then you can use another transistor in Darlington configuration.
 
The transistor is going to dissapate a fair amount of power across the collector emitter connections as well. Assuming a Vce(sat) = 0.2 V, a collector current of 4 A will cause a power of 0.8 W. This is in addition the power from the base current, P = V x I = 0.7 V x 0.4 A = 0.28 W. Now the part is rated to 75 W, so it can handle the power with sufficient heat sinking. Without heatsinks, 1 W of power will still cause the part to get HOT.
 

Thread Starter

wrcsubers

Joined Oct 16, 2015
4
Thanks everyone for the replies, that information is quite helpful.

Can anyone tell me why the Transistor would be getting so much hotter when using a larger value resistor on the Base? I don't quite understand what is happening in this scenario...
 

dl324

Joined Mar 30, 2015
9,151
I'm using a Lego NXT to switch multiple fans On and Off. I have a 12v NiMH Battery Supply that is connected to a 12v 4.0amp fan and the NXT is switching the transistor with a 7.5v supply. Currently I'm using a single MJE3055T Transistor. My issue is this: When using ANY value base resistor the E3055T rapidly generates heat, if I'm not using a resistor it gets hot but MUCH slower than it does with a resistor.
Measure the voltages at the transistor collector, base, and, when using a base resistor, the NXT output . I think you'll find that the NXT can't provide anywhere near 400mA and the issue you're seeing is that without the resistor the transistor is driven a little harder and it's dissipation is slightly lower. Without a heatsink, the longer the transistor is on, the hotter it will get. BTW, safe power dissipation will still be uncomfortably hot to the touch.
How can I get this transistor in to saturation without overheating? Do I need to be using a Darlington?
Assuming the issue is drive capability from the NXT, a split Darlington would let you drive the power transistor into saturation and not raise the saturation voltage.
 

dl324

Joined Mar 30, 2015
9,151
Can anyone tell me why the Transistor would be getting so much hotter when using a larger value resistor on the Base? I don't quite understand what is happening in this scenario...
Read post #6. With a larger resistor the NXT isn't able to provide as much drive current so the transistor is further from being saturated and power dissipation increases. Measure and provide voltages so we don't have to guess so much.
 

GS3

Joined Sep 21, 2007
408
Can anyone tell me why the Transistor would be getting so much hotter when using a larger value resistor on the Base? I don't quite understand what is happening in this scenario...
The transistor is not saturated and the C-E voltage is too high. Measure it.
 

MikeML

Joined Oct 2, 2009
5,444
...
Can anyone tell me why the Transistor would be getting so much hotter when using a larger value resistor on the Base? I don't quite understand what is happening in this scenario...
Look at this:

286.gif

The collector voltage V(c) red and Power dissipated in Q1 green is plotted vs the value of R2 in Ohms.

Note that R2 must be less than ~27Ω for the transistor to be saturated. If it is saturated, then the dissipation is 2W, which will make the transistor stinking hot if it is not bolted to a heatsink. For values of R2 >30Ω, the transistor comes out of saturation and gets even hotter.

In the extreme, if R2 is less than ~11Ω, the base to emitter current causes additional dissipation in the transistor.
 

MikeML

Joined Oct 2, 2009
5,444
Contrast the above with using an NFET:

286f.gif

Note that the NFET M1 dissipates P = IE = (Id*V(c)) = 3.977*0.066 = 260mW, which it can do with no heatsinking...
 

Thread Starter

wrcsubers

Joined Oct 16, 2015
4
Wow Mike... That graph is extremely helpful to understanding what is happening. I'll be fashioning a heatsink I guess! I'll also be picking up some low value resistors.

Thanks for the help I really appreciate it. I'll keep you posted on how it goes =)
 

dl324

Joined Mar 30, 2015
9,151
graph is extremely helpful to understanding what is happening. I'll be fashioning a heatsink I guess! I'll also be picking up some low value resistors.
You still need to determine the drive capabilities of your NXT. As I mentioned, I doubt it can provide 400mA.

If you use a Split Darlington, you can decrease the current required from the NXT by 2 orders of magnitude. Or you could use a power N MOSFET which replaced BJTs as the preferred switch when current capabilities became sufficient...
 

Thread Starter

wrcsubers

Joined Oct 16, 2015
4
You still need to determine the drive capabilities of your NXT. As I mentioned, I doubt it can provide 400mA.

If you use a Split Darlington, you can decrease the current required from the NXT by 2 orders of magnitude. Or you could use a power N MOSFET which replaced BJTs as the preferred switch when current capabilities became sufficient...
So the NXT can drive motors up to 0.7 amps. However, I'll be running 4 total circuits, and while they all won't be on at the same time, that could potentially be 2.8A; which I am guessing is too much for the NXT.
 

dl324

Joined Mar 30, 2015
9,151
So the NXT can drive motors up to 0.7 amps. However, I'll be running 4 total circuits, and while they all won't be on at the same time, that could potentially be 2.8A; which I am guessing is too much for the NXT.
You can replace the 2N3055 with a Split Darlington which will decrease current requirement from NXT to around 4mA or you can replace the BJT with an N MOSFET. If your NXT can sink and source current, you don't need any resistors on the base of the MOSFET. Just make sure the NXT output provides enough voltage to give a low on resistance.
 
Top