How can I properly select and use a transistor for a motion sensing device?

[kerputer]

Hi Everyone,

First of all, thanks for such a great resource. I have been lurking for a bit and just made an account to ask for help with a little project I am working on.

What I am trying to accomplish is a motion sensing device that performs a function when something enters the detection space. To be a bit more specific;

I have a 5.5v power source
I have a IR sensor that runs on 5.5v and depending on distance of object in its LOS, outputs between .3-3volts (for this use anything that enters will be minimum of 1v output)
I have a motor that runs on 5.5v and draws about 300mA while running (Can not find stall amperage)

I need to find the correct components to put into all of this and make it work correctly. I have some random Transistors laying around and got one to work, but as a stress test held the circuit closed for 2 minutes and burned it out. I am a novice with identifying Transistors, but even looking around online could not find exactly the ratings for the one I was using (It says 8050 D128).

I know this was not the "right" way to do this, but this is what I tried that had success.

5.5v + went to Collector
5.5v + went to IR + terminal
5.5v - went to IR - terminal
IR output went to Base
Emitter output went to motor + terminal
5.5v - went to motor - terminal

I know it was wrong to not use any resistors, but I didnt have anything that worked on hand (i tried a 1k between 5.5v+ and the collector, did not work)

This transistor seems to not be reaching full saturation as the output voltage is around 2.5v (but still runs the motor as proof of concept) and when I kept the my hand in the sensor for 2 minutes to run the motor continuously, the transistor became incredibly hot, and was damaged (permanently closed circuit regardless of what's at the base)

I tried to include as much information as I could, please let me know if I can provide more information that could be helpful.

Thanks!

 

[Simon Bridge]

S8050 are a family of small-signal pnp transistors aren't they?
If you just want to use it as a switch, you could try a PNP power transistor.
Only this time, don't stress-test until you know you have a replacement handy.
Current limiting is your friend here though.

It is so hard to find a "proper" transistor these days - they are so base: hardly any manners at all. I blame their diodes.

 

[meBigGuy]

Basically you are using the transistor as a darlington. The voltage out of the IR probably doesn't reach 5.5, so there will be significant voltage drop across the transistor, which means power consumption.

300ma is way high for a small signal transistor. Even if it is saturated it would be > 1 watt, and lots more during startup and a stall. I think you want a power FET. Either a pchannel to 5.5 or an N channel to ground. Look up fet switch or transistor switch on google images. Whatever you choose, determine the power consumption and whether you need a heat sink. A part in a to220 like package may work without a heatsink.

Put a reversed diode across the motor or it will blow things out when you remove power.

 

[Simon Bridge]

@kerputer: any of this useful yet?
I've been guessing you know your circuit needs a redesign right?

 

[kerputer]

Hi Simon,

Yes it is helpful. I know the basic functions of many of the little components I am using, but I learn every day how poorly equipped I actually am to design a quality circuit. I run mainly on trial and error :). And the S8050 I had was NPN small signal transistor, and from my experience with it, could not handle the amperage pull of the motor for any duration over probably 3 seconds (it starts getting hot really fast, 3 seconds is to hot to touch).

That being said I do need to rework everything, my next version will run a diode across the motor to help dissipate the stall amperage spike to not damage the circuit. And I am going to attempt using a TIP31C power transistor as it appears to fit the requirements of the project. One thing I could really use clarification on is how to discern what voltage/amperage is required at the base of any transistor in order to close the gate and make the circuit closed. Since I have such a low voltage and incredibly low amperage out of the IR device, and what runs the motor is such a relatively high voltage and amperage, I am not sure how to pick a suitable component.

I think I will begin teaching myself a bit about Arduino and microcontrollers, where I can just program the functions I need, but until then I am still stuck in a hardware world.

Thanks!

 

[Simon Bridge]

In my day we had this thing called the Radioshack Engineers Cookbook - covered integrated circuit electronics.
Transistor stuff was my dads generation - there were similar books for that, for switching to radio circuits.
They were non-academic resources for people who wanted to just use the components.

Looking around today... it does seem that microcontroller programming is what people are doing so I'd encourage that direction. If you need to know more about transistors and other active components, then you want to hunt for information at your education level... i.e.
http://www.marulaberry.co.za/index.php/tutorials/electronics/beginners-guide-to-transistors/

For Op-Amps:
http://www.societyofrobots.com/robotforum/index.php?topic=9840.0

How the transister responds to the base bias depends on the transister - you have to ask the manufacturer.
Usually these things are published in tables - google the transister name.

 

[meBigGuy]

TIP31C lowest Beta at 1A is 25. That means that for a 1 amp collector current you will need to drive 1A/25 = 40ma which will probably be an issue. But you are headed in the right direction. At one amp it will dissipate a watt or so. You can connect a small signal transistor as a darlington to get the drive current you need but then you need more voltage to turn it on.

If you really want to turn on with 1V output, you will need to use a comparator or something as a level shifter.


But, You are much better off with an P channel power FET and a driver transistor. You drive your IR into a NPN transistor that turns on a P channel FET.

http://www.onsemi.com/pub_link/Collateral/AND9093-D.PDF explains it. Look at figure 3. Q1 could easily be a transistor. A transistor is better for you because the turn on voltage will be lower.

An OK PFET would be the http://www.fairchildsemi.com/ds/FQ/FQP47P06.pdf You may be able to get away with a smaller device.

 

[ScienceGeyser]

Transistors tend to fail shorted. For this reason it is always a good idea to use some sort of current limiting so you don't toast other components in the circuit. Since the motor will run directly on the supply voltage, there is no need to put any current limiting in series with it. The output of your IR sensor, on the other hand, may not be able to handle a high current and you should probably put a small (330 ohm or so) resistor in series between it and the driven circuit.

If you are using an NPN transistor, you should put it between the load and ground. If you use a PNP then you will want to use it between the supply and the load. In the case of the NPN, a voltage below ~0.5V will keep the transistor off. As the voltage increases and current starts to flow thought the base, the C-E series resistance will lower. You are not likely to get more much than 1V on the base before the transistor saturates. If you connect the output of the IR sensor directly to the base of a transistor and the output is capable of 3V open circuit, then you are most likely exceeding the output current of the IR sensor and need another transistor to amplify the output current. If you are unable to supply the necessary current to the base of a bipolar junction transistor (BJT), then you will never get it to reach saturation.

If you use an NPN transistor to control the supply side of a load, then your base voltage will "float" with the load voltage drop. If there is no current flowing through the motor and the voltage drop is zero, then you can turn on the transistor with 0.6V. But, as current starts to flow and the motor starts to turn, back EMF in the motor will cause the voltage drop to increase which will require that the base voltage also increase in order to maintain the base current. Once the voltage drop across the motor reaches about 0.6V below the supply voltage, you will not be able to drive more current and the system will be in some arbitrary current limiting mode.

Field Effect Transistors (FETs) are a great choice for switching controls like are most often used on motors but if you are trying to "get er done" with "what you got", I guess that might not be possible. FETs allow you to drive circuits with voltage rather than current but usually require a larger voltage change than do BJT's. FET's can be driven directly in most cases but it is still a good idea to use a series resistance on the gate to protect the driving circuit.

The TIP31 should work for you if you are able to drive enough current through the base to get the motor current that you require. Remember that the motor current will increase as you load it and by the time you lock the rotor you are drawing several times the unloaded motor current.

Also, if you look at most transistors you will see that they say that 150C is the absolute maximum junction temperature. What they often neglect is the time it takes to dissipate heat to the outside of the case. Usually, by the time you think the transistor is too hot, you've already damaged the junction. Attaching a heat sink will give you some more time before total destruction.

My apologies if that was to windy.