I know this video here is a year old, I just wanted to say how happy I was to see your channel is still kicking. Thanks for all you do John I have learned quite a bit here
My home turf is the digital realm, but... electronics is electronics and engineers are curious by nature. Even though JFETs are not something I normally work with, thank you for teaching me about them!
Yes I did enjoy that. Thank you John. These JFETs are new to me and you teaching me about them made it interesting, reminded me of a good colledge lecture. Thanks again for the lesson. I believe you have a great manner for teaching.
I can't be the only one that likes these nice simple explanations that don't go off to the stratosphere. Finally. Someone makes a video so easy to understand and grasp it makes it a joy to watch. No super technical stuff. Just a simple introduction with clear and understandable theory and a working model of its operation in easy to understand language and terms. The only thing I could ask for is if you could put in the description a paypal or subscribe star. Not to hot on Patreon after what they did to the canadian truckers that bring in all our goods. Thank you for the wonderful video john.
Thank you. You are exactly on topic, for my present learning curve. You provided a whole lot of practical information that is not in other JFET introductory tutorials = kudos!
Looking forward to amplifier application. A friend built me a jfet preamp in the 90s, from a local electronics magazine circuit. I notice in forums people still regard it highly and warrants rebuilds. Though some people find the sound not to their taste. Love your vids and try not to miss any.
There were never as many different common JFETs available as there were BJTs. I only used the 2N3819 and the very similar MPF102 for many years for the odd application where an unmatched JFET was required. Usually the high input impedance was needed as part of a matched pair and then a JFET opamp was the practical solution. One of the issues with JFETs is the wide range of their key parameters Idss (the drain current at Vgs=0) and Vp (the pinch-off voltage). For example, the both the MPF102 and the 2N3819 have a specified Idss range of 2mA to 20mA, while the Vp range could be -0.5V to -7.5V according to their datasheets. However, I've recently bought 20 cheap J112 JFETs and found them well-matched in characteristics with a Vp around -3.5V. Modern fabrication processes probably produce more consistent results. Or possibly the J111/J112 are roughly tested versions of the same fabrication and "binned" into two variants. In practical circuits, the variation in parameters is usually reduced by negative feedback, the simplest being a resistor between source and ground, but that reduces the output swing available. For tighter control of dc operating points, negative feedback from drain to gate is necessary, but that then reduces the input impedance. Nevertheless, when amplifying a small signal from a high source impedance, the JFET can produce gain with low noise comparable, or superior, to the best bipolar designs.
Just a side note, Actually the bjt is a voltage control device. Many thinks that the collector/emitter current depends on the base current, but that wrong. The Ebers-Moll models show the term Vbe as the most important parameter in the model. Even this Vbe parameter determine the concept of small sigmall on analog eletronics. Other than that, I really apreciate the content you are doing and thank you for doing such videos. Cheers!!!
I agree that Ebers-Moll is a more precise way to model the BJT. Beginners are probably most familiar with base current controlling collector current. While less accurate, it is good enough to make the distinction between BJTs and JFETs in an introductory video.
@@misterjorous The average hobbyist is not interested in Ebers-Moll models or theoretical engineering, and it's just adding confusion to this discussion and it doesn't belong here. There was no discussion about BJTs!
@@acmefixer1 oh yeah, right... I forgot that the average hobbyist union had elected you to decide what should be and what shouldnt be a topic to discuss. If it is making a confusion, I sorry, just sit your god damn ass and study it. Been a hobbyist or not is not a excuse if you wanna talk about something.
Hello, in your picture you show the gate not connected to the source, but on your breadboard you mentioned that you short the gate to the source. What does that do to/for the operation of the FET, in particular the drain to source current flow? Thank you for this video.
An N-channel depletion JFET operates very much like a triode vacuum tube, but at lower voltages and current. Since it is a majority carrier device, the noise will be low.
check out using jfets in the bjt szlikai; cascoding the collector of the driven bjt with the source of a jfet (and the gate to emitter, and drain to the drivers emittor) makes for a very high gain, speedy, and linear small signal gain stage by reducing the miller effect of the driven transistor. connecting the gate of a jfet to the emitter of the driver and its source to the emittor to the driven bjts emittor decreases the early effect of the driver and greatly reduces distortion - especially if the driver is a jfet instead of a bjt! personally im using a bjt as cascode for the driver, since with a jfet as driver the stage cant handle much voltage.
11:20 The drain and source are interchangeable in ALL jfets. The only difference being: 1. In some jfets the P channel is evenly distributed over the N channel (or vice versa in P jfets). For these the datasheet will clearly explicitly state that the drain and source are interchangeable as there is literally no difference between them. They are literally symmetrical. 2. In some jfets the P material is unevenly distributed over the N channel (its volume is centered more towards the source). In these jfets applying the reverse voltage between G-S will much more strongly pinch the current flowing between D and S, whereas applying the same reverse voltage between G-D will pinch the currently much more weakly. But in both cases you can interchange D and S. It's just that the ones that DON'T say they are interchangeable are more easily controlled by G-S than by G-D voltage. And the ones that DO say they are interchangeable are equally easily controlled from both ends. D and S are always internally "shorted" in jfets.
I remember reading somewhere that FETs can switch low-power AC signals. If that's true, reversing the drain and source won't matter, and that may be why it's not obvious what makes drain versus source.
A basic fet is symmetrical (on a silicon layout level), and drain and source therefore inteechangeable (when layouting ICs you can just mirror them and change signals around). They start getting asymetric when you try to max out your specs and tweak every little thin. Additionally, they are not symmetric when the bulk is internally wired to the source
My main power amp has J-Fet output stages with tube driver circuit. A Luxman LV-103 amp from the 80's. Reel fast (nano sec on and off) amp with enormous punch in the bass. Does good at the whole spectrum. Those transistors are not available any more. No plugin replacements exists. I have the spare tubes.
A lot of JFETs are not only expensive but are getting very difficult to obtain. And due to their very wide spread of characteristics you may have to buy 2 or more to find one that works properly in a circuit. Therefore JFETs are not popular and are frustrating for most hobbyists. Try to find a MPF102 or similar RF JFET, then you have to fiddle with the bias voltage to get the circuit working. Thanks and I hope to see the second part soon.
The body diode is an artifact of enhancement mode fets where a doping well is used to make the device have an off state with 0 VGS. The construction of typical JFET is such that there is no parasitic body diode.
"Negative" is simply referenced to the voltage on the source. If the voltage on the source is higher than the voltage on the gate then you have a negative Vgs, even though none of the voltages are below the 0v rail. Adding a resistor into the source causes a voltage drop, i.e. it raises the source voltage. If the gate is tied to 0v the gate is then reverse biased.
@@ferrumignis Of course you need to tie the gate to GND with a big resistor in an amplifier curcuit to have a high load impedance and not just a short. That resistor won't drop any DC voltage because the gate is technically an open curcuit and so is the coupling cap before it (for DC).
Please try TPA3116d2 based board XH-M189 is lower version of XH-M190 please try this. XH-M189 Give 50w + 50w. Please try this board and full test please I am from India. ❤️
@@twotone3070 thanks. yes I asked because being in one place does not necessarily implies one is speaking that place accent. Arnold Schwarzenegger, for example... 😊
I know this video here is a year old, I just wanted to say how happy I was to see your channel is still kicking. Thanks for all you do John I have learned quite a bit here
Thanks for the kind comment!
My home turf is the digital realm, but... electronics is electronics and engineers are curious by nature. Even though JFETs are not something I normally work with, thank you for teaching me about them!
That was an interesting introduction to a component I know very little about. Looking forward to the follow-on video.
Yes I did enjoy that. Thank you John. These JFETs are new to me and you teaching me about them made it interesting, reminded me of a good colledge lecture. Thanks again for the lesson. I believe you have a great manner for teaching.
That was interesting. Looking forward to the simple amp designs.
I can't be the only one that likes these nice simple explanations that don't go off to the stratosphere. Finally. Someone makes a video so easy to understand and grasp it makes it a joy to watch. No super technical stuff. Just a simple introduction with clear and understandable theory and a working model of its operation in easy to understand language and terms. The only thing I could ask for is if you could put in the description a paypal or subscribe star. Not to hot on Patreon after what they did to the canadian truckers that bring in all our goods.
Thank you for the wonderful video john.
L
L
Thank you. You are exactly on topic, for my present learning curve. You provided a whole lot of practical information that is not in other JFET introductory tutorials = kudos!
Excellent explanation John, thanks for all your great work.
Looking forward to amplifier application. A friend built me a jfet preamp in the 90s, from a local electronics magazine circuit. I notice in forums people still regard it highly and warrants rebuilds. Though some people find the sound not to their taste. Love your vids and try not to miss any.
Nice video about these. I take JFET as the closest thing to vacuum tube you can get in silicon.
This was great! Thank you. I’m trying to learn more about Jfets and their role in audio circuits.
2:35 those SiC power JFET's look amazing, makes me think of the old VFET's.
Always a pleasure. Thank you!
Cool explanation, thank you very much!
There were never as many different common JFETs available as there were BJTs. I only used the 2N3819 and the very similar MPF102 for many years for the odd application where an unmatched JFET was required. Usually the high input impedance was needed as part of a matched pair and then a JFET opamp was the practical solution.
One of the issues with JFETs is the wide range of their key parameters Idss (the drain current at Vgs=0) and Vp (the pinch-off voltage). For example, the both the MPF102 and the 2N3819 have a specified Idss range of 2mA to 20mA, while the Vp range could be -0.5V to -7.5V according to their datasheets. However, I've recently bought 20 cheap J112 JFETs and found them well-matched in characteristics with a Vp around -3.5V. Modern fabrication processes probably produce more consistent results. Or possibly the J111/J112 are roughly tested versions of the same fabrication and "binned" into two variants.
In practical circuits, the variation in parameters is usually reduced by negative feedback, the simplest being a resistor between source and ground, but that reduces the output swing available. For tighter control of dc operating points, negative feedback from drain to gate is necessary, but that then reduces the input impedance. Nevertheless, when amplifying a small signal from a high source impedance, the JFET can produce gain with low noise comparable, or superior, to the best bipolar designs.
Just a side note,
Actually the bjt is a voltage control device. Many thinks that the collector/emitter current depends on the base current, but that wrong.
The Ebers-Moll models show the term Vbe as the most important parameter in the model. Even this Vbe parameter determine the concept of small sigmall on analog eletronics.
Other than that, I really apreciate the content you are doing and thank you for doing such videos.
Cheers!!!
I agree that Ebers-Moll is a more precise way to model the BJT. Beginners are probably most familiar with base current controlling collector current. While less accurate, it is good enough to make the distinction between BJTs and JFETs in an introductory video.
@@JohnAudioTech indeed
@@misterjorous
The average hobbyist is not interested in Ebers-Moll models or theoretical engineering, and it's just adding confusion to this discussion and it doesn't belong here. There was no discussion about BJTs!
@@acmefixer1 oh yeah, right... I forgot that the average hobbyist union had elected you to decide what should be and what shouldnt be a topic to discuss. If it is making a confusion, I sorry, just sit your god damn ass and study it. Been a hobbyist or not is not a excuse if you wanna talk about something.
lovely this, I just got me some vintage mpf102's they are golden for the ruby lm386 guitar amp.
I like the sound of Jfet s in amplifier and pedal's especially the 2nd harmonic 🙃
Thank you John
Hello, in your picture you show the gate not connected to the source, but on your breadboard you mentioned that you short the gate to the source. What does that do to/for the operation of the FET, in particular the drain to source current flow? Thank you for this video.
An N-channel depletion JFET operates very much like a triode vacuum tube, but at lower voltages and current. Since it is a majority carrier device, the noise will be low.
Does there need to be a negative threshold voltage from 0v from the jfet to start depleting the channel ?
check out using jfets in the bjt szlikai; cascoding the collector of the driven bjt with the source of a jfet (and the gate to emitter, and drain to the drivers emittor) makes for a very high gain, speedy, and linear small signal gain stage by reducing the miller effect of the driven transistor. connecting the gate of a jfet to the emitter of the driver and its source to the emittor to the driven bjts emittor decreases the early effect of the driver and greatly reduces distortion - especially if the driver is a jfet instead of a bjt! personally im using a bjt as cascode for the driver, since with a jfet as driver the stage cant handle much voltage.
11:20 The drain and source are interchangeable in ALL jfets. The only difference being:
1. In some jfets the P channel is evenly distributed over the N channel (or vice versa in P jfets). For these the datasheet will clearly explicitly state that the drain and source are interchangeable as there is literally no difference between them. They are literally symmetrical.
2. In some jfets the P material is unevenly distributed over the N channel (its volume is centered more towards the source). In these jfets applying the reverse voltage between G-S will much more strongly pinch the current flowing between D and S, whereas applying the same reverse voltage between G-D will pinch the currently much more weakly.
But in both cases you can interchange D and S. It's just that the ones that DON'T say they are interchangeable are more easily controlled by G-S than by G-D voltage. And the ones that DO say they are interchangeable are equally easily controlled from both ends. D and S are always internally "shorted" in jfets.
GREAT video!
I remember reading somewhere that FETs can switch low-power AC signals. If that's true, reversing the drain and source won't matter, and that may be why it's not obvious what makes drain versus source.
A basic fet is symmetrical (on a silicon layout level), and drain and source therefore inteechangeable (when layouting ICs you can just mirror them and change signals around). They start getting asymetric when you try to max out your specs and tweak every little thin. Additionally, they are not symmetric when the bulk is internally wired to the source
@@schabingcraft Cool to know. Thank you for the clarification!
Ah, the good old Fire Emitting Transistor.
Interesting!!
Make a video on how you make those acrylic cases for chip amps.
I definitely plan on doing this.
My main power amp has J-Fet output stages with tube driver circuit. A Luxman LV-103 amp from the 80's.
Reel fast (nano sec on and off) amp with enormous punch in the bass. Does good at the whole spectrum.
Those transistors are not available any more. No plugin replacements exists. I have the spare tubes.
Thank you
A lot of JFETs are not only expensive but are getting very difficult to obtain. And due to their very wide spread of characteristics you may have to buy 2 or more to find one that works properly in a circuit. Therefore JFETs are not popular and are frustrating for most hobbyists.
Try to find a MPF102 or similar RF JFET, then you have to fiddle with the bias voltage to get the circuit working. Thanks and I hope to see the second part soon.
Regarding swapping drain and source might there be some sort of body diode ?...cheers.
The body diode is an artifact of enhancement mode fets where a doping well is used to make the device have an off state with 0 VGS. The construction of typical JFET is such that there is no parasitic body diode.
Interesting.. Im stuck on the "more negative" part. How can their be a more negative potential then the rail? Great video!
Because it may not be the rail. Add a resistor and you get a more positive voltage than the rail.
Like 14:50
"Negative" is simply referenced to the voltage on the source. If the voltage on the source is higher than the voltage on the gate then you have a negative Vgs, even though none of the voltages are below the 0v rail. Adding a resistor into the source causes a voltage drop, i.e. it raises the source voltage. If the gate is tied to 0v the gate is then reverse biased.
@@ferrumignis
Of course you need to tie the gate to GND with a big resistor in an amplifier curcuit to have a high load impedance and not just a short. That resistor won't drop any DC voltage because the gate is technically an open curcuit and so is the coupling cap before it (for DC).
thanks
Good information. especially with the cooling experiment.
this jfet confused my butt for a solid 2 days cause its causing noise on my diy amp mistook it for a common transistor
Please try TPA3116d2 based board XH-M189 is lower version of XH-M190 please try this.
XH-M189 Give 50w + 50w.
Please try this board and full test please
I am from India. ❤️
A jfet is basically a varient of mosfet
OT but can someone who is American describe the accent? It reminds me a certain politician, but I do not want to bring such names in here.
I believe he is in Ohio, but that doesn't mean he's from there. I like how TOT reminds me and others of Alan Alda.
@@twotone3070 thanks. yes I asked because being in one place does not necessarily implies one is speaking that place accent. Arnold Schwarzenegger, for example... 😊
Glue a Jfet to a Bipolar. As one thermally runs away the other automatically clamps it. lol