EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: edy on September 02, 2013, 06:13:11 pm
-
I have been playing with what I thought was a typical transistor that I scavenged from a power supply. I figured the middle pin was the base, and the ends were collector and emitter. So I wired up a +5V source to the collector, and an LED was grounding the emitter. When I shorted the base to ground, I got no LED light (switch off state) and when I gave my base a slight charge it would turn it on and I'd get a light.
Now, I tried hooking up a capacitor to the base, hoping that as it would charge it would reach a threshold and trigger the switch on and LED light. The same cap would also be hooked to collector itself so the trigger would also drain the cap, and cycle would start again. Well, I couldn't get this behaviour and I'm still trying to work out what I'm doing wrong.
Meanwhile, I was rewiring and ended up with this interesting setup which I'm trying to learn about so I know what is going on. The video below shows the results. I have my 5V source hooked to base and the LED from emitter to ground. That should turn it on, but since I have nothing hooked to collector, there is little current flow. If I hook collector to +5V then I get light on, and when I connect it to ground I get light off. But when I touch collector I can toggle the light on and off.
I figure the +5V to the base is leaking to the emitter and eventually it opens to full flow and lights up LED. Somehow various states of the collector alter the path of current or suppresses the leakage. Unless I have the pins all wrong and have no idea what is happening, which is probably the case. I am a beginner so be gentle. :-). I've attached the video below and datasheet for the part.
MOSFET weirdness (https://www.youtube.com/watch?v=NjnXFZ-neMI#)
-
Never assume!
The UTC 1N60A is a high voltage MOSFET and is designed to
have better characteristics, such as fast switching time, low gate
charge, low on-state resistance and have a high rugged avalanche
characteristics. This power MOSFET is usually used at high speed
switching applications in power supplies, PWM motor controls, high
efficient DC to DC converters and bridge circuits.
http://www.datasheetarchive.com/indexer.php?file=DSA00376213.pdf&dir=Datasheet-072&keywords=1N60AL&database=user-highscore# (http://www.datasheetarchive.com/indexer.php?file=DSA00376213.pdf&dir=Datasheet-072&keywords=1N60AL&database=user-highscore#)
-
Yes thank you. I realize that components have very different uses and ratings for different applications. Nevertheless, it is interesting to look at this circuit and figure out why it is behaving that way. I want to explain the results in the video above.
For example, I can make the LED turn on or off, or leave it on, or leave it off by touching it. I would have expected that it would either be ON whenever I was touching it, or OFF as soon as I let go. However, I can leave the LED on if I just let lt go a certain way. Why?
The base is at 5V and is probably at the limit of the base-emitter threshold at which it breaks through and begins to allow current through. Once the current gets going, it keeps going and lights up the LED. What effect am I having by touching the collector/source pin? Could I be affecting the threshold breakdown voltage? And how is it possible that I am able to touch and release but have it turn off, or touch and release and have it turn on?
Perhaps some current also goes reverse bias and from base into collector (my finger) and that temporarily reduces the amount of current flowing to the emitter, so it goes below the breakdown voltage for the base-emitter and stops the LED. That turns the LED off. But then when I touch the collector again, maybe I am now charged up (like a giant capacitor) and feeding in electrons and inducing breakdown again?
-
First, the device is a MOSFET, not a bipolar junction transistor. No base, no collector, no emitter therefore. It has a source, drain and gate. Don't expect it to behave like a bipolar junction transistor because it isn't one.
The gate is insulated from the source and drain and forms a capacitor (gate to source & drain) that can hold a charge for a surprisingly long time. If you charge the gate with sufficient charge of the correct polarity, the MOSFET will demonstrate low resistance between drain and source and may well maintain that low resistance state for considerable time as the gate charge slowly leaks off. The charge may be supplied by a power supply or by static electricity if you touch the gate pin.
-
Thank you for explaining this. The pin I was touching must have been the gate (was at far left, although I expected it to be in the middle). I had a look at the datasheet and indeed pin 1 is a gate. Pin 2 turns out to. A drain and pin 3 is the source.
I had 5V running to the drain pin (middle pin) and actually had my source pin through the LED to ground (at far right). This is backwards but i guess it doesn't matter since it isn't acting as a diode? So when I touch the gate it is altering the capacity, perhaps via static build up I was able to push or draw electrons out of it, affecting the flow of the drain-source current.
Before I was using it like a transistor and it seemed to behave like one. When I confused it for a BJT NPN, I had the far left pin 1 gate labeled collector and hooked up to 5V, my far right pin 3 source labeled emitter and hooked up to the LED and ground, and was able to turn it on by touching the middle pin 2 drain to some positive voltage. Basically, I had 5V going into the gate, and my drain was hooked to ground and my source was actually connected through the LED to ground as well. Basically, competely backwards. The LED would be off until I touched pin 2 to a positive voltage, but it would only turn on half brightness (not what I would expect if it was getting the full 5V).
So, I assume I was essentially powering the LED due to the pin2 to pin3 current flow (drain to source, although backwards) although pin2 voltage alone could not explain the LED turning on, since I could touch it to my hand whilie holding a 5V lead in the other hand... Could there be enough voltage drop from my body to light up the LED? And the gate pin1 needed to be connected to 5V otherwise it wouldn't work. Unless there was some leakage from the gate into the circuit.
-
The gate is open circuit so no current needs to flow. All is happening is it's picking up the charge off your body, most likely the mains. If you touch it and hook up the drain to a 'scope, you'll probably find it's oscillating between off and on at the mains frequency.
The maximum rated gate voltage is only 30V, yet it's possible for you to be charged up to 30kV of static electricity which could zap the gate. You can protect the gate by connecting a 20V zener diode between the gate and source (cathode to gate).
In your previous post, you mentioned attempting to build some kind of oscillator but it won't work with a single transistor and only one capacitor because you won't have enough phase shift. You can use two transistors, a transformer or three RC networks to get enough phase shift although the latter may not work with a single MOSFET as you might not have enough gain.
-
Thanks for the explanation Hero. I do have some Zeners that I salvaged, I could tell by the component tester on the scope as I managed to get the classic shape with breakdown at both ends (rather than just an "L" with a regular diode). The MOSFET also had the same shape as a Zener, it must have been allowing current flow in either direction (drain-source) once it was over the gate threshold? Until I downloaded the datasheet I just treated it like a transistor, hence my confusion.
Anyways, will have to pick up some "real" transistors and caps from my local supply shop. I thought I could salvage a few bits from scraps at home, but it will be too hard for a beginner like me to work with. I'm going to try both the 2-transistor version oscillator and also a 555 timer version from specs I've seen online. But for that, I'll need to get the proper parts with known values and just follow step-by-step instructions from known working schematics.
-
If you look at the data sheet, you will note a parasitic diode is shown between source and drain - this is very common for MOSFETs. If you connect the source and drain such that the diode is forward biased, then current will flow through the MOSFET device, and the gate no longer will do anything to control the MOSFET. This is NOT the normal connection of course.
Also, the gate voltage is, by convention, defined as between the gate and source. Vgss.
And, it's very easy to build an unwanted inadvertent oscillator with a MOSFET when breadboarding a circuit with long leads on a plug board. (I'll tell a story on myself about that ... I was trying to measure the switching time and some other parameters for a TO-220 packaged MOSFET a few years ago and I didn't have time to do a proper test fixture, so I threw one together with some clip leads and plug board. I happened to have the shortwave receiver running in the background to listen to the 7 MHz amateur radio band, and as I was running the power supply voltage up and down (the one connected to the drain load resistor), I heard a loud buzzing signal sweep by the frequency I was monitoring. With a careful tweak to the power supply voltage I was able to tune the unwanted parasitic oscillation up and down the 7 MHz band. The oscillation, of course, also meant that the data I was trying to collect was bogus, I did what I should have in the first place and made up a proper fixture with appropriate short lead bypass capacitors, etc. No parasitics and good data.)