This video shows how a little basic knowledge can be used to get all the information you need to characterize most components, even unknown ones. Although I do recommend people buy one of those cheap ($7 to $15) component identifier/testers. Even with my bench full of good test and measurement I find the little tester very quick, handy, and reasonably accurate to sort through new, surplus, and pulled parts. It's amazing how cheap useful test equipment has become.
The trouble with those is there's a ton of clones, and not all the clones support the more advanced firmware that has some extra functionality in it. It's nearly impossible to tell what hardware you'll get (and thus if it's compatible with the extended feature firmware) until you order it and open it up yourself these days! I'd really like to get one, but got bogged down trying to figure out which one to get, and the consensus now is that it's impossible to tell...
This sort of stuff is great and there isn't nearly enough anymore. Stuff like Jerry Walkers channel, he's doing a step by step design of a Switched Mode PSU. It's awesome to get walked through analogue electronics I personally am not enjoying MCU's as much as I did so I found this video very refreshing and would like to see more and of course "everybody loves chip of the day ! "....cheers
I'd recommend watching the current as you turn the knob and go in the increments of current, up to some max that is safe for the device, noting the voltage at each increment. This way you are less likely to overheat it and will get more useful data points (due to the non-linearity of the I/V curve).
I recently did the same with some OpAmps. Kept them in a very basic circuit with only two pots to set the gain and running on 2 x 12V batteries. No reactive components except the ICs. Set there gain to x10 (20db) at 500Hz. Increased the freq, but left the gain controls alone. Graphed the results. Unity gain for my TL071 was about 3MHz as it should be. OPA637s and 741s were rubbish at 200kHz and 100kHz. Oh well you get what you pay for. Least I know what to throw out. Interesting exercise in spite of being a bit disappointing. Might do the same for other OpAmps in my tray.
At 2:57 "... a maximum of 1 amp, so we're not going to blow up anything, okay?" Not okay. Did you miss the part of the datasheet (see 0:32 ) that states Id max = 500mA continuous? Or the part that gives maximum power dissipation of 830mW (since a 10V supply and a 10R load will allow a worst-case power dissipation in the mosfet of 5V x 0.5A = 2.5W)? At 7:40 that wasn't just the resistor getting hot, it was the mosfet passing 450mA with 5.5V across it, thus dissipating almost 2.5 W in a TO-92 package! That will give you a "junction" temperature around 400°C. Small signal mosfets are meant to be used with moderate currents. Use a 100R 1W resistor and measure up to Id = 100mA. It's just as useful a demo of the characteristics.
Pfft, datasheets - you only need to follow those if doing designs. This is more like destructive testing / education / experimentation. How do you think they come up with the datasheets in the first place on a new design? They basically do it like this (but with more automated test setup). Following the datasheets too closely leads to very boring circuits with a distinct lack of smoke, fire, and excitement.
@@gorak9000 What you say is accurate if you only design circuits to blow up and set on fire. However, for the rest of the world (aircraft industry excepted) ...
@@RexxSchneider You missed my point entirely - the purpose here is not to build circuits, it's to characterize the device. You want to see what the device can and can't do. You're not designing a circuit that you want to still be working in 100 years, you're pushing the limits of the device.
@@gorak9000 Yes, you're right: I missed your point entirely. I thought the point of the video was to build circuits. I didn't realise the point was to ignore what the manufacturers who wrote the datasheets had already told us and to independently discover what it takes to destroy a particular sample of the device.
Nice experiment, just you have to be fast enough. In the setup with 10V and 10 Ohms you pass through the point of 5V 0.5A which is 2.5W at the transistor. I was watching the drift of the output voltage to see when the transistor will give up and send a smoke signal, but it was finished soon enough and it survived. I propose tho do the experiment with a 100 Ohm resistor.
Cool, would you do that again vs BF256, MMBFJ202, comparison mosfet vs fet, to see how mosfets differ from fets. The simulator shows you can build differential amplifier with mosfets, but they use fets inside opamps.
@@simontay4851 No, it's not MOSFET = "Metal Oxide" Semiconductor Field Effect Transistor and have a body diode. The sell both types at mouser and so within their own category.
@@simontay4851JFETs are different from MOSFETs. The MOSFET demonstrated here is an N-Channel MOSFET which can be turned on with a positive gate signal relative to Source. An N-Channel JFET (Junction Field Effect Transistor) works a bit different. When the Gate is at the same potential/voltage as the Source the transistor is fully switched on. To switch the JFET off the gate must be pulled negative compared to the Source. Obviously with P-Channel FETs everything is turned on its head.
Nice! At $20/meter you might as well have 2! That will make your measurements a bit quicker since you don't need to move your test leads around so much. Just twiddle the pot and read off the voltage across your test resistor!
Remember that when the gate is on enough to draw .5A through the drain-source, the voltage across the drain-source will be 5V. 5V times .5A equals 2.5 watts. I'm pretty sure that the device was not rated for 2.5 Watts (at least not without a substantial added heat sink) and that the device was what you noticed getting hot. I'm kind of surprised that it survived the experiment.
I don't have any idea what a 4 quadrant Triac is. But I can answer the second question. the 4 quadrants are voltage/current +v+i +v-i -v+i -v-i ua-cam.com/video/qFVhe_uzxnE/v-deo.htmlsi=MQDmHbTZhPJZAYBf
Not generally. For higher frequency RF current, this begins to be the case. For DC, not. It's call "skin effect". As a ham radio operator using sketch antennas in the field with a lot of RF "floating around" on the antenna cables, the other "skin effect" describes the RF burns you can get as the RF prefers you as the path to ground.
The maximum current through the transistor occurs at the lowest VDS. Power is just the voltage times the current through/across the junction. If the drain and source had a voltage of say 5V at 1A, the power would be a not-so-good 5W.
Nice demo, but I think it would have been much better if the measurements were made abruptly. Constant current of 500mA at 5 Volts , that's 2.5 watts. Way too much for a little TO92 . The internal junction temperature must have raised near max. And , DMM readings do not reflect measurements for a fixed temperature, as the datasheet graph is proposing.
You could also use an Arduino to control the voltage supplied to the Gate, and to measure the Gate-Source voltage. Step trough the voltages in small increments, and output the results as a csv-file that you can open in Excel and make graphs out of. Basically do the same as a CurveTracer do, just a lot cheaper.
For us that don’t have the basics down this is a very helpful format. We can transition to more technical videos as we learn. Thank you!!!
Love this kind of stuff. Makes me appreciate the time savings provided by fancy tools.
The FET is getting hot 🔥 Lol.
This video shows how a little basic knowledge can be used to get all the information you need to characterize most components, even unknown ones. Although I do recommend people buy one of those cheap ($7 to $15) component identifier/testers. Even with my bench full of good test and measurement I find the little tester very quick, handy, and reasonably accurate to sort through new, surplus, and pulled parts. It's amazing how cheap useful test equipment has become.
The trouble with those is there's a ton of clones, and not all the clones support the more advanced firmware that has some extra functionality in it. It's nearly impossible to tell what hardware you'll get (and thus if it's compatible with the extended feature firmware) until you order it and open it up yourself these days! I'd really like to get one, but got bogged down trying to figure out which one to get, and the consensus now is that it's impossible to tell...
This sort of stuff is great and there isn't nearly enough anymore. Stuff like Jerry Walkers channel, he's doing a step by step design of a Switched Mode PSU. It's awesome to get walked through analogue electronics I personally am not enjoying MCU's as much as I did so I found this video very refreshing and would like to see more and of course "everybody loves chip of the day ! "....cheers
I'd recommend watching the current as you turn the knob and go in the increments of current, up to some max that is safe for the device, noting the voltage at each increment. This way you are less likely to overheat it and will get more useful data points (due to the non-linearity of the I/V curve).
I recently did the same with some OpAmps. Kept them in a very basic circuit with only two pots to set the gain and running on 2 x 12V batteries. No reactive components except the ICs. Set there gain to x10 (20db) at 500Hz. Increased the freq, but left the gain controls alone. Graphed the results. Unity gain for my TL071 was about 3MHz as it should be. OPA637s and 741s were rubbish at 200kHz and 100kHz.
Oh well you get what you pay for. Least I know what to throw out. Interesting exercise in spite of being a bit disappointing. Might do the same for other OpAmps in my tray.
That's how we did it in high school.
The transistor is getting too hot to get a good reading - it changes specs too !
At 2:57 "... a maximum of 1 amp, so we're not going to blow up anything, okay?" Not okay. Did you miss the part of the datasheet (see 0:32 ) that states Id max = 500mA continuous? Or the part that gives maximum power dissipation of 830mW (since a 10V supply and a 10R load will allow a worst-case power dissipation in the mosfet of 5V x 0.5A = 2.5W)? At 7:40 that wasn't just the resistor getting hot, it was the mosfet passing 450mA with 5.5V across it, thus dissipating almost 2.5 W in a TO-92 package! That will give you a "junction" temperature around 400°C.
Small signal mosfets are meant to be used with moderate currents. Use a 100R 1W resistor and measure up to Id = 100mA. It's just as useful a demo of the characteristics.
Part of the instructional process to gain acquaintance with the "hot electronic smell" that ever experimenter should be aware of.
Pfft, datasheets - you only need to follow those if doing designs. This is more like destructive testing / education / experimentation. How do you think they come up with the datasheets in the first place on a new design? They basically do it like this (but with more automated test setup). Following the datasheets too closely leads to very boring circuits with a distinct lack of smoke, fire, and excitement.
@@gorak9000 What you say is accurate if you only design circuits to blow up and set on fire. However, for the rest of the world (aircraft industry excepted) ...
@@RexxSchneider You missed my point entirely - the purpose here is not to build circuits, it's to characterize the device. You want to see what the device can and can't do. You're not designing a circuit that you want to still be working in 100 years, you're pushing the limits of the device.
@@gorak9000 Yes, you're right: I missed your point entirely. I thought the point of the video was to build circuits. I didn't realise the point was to ignore what the manufacturers who wrote the datasheets had already told us and to independently discover what it takes to destroy a particular sample of the device.
Nice experiment, just you have to be fast enough. In the setup with 10V and 10 Ohms you pass through the point of 5V 0.5A which is 2.5W at the transistor. I was watching the drift of the output voltage to see when the transistor will give up and send a smoke signal, but it was finished soon enough and it survived. I propose tho do the experiment with a 100 Ohm resistor.
Did exactly the same : view the drift and know that the transistor was feeling pretty hot...
Cool, would you do that again vs BF256, MMBFJ202, comparison mosfet vs fet, to see how mosfets differ from fets.
The simulator shows you can build differential amplifier with mosfets, but they use fets inside opamps.
FET is short for MOSFET. So a FET the same thing. There is no vs.
@@simontay4851 No, it's not
MOSFET = "Metal Oxide" Semiconductor Field Effect Transistor and have a body diode.
The sell both types at mouser and so within their own category.
@@simontay4851JFETs are different from MOSFETs.
The MOSFET demonstrated here is an N-Channel MOSFET which can be turned on with a positive gate signal relative to Source.
An N-Channel JFET (Junction Field Effect Transistor) works a bit different. When the Gate is at the same potential/voltage as the Source the transistor is fully switched on. To switch the JFET off the gate must be pulled negative compared to the Source.
Obviously with P-Channel FETs everything is turned on its head.
Nice! At $20/meter you might as well have 2! That will make your measurements a bit quicker since you don't need to move your test leads around so much. Just twiddle the pot and read off the voltage across your test resistor!
There is no such thing like too many multimeters!
@@yakovkonovalchukov2400
True 👍
Remember that when the gate is on enough to draw .5A through the drain-source, the voltage across the drain-source will be 5V. 5V times .5A equals 2.5 watts. I'm pretty sure that the device was not rated for 2.5 Watts (at least not without a substantial added heat sink) and that the device was what you noticed getting hot. I'm kind of surprised that it survived the experiment.
No current flows in the gate in static conditions on a MOSFET. The current is flowing in the drain-source channel.
@@misterhat5823 Yes and that drain-source channel current is .5A and it has 5V across it so it's still doing 2.5W and getting really really hot.
@@seadon99 Yes. But, it's not the gate dissipating that power.
Can I use the BS170 instead of a JFET like BFW10 or something ?
Could you do something like this but with a 4 quadrant Triac and how and why there are 4 ?
I don't have any idea what a 4 quadrant Triac is. But I can answer the second question. the 4 quadrants are voltage/current +v+i +v-i -v+i -v-i
ua-cam.com/video/qFVhe_uzxnE/v-deo.htmlsi=MQDmHbTZhPJZAYBf
Is pretty hard for TO92 to put 2.5W in for many second!!!😧
But the setup work... 🙂
Why is it called "transfer characteristics"?
things go from the input to the output
Is it true that current move only on the surface of conductor.please I need answer
Not generally. For higher frequency RF current, this begins to be the case. For DC, not. It's call "skin effect". As a ham radio operator using sketch antennas in the field with a lot of RF "floating around" on the antenna cables, the other "skin effect" describes the RF burns you can get as the RF prefers you as the path to ground.
I'm surprise that he didn't blow his transistor. If the big resistor become hot how come the small transistor didn't blow up?
The maximum current through the transistor occurs at the lowest VDS. Power is just the voltage times the current through/across the junction. If the drain and source had a voltage of say 5V at 1A, the power would be a not-so-good 5W.
Nice demo, but I think it would have been much better if the measurements were made abruptly. Constant current of 500mA at 5 Volts , that's 2.5 watts. Way too much for a little TO92 . The internal junction temperature must have raised near max. And , DMM readings do not reflect measurements for a fixed temperature, as the datasheet graph is proposing.
Anybody else hear FET and think of a JFET and not a MOSFET?
Doing this procedure by hand a few times will incentivize just about anyone to run out and buy a curve tracer...
Except they're called SMU's now, cause we're fancy like that
V vds ❤
You could also use an Arduino to control the voltage supplied to the Gate, and to measure the Gate-Source voltage.
Step trough the voltages in small increments, and output the results as a csv-file that you can open in Excel and make graphs out of.
Basically do the same as a CurveTracer do, just a lot cheaper.
And it’s nice that you can control the granularity to about 12-bit places.
There are also some nice peripheral DACs with low impedance outputs and very high resolution too.