Go Down

Topic: What did I not understand about transistor amplification? (Read 2120 times)previous topic - next topic

a_lehmann

Hey there,

this my first transistor amplification and I'm running into some trouble of course.
In my original circuit I'm using a NPN transistor to pulse an IR LED with higher current than the arduino can provide.
The power source is 4 AAA batteries, so it's 6 V and up to 4,5 A (short circuit, limited by inner resistance).

The amplification was not that good so I made this easy circuit (attachment) to measure where the problem is.
When I got the transistor thing right there must be a high resistance between collector and emitter when there is no voltage on the base and nearly no resistance when I put voltage on the base. Right?

On my breadboard I have 30 mV when the base has no connection to anything and the full 6 V when I connect the base to +6V with a 10 K-Ohm resistor. That's cool.
But, now I'll describe my problem, there is only 100 mA flowing where the LED is. I measured it by taking the LED out, so there is only my multimeter, the transistor and the 10k resistor on the base. Nothing else. I expected the 4,5 A (which I had at the short circuit of the 4 AAA) but only got this 100 mA. Why?

The transistor is rated for 15 A / 80 V (it's a 2N6488). But I've also tested another npn with the same result.

So I would be glad if you could tell me what I didn't get about it. And how can I achieve to draw the full 4,5 A.

Greetings

Alex

pwillard

#1
Jul 21, 2011, 06:03 pmLast Edit: Jul 21, 2011, 06:26 pm by pwillard Reason: 1
1) Never a good idea to short circuit a transistor, regardless of rating.
2) I doubt you will find this exact configuration you have in ANY amplifier designs
3) Never assume there is some "inner resistance" in a transistor.

In short, I feel that you have less understanding about NPN transistors than you believe you do.  That's OK... but I'm not really going to try to explain transistors in a sentence or two. Are you sure you mean "amplify"?

How about we give an example of how you can get an NPN to amplify?  Good place to start?   Maybe I will draw something and do a follow up post.

A problem you have right now is this.  A transistor can SWITCH ON and OFF and it can amplify... sometimes it can both... but it's related to base biasing.   You *can* make a voltage amplifying switch this way, (if the collector V+ voltage is higher) though there are some issues with how you have things connected.  in other words... you have sort of the right idea... but a less than optimal hookup.  You can make a current handling switch in cases where you want a device to handle more current than a micro can handle.

Ok, rather than draw up something new... I'll give you a drawing I already created that does what you want.   Obvuously, change out my values of transistor and LED for yours.

http://pwillard.com/files/driver2.pdf

Erni

http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1230677423

Ernst

a_lehmann

I'm glad you try to help me!

Quote from: pwillard
1) Never a good idea to short circuit a transistor, regardless of rating.

Why not?

Quote from: pwillard
2) I doubt you will find this exact configuration you have in ANY amplifier designs

Yes, I know, I only build this to measure the current.

Quote from: pwillard
3) Never assume there is some "inner resistance" in a transistor.

How else would you tell it?

Quote from: pwillard
In short, I feel that you have less understanding about NPN transistors than you believe you do.

That's right I don't know much about any component. Very new to it.

Quote from: pwillard
Are you sure you mean "amplify"?

As you say it... No? I'm sorry, my English has got rusty. I don't want to amplify something like an audio signal. I just want more power than the arduino can provide. Like you say, an ON/OFF SWITCH . I'm using a timer interrupt to switch the LED ON/OFF every 13 us to achieve the 38 kHz the reciever needs. It already works on my breadboards but not with the desired current.

Quote from: pwillard
How about we give an example of how you can get an NPN to amplify?  Good place to start?   Maybe I will draw something and do a follow up post.

It would be great if you could help me to solve my problem.

Quote from: pwillard
A problem you have right now is this.  A transistor can SWITCH ON and OFF and it can amplify... sometimes it can both... but it's related to base biasing.  Using just 1 resistor won't really do it for you.  You *can* make a voltage amplifying switch this way, (if the collector V+ voltage is higher) though there are some issues with how you have things connected.  in other words... you have sort of the right idea... but a less than optimal hookup.

Yeah unfortunatley I don't get what I have to change yet. Do you understand me at least? I want this IR LED to be pulsed at 38 kHz. In the datasheet there is a description that says that you can use it with 3 A (for 10 us in the example). Because you don't get that much from an arduino pin I wanted to use a transistor. But I only get about 100 mA.

Quote from: pwillard
Ok, rather than draw up something new... I'll give you a drawing I already created that does what you want.   Obvuously, change out my values of transistor and LED for yours.

At first view I would say I've done it this way.

So why do I get only 100 mA when I short circuit a transistor? How do I get the transistor to let the whole current pass?

Quote from: Erni

http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1230677423

Do I have to use a MOSFET instead?

pwillard

#4
Jul 21, 2011, 07:22 pmLast Edit: Jul 21, 2011, 07:25 pm by pwillard Reason: 1
Starting at bottom.  Mosfet not required.

Quote
At first view I would say I've done it this way.

Look again.

1) you have a high value resistor when 1K is better.  You want the transistor to conduct.
2) You have connected the Emitter and Collector differently.
3) you do not have a current limiting resistor on the LED.

Essentially, let the Collector feed the load... (LED), connect Emitter to GND potential.

Quote
Never a good idea to short circuit a transistor, regardless of rating.

To put it simply... that's how you let out the magic smoke that makes it work.  A transistor will not limit itself to it's maximum rated current.  It will break (go poof) when it's maximum rated current is exceeded.  A short circuit will allow these conditions to exist.

This is also why you need to always limit LED current (with a resistor) to keep the  value below the maximum rated value for that LED.

An LED with no resistor current limiting will allow an effectively unlimited amount of current to flow through it... until it "fries" itself.  Same is true of transistor.

a_lehmann

#5
Jul 21, 2011, 08:14 pmLast Edit: Jul 21, 2011, 08:16 pm by a_lehmann Reason: 1
Quote from: pwillard
1) you have a high value resistor when 1K is better.  You want the transistor to conduct.

I think now I get it. I've just tried resistors from 1K over 500 down to 150 Ohm and the current was increasing. The the conduction depends on the voltage on the base. That's the part I didn't understand. I was believing it behaves like discrete 1 and 0 gate with the npn/pnp transistors.

One step further: Could I simply hook up the base to the Arduino pin without any resitors?

Quote from: pwillard
2) You have connected the Emitter and Collector differently. [...] Essentially, let the Collector feed the load... (LED), connect Emitter to GND potential.

Alright, I didn't know this is important.

Quote from: pwillard
3) you do not have a current limiting resistor on the LED.

Yes, I wanted the LED to go poof. But in my application I have a resistor for my IR LED.

Thank you so much, you really really helped me a lot in understanding this topic.

Quote from: pwillard
In short, I feel that you have less understanding about NPN transistors than you believe you do.

You were so right...

retrolefty

#6
Jul 21, 2011, 08:22 pmLast Edit: Jul 21, 2011, 08:26 pm by retrolefty Reason: 1
Quote
The the conduction depends on the voltage on the base. That's the part I didn't understand. I was believing it behaves like discrete 1 and 0 gate with the npn/pnp transistors.

More correctly the collector conduction depends on the current flowing through the emitter base junction times the current gain of the transistor. A npn or pnp transisor is a current operated device, not a voltage operated device. The base current times the current gain (beta value) of the transistor determines the collector (and therefor the load's) current that can flow.

Once the base voltage exceeds the forward voltage drop rating of the base/emitter junction it does not increase any further, it's the base current limiting resistor and the voltage source driving the resistor that determines the base current. Lower resistor, more base current, higher driving voltage, more base current. An arduino has a fixed output voltage when HIGH, so your only handle on controlling the base current and therefor the collector current is the base resistor.

That make sense?

Lefty

a_lehmann

Quote from: retrolefty
That make sense?

Thank you for your detailed explanation. I had to read it four five times, but yes, I get it. And it works great now.
(Don't know if anyone's interested but the IR LED now is so "bright", the reflection from the wall through a half opened door is enough to trigger the reciever... And I'm very satisfied now!)

So thank you very much you both!

retrolefty

Great.

One last point:

Quote
Could I simply hook up the base to the Arduino pin without any resitors?

Definitely not. Once the base voltage is higher then the base emitter junction voltage there is nothing to control the amount of base current flowing from the arduino output pin without a resistor. This will certainly rise to more then the absolute maximum output pin current rating of 40ma and will then proceed to burn out the output pin driver circuitry. Your circuit components are what determines current flow according to ohm's law, so the base resistor is a required component.

Lefty

a_lehmann

Quote from: retrolefty
Definitely not. Once the base voltage is higher then the base emitter junction voltage there is nothing to control the amount of base current flowing from the arduino output pin without a resistor. This will certainly rise to more then the absolute maximum output pin current rating of 40ma and will then proceed to burn out the output pin driver circuitry. Your circuit components are what determines current flow according to ohm's law, so the base resistor is a required component.

In the datasheet of the 2N6488, which I'm using, it says the Current Gain is somewhere between 5 and 20 at 4 V (with Ic = 5 A up to 15 A). So if I want 4 V / 40 mA from my Arduino pin, I'll need a 25 Ohm resistor, right? Than I get like 200 mA through the collector... But I want something around 3 A.

So, does this mean I need a series of transistors until I get the right Collector current? I guess yes. Just asking to be on the safe side.

Is there maybe an alternative component to this? I don't know if you recognized it but at the beginning of the thread I mixed up the characteristics of transistors and relays. But I guess there is nothing like a relay what can switch its states at 13 us for about some milliseconds.

retrolefty

#10
Jul 23, 2011, 10:28 amLast Edit: Jul 23, 2011, 10:32 am by retrolefty Reason: 1
Well I personally now a days only use LOGIC LEVEL N-channel MOSFET transistors for any kind of higher power switching applications.
http://en.wikipedia.org/wiki/MOSFET

A mosfet is a pure voltage operated transistor so you don't have to deal with current gain calculations. A Logic level mosfet is designed to full turn-on with just +5vdc on it's gate lead (with grounded source lead). Because a mosfet gate is insulated from the source and drain leads, there is no current flow through the gate during steady states.

Here is an example of a nice logic level n-channel mosfet with very beefy voltage and current ratings that can be driven directly by a arduino digital output pin.

http://www.sparkfun.com/products/10213

http://www.sparkfun.com/datasheets/Components/General/RFP30N06LE.pdf

a_lehmann

Quote from: retrolefty
A Logic level mosfet is designed to full turn-on with just +5vdc on it's gate lead

Oh thank you sir, this might be just the thing I was seraching for!

Quote from: retrolefty
Here is an example of a nice logic level n-channel mosfet with very beefy voltage and current ratings that can be driven directly by a arduino digital output pin.
http://www.sparkfun.com/products/10213

I couldn't find this part at my mail order of choice (I'm here in Europe) but I'm sure to finde the right thing with the datasheet you gave me.

Thank you very much!

retrolefty

Quote
I couldn't find this part at my mail order of choice (I'm here in Europe) but I'm sure to finde the right thing with the datasheet you gave me.

Thank you very much!

Your welcome. Just be sure you are selecting a 'logic level' N-channel mosfet. Standard mosfets require +10vdc to fully turn on a mosfet and therefore don't work with the arduino's +5vdc output pins.

Lefty

Go Up

Please enter a valid email to subscribe