2:00: sensitivity, voltage amplification. The preamp is to keep the signal level above the background noise level and match the next stages input impedance Power output gain is for the power amp. RS is negative feedback gain stability.Thanks for taking the time to make your presentation.
This is pretty much how most of studio condenser microphones operate. + something like 10:1 output transformer for low output impedance. Rg in this case is by default something like 1Giga Ohm. Many think this value is a mistake in schematics. The signal is produced by a variable capacitor (microphone capsule) which varies something like 1pF in capacitance, and produces signal. The circuit is often called "impedance converter", as there isn't much amplification going on.
Excellent series John! Unfortunately we didn’t spend much time in my Electronic Devices class on JFETs. Most of the class was spent on BJTs and MOSFETs. Thx for your presentation!
The way you went at it as just fine. From the output impedance, calculate your drain current, then from your estimated Vgs, estimate your source resistor. Then make up the circuit and find the proper value of for Rs. There's really no way around that with a JFET because of the wide variance of Vp between samples. One issue to note: at 13.46 you annotate with the legend "1.4 seimens in this example". However, you had a hypothetical (2.3-0.9)mA / (1.5-0.5)V which works out to 1.4 *milli-Siemens,* and that's quite an important difference! The reciprocal transconductance is then 714 ohms, which of course is bigger than your hypothetical value of Rs. Also, whatever voltage you drop across Rs becomes unavailable for the negative-going output swing and reduces your headroom. This can become significant with low supply voltages and JFETs with high Vp. For a J111, it can be as much as 10V. Now, in the real case, your MPF102 has a higher Vp than your hypothetical example, hence the higher value of Rs required, and its gm is also higher, maybe around 5mS at currents around a couple of mA. Nevertheless, it's unusual with small-signal JFETs to be able to ignore the transconductance when calculating the expected gain. By the way, the gm varies with drain current in a quadratic relationship, giving some mild second harmonic distortion when 1/gm is comparable to Rs (but not the third harmonics you get with BJTs when re is comparable to Re). The only fix is a higher supply voltage, and that allows more gain (which is effectively capped at half the supply voltage divided by the JFET's cutoff-voltage), which you can then sacrifice by increasing Rs at the same Id.
I've found that an optimum Q point for a JFET amplifier lies around 0.382 Vp, which corresponds to 0.382 Idss (magically). With that Q point Rs is simply Vp / Idss, and Rd ~= (Vdd+Vgs) / (2*Id), where Id = .383 Idss. Vp and Idss of a given JFET can be easily obtained with a simple measuring circuit which is available on the web.
Hey John, I'm not sure if you've ever checked out any of the depletion-mode mosfets available or not but, I ended up acquiring a few higher-power ones and they have really great specs and sound great for power amplification. The ones I picked up are the IXYS "IXTP6N50D2" which is a NPN in a TO-220 package and specs like Vds: 500V, Id: 6A, Vgs: -20V, +20V, Rds On: 550 mOhms (YES! milliohms!), Pd: 300W, etc. Go to Mouser (or your favorite place) and check out the full specs. They aren't expensive for what you get IMO at around $7 each. There's not many depletion mode mosfets but, you can still get ones with even higher power dissipation like 800 watts! They sound much better than the "enhancement mode" FETs IMHO. Seriously go check out the specs. I would share a link to the PDF but, TH-cam is getting weird with them.
Ya, you need to be cautious when selecting an opamp, jfet opamp's don't need a resistor when used as voltage follower, but transistor based input opamps do.
Good video. I would think you would want to characterize your jfet before doing your calculations. I would find the pinch off voltage and the Odds, then draw the graph. After that your calculations will be much closer to the real world response.
Hello sir, since you are now working on amplifiers and preamplifiers i would like to ask you? you may be able to show a microphone amplifier with time. ( schematic ) to make a condenser microphone preamp with: 1: maximum gain 2: as little sensitive as possible to high frequency Friendly greetings from The Netherlands! Rob
There will be a small bias current flowing out of the gate of the n-type JFET. The bias resistor has to be sized to keep the resulting voltage within a few tens of mV of ground...
True, but as it's only of the order of 1 or 2nA at room temperature, that is still only 10 or 20mV with a 10M resistor. There are probably more compelling reasons for keeping the gate resistor as low as is needed to set the input impedance, such as the problems of picking up stray noise at a high impedance point.
@@RexxSchneider Yes, 1 or 2 nA at room temperature, but increasing exponentially at higher temps. In sonar applications with small piezoelectric sensors (~10nF PZT ceramic capacitors), the bias resister is a significant low frequency noise source and reducing the bias resistance increases this noise, so values between 20Meg and 50Meg are most common. I wouldn't be surprised if this value was lower in other applications.
Ive messed around with making a jfet hifi preamp, never could get it lower than .05 percent distortion without it loosing its little bit of tube vibe, in the end its just easier to get the tube sound with tubes 😁
how about the following amplifier/dc-power-source: transformer (volts up) -> capacitive voltage divider (tunable amplifier output level, dc-remover) -> full bridge rectifier + smoother capacitor (dc-output stage), should be very tunable and very frequency independent, linear amplification, dont know what class amplifier that is, but no transistors required
Howdy. In my mind there was an error about impedances. When the impedances match half the power is consumed in the driver transistor + driver resistor and half in the load. The voltage drop is 1/sqrt2, not ½ which is -3dB. Regards.
No, it is 1/2 as stated (not exactly 1/2 in a real circuit because the transistor is not a perfect device). Remember the transistor is functioning as a current source (or sink in this case) with very high drain impedance, therefore Rd is the output impedance. If you need confirmation, build the circuit and measure it.
What year did you buy that MPF102, and do you have any recommendations for current production jFETs? All the old standbys are obsolesced and getting rarer than hens' teeth. Most of the available through-hole jFETs are dirty-grade choppers. Miller effect on g-d capacitance at 12V is negligible at audio frequencies. It's significant in triodes, which have higher g-a capacitance and operate with anode voltages an order of magnitude higher; in fact some designers like to add an external capacitor from gate to drain on jFET stages to approximate the Miller capacitance of triodes.
The MPF102 hasn't been available from Mouser for some time. You can still get the similar 2N3819 from them, but I tend to baulk at paying around £1 for a single small-signal transistor. On the other hand, it's near-impossible to design a sensible bias point for these JFETs as their Vgs @0.2mA is specified as somewhere between -0.5V and -7.5V. Try selecting Rs to cater for that without applying negative feedback from drain to gate (which is actually the only sensible solution to get a repeatable quiescent drain voltage). I've been experimenting lately with some of those dirty-grade chopper JFETs, specifically the J111/J112, which are cheaply available, and I've been impressed with the tightness of parameters across the samples I've tried so far. I still doubt that I'll ever find an application for them as amplifiers, though, as modern JFET opamps are far easier to design with, and give as good performance with high-impedance sources as any discrete design.
@@RexxSchneider Only reason I can think of for using a discrete jFET instead of a FET-input op amp is to introduce some of "that sweet, sweet harmonic distortion" by operating them in their triode region (the part of the response curve described in the video) to simulate a triode gain stage. In the case of the J11x, this is typically less than 2V, so you'd need really low system noise to get anything out of it. There are other ways of introducing this distortion (e.g. a diode and resistor from output to ground) that are more tuneable and less dependent on the transistor's individual characteristics and can be used with op amps, so there's another nail in the jFET gain stage coffin. Still, it would be nice if anyone was still making the MPF-102 or J201, if only for the people who like these legacy designs.
@@AnalogDude_ I have some 2N4085 from many years ago when I needed a high impedance input with minimal temperature drift. That was before we had JFET opamps, and they were all that was available to do the job. I see you can still get them for $20+ apiece, but they don't really make sense anymore when an OPA2134 is half the price.
@@petersage5157 The higher the Vp, the lower the maximum stage gain for a given supply voltage, so using JFETs with higher Vp doesn't actually offer any improvement in noise margin. The 2N3819 is a drop-in replacement for the MPF102 with near identical characteristics and it's still currently manufactured by Central Semiconductor Corp.
Hi John. Could you please do the same video with a MOSFET? Sorry if you did allready. Could you please point me to it in that case? By the way: it is "Siemens" not seimens.
You have a problem again with the variability of the threshold voltage across samples of a MOSFET. I suggest using negative feedback from drain to gate and a significant source resistor to stabilise the voltage at the drain, which means you'll need a fairly high supply voltage. Try using a 15V supply, with Rd=4K7 and aim for Id around 1mA. Then connect 100K from drain to gate and 100K from gate to ground. That will set roughly 10V at the drain and 5V at the gate, so the source will be a couple of volts below that for a small-signal MOSFET like the 2N7000. You'll then need a source resistor around 3K, but you can split it and bypass the larger part to give some ac gain while retaining dc stability. Perhaps try 150R + 2K7, with the 2K7 bypassed with 220μF. Make the 150R bigger to decrease gain and improve linearity.
@@thomasr.5443 You're welcome. If you mean his video using a power MOSFET as a class A amplifier driving a speaker ( th-cam.com/video/D4tHgHfljyY/w-d-xo.html ) then you can see the problems of setting a bias point which he tackles by using a trimmer resistor. Obviously that's a suitable solution for the hobbyist, but doesn't scale well for production. You can of course apply the same principle of trimming to a low-power pre-amp design using something like the 2N7000, and you should be able to get x10 gain with low noise and decent linearity even from a high impedance source. For low impedance sources, I still recommend using a BJT like the BC549C.
8:05 Would you be so kind to explain, how the Ohm's formula is used to calculate voltage drop over a resistors, i forgot how it's done. 12V - 3K3 = 6 volt? 12÷3300 = 0,003636364 Amps. 12÷3690 = 0,003252033 Amps (if the fet wasn't there). Roland uses something similar in their System 100M VCA (yusynth). Rg = 33K, no Rd.
Voltage drop across a resistor is a function of current, I*R=V. In the video, I determined what I needed the resistor value (3k3) and the voltage at the output node (6v). From that, I needed the current, V/R=I
@@JohnAudioTech Thank you very much for your answer, but ... 12V÷3300R = 0,003636364 Amp × 3300R = 12V again. using a "voltage divider calculator" assuming the FET aint there: 3K3 and 390R = 1.268 Volts, so the voltage drop would be 10,732 V or am i missing something? i tried your circuit in the Falstad simulator: tinyurl/2bg4httk, normally it works perfect, accept i don't calculated the beta for the JFET you're using.
@@AnalogDude_ From the supply (12v) to the output node (source) I needed 6v. 12-6=6v, therefore 6v across Rs (3k3) which results in a current of 1.8ma. Now that I know the idle current the transistor needs to conduct at, I can look at the graph to find the gate voltage needed for 1.8ma. It was around -1.2v for the real device, so I need 1.2v drop across the source resistor (Rs). 1.2/.0018=666 (680 is closest common value). Beta is not the correct term for JFETs. Were you calculating the voltage gain of the circuit?
@@JohnAudioTech the Fallstad simulator has a calculator for that of you click on the JFET part. I'm looking at the graph of the 2SK30A at the same graph you used in the video, it has multiple plots, would it be the same -1.2V?
@@AnalogDude_ John did explain that JEFTs tend to have quite a wide spread of Vgs values, which is why you can only get a ballpark value using the nominal datasheet values. For a real circuit where biasing is important it's likely you will have to modify the resistor values to correct the biasing.
Newby, pertnear 70, gave up on electronics in HS because the study of logarithms didn't harmonize with puberty :-). A mental block for me now is the concept of impedance (type of resistance or "thirst" or combination?). It seems directly related to the resistance at an input(?) but what the hell is "output" impedance? Seems a contradiction. Oh well, maybe it'll come to me if I live long enough :-) since I'm not very studious. Flashing lights, sirens and motors are kinda fun. I don't think I'll be making any breakthroughs :-)
Output impedance determines how much current the circuit can deliver into a load., imagine a resistor in series with a perfect voltage source. If a circuit has a low output impedance it can drive low impedance load e.g. an audio power amplifier needs a very low (fractions of 1 ohm) output impedance to deliver useful power into a 4 ohm speaker. A high impedance output can only deliver small currents, so is useful only for driving a high impendence load. Impedance is synonymous with resistance in DC circuits and simple AC circuits with only pure resistive loads, but in AC circuits you can have complex loads (i.e. capacitive and inductive) which causes a phase shift between current and voltage, so impedance becomes a vector (magnitude and phase).
I have been electronic engineer for 35 yrs and can honestly say never bothered with JFETs, they really are useless and have been obsolete since the 80s, seriously an OPAMP is simpler and easier to get hold of, as for input impedance there are plenty of high input impedances opamps. One job they might be really good at is a voltage controlled resistor but again these days plenty of other options.
It's harder to match J-fets than bipolars in case of power amp input stage. Today all of the popular Toshiba J-fets are out production. But for some preamp, headphone amp I would try some of the new chips. Other than that it depends on application.
I started working as an electronic design engineer 50 years ago, and I've also never found an application where a single discrete JFET was the solution. I suppose that a high impedance source requiring low noise amplification at low cost might be one possible application, but it would require bias trimming and/or lots of dc feedback to give a repeatable operating point.
2:00: sensitivity, voltage amplification. The preamp is to keep the signal level above the background noise level and match the next stages input impedance Power output gain is for the power amp. RS is negative feedback gain stability.Thanks for taking the time to make your presentation.
Always informative and easy to digest...cheers.
Excellent presentation. The N-channel JFET characteristics are very similar to a vacuum tube triode but with lower voltages and currents.
Thankyou. Great explainer in an easy to understand format 👍👍
You always teach me new stuff. Really good channel. Thanks.
This is pretty much how most of studio condenser microphones operate. + something like 10:1 output transformer for low output impedance. Rg in this case is by default something like 1Giga Ohm. Many think this value is a mistake in schematics. The signal is produced by a variable capacitor (microphone capsule) which varies something like 1pF in capacitance, and produces signal. The circuit is often called "impedance converter", as there isn't much amplification going on.
If you take the output at the Source then you have low Z out without a transformer. Very handy!
@@markanderson8066 True, some are made like this. However transformer is usually used because of signal balancing and cmmr.
This looks like a standard tube layout.
Excellent series John! Unfortunately we didn’t spend much time in my Electronic Devices class on JFETs. Most of the class was spent on BJTs and MOSFETs. Thx for your presentation!
Thanks for all your wonderful content! I love the way you break it down on paper then build and demo it!
The way you went at it as just fine. From the output impedance, calculate your drain current, then from your estimated Vgs, estimate your source resistor. Then make up the circuit and find the proper value of for Rs. There's really no way around that with a JFET because of the wide variance of Vp between samples.
One issue to note: at 13.46 you annotate with the legend "1.4 seimens in this example". However, you had a hypothetical (2.3-0.9)mA / (1.5-0.5)V which works out to 1.4 *milli-Siemens,* and that's quite an important difference! The reciprocal transconductance is then 714 ohms, which of course is bigger than your hypothetical value of Rs. Also, whatever voltage you drop across Rs becomes unavailable for the negative-going output swing and reduces your headroom. This can become significant with low supply voltages and JFETs with high Vp. For a J111, it can be as much as 10V.
Now, in the real case, your MPF102 has a higher Vp than your hypothetical example, hence the higher value of Rs required, and its gm is also higher, maybe around 5mS at currents around a couple of mA. Nevertheless, it's unusual with small-signal JFETs to be able to ignore the transconductance when calculating the expected gain. By the way, the gm varies with drain current in a quadratic relationship, giving some mild second harmonic distortion when 1/gm is comparable to Rs (but not the third harmonics you get with BJTs when re is comparable to Re). The only fix is a higher supply voltage, and that allows more gain (which is effectively capped at half the supply voltage divided by the JFET's cutoff-voltage), which you can then sacrifice by increasing Rs at the same Id.
My bad, I left the milli off. Thanks for the response.
I've found that an optimum Q point for a JFET amplifier lies around 0.382 Vp, which corresponds to 0.382 Idss (magically). With that Q point Rs is simply Vp / Idss, and Rd ~= (Vdd+Vgs) / (2*Id), where Id = .383 Idss. Vp and Idss of a given JFET can be easily obtained with a simple measuring circuit which is available on the web.
I use this circuit as a pre-amp to feed line level amp boards for my boutique guitar amps!
very cool...i was able to get a gain of 9 with an MPF102 by adding a source degeneration capacitor.
Excellent video John.
Thanks for all the videos, John
Wow! Thanks for the contribution. I just happened to find this. I didn't see a notification.
Thanks John, that was VERY clear indeed!
Hey John, I'm not sure if you've ever checked out any of the depletion-mode mosfets available or not but, I ended up acquiring a few higher-power ones and they have really great specs and sound great for power amplification. The ones I picked up are the IXYS "IXTP6N50D2" which is a NPN in a TO-220 package and specs like Vds: 500V, Id: 6A, Vgs: -20V, +20V, Rds On: 550 mOhms (YES! milliohms!), Pd: 300W, etc. Go to Mouser (or your favorite place) and check out the full specs. They aren't expensive for what you get IMO at around $7 each. There's not many depletion mode mosfets but, you can still get ones with even higher power dissipation like 800 watts! They sound much better than the "enhancement mode" FETs IMHO. Seriously go check out the specs. I would share a link to the PDF but, TH-cam is getting weird with them.
2 of my favorites for class a amp building are 2sc5200 and ifrz250 powerful and sound soooo soooo good
Ya, you need to be cautious when selecting an opamp, jfet opamp's don't need a resistor when used as voltage follower, but transistor based input opamps do.
Thanks John, do you need to be concerned about noise contributed by the high impedance of Rg?
Sir what will be the voltage between Gate and source if we measure with a multimeter
Good video.
I would think you would want to characterize your jfet before doing your calculations.
I would find the pinch off voltage and the Odds, then draw the graph.
After that your calculations will be much closer to the real world response.
what resistance would RD be to use 8 ohm speakers
Beautiful 🫡
Hello sir,
since you are now working on amplifiers and preamplifiers i would like to ask you?
you may be able to show a microphone amplifier with time.
( schematic )
to make a condenser microphone preamp with:
1: maximum gain
2: as little sensitive as possible to high frequency
Friendly greetings from The Netherlands!
Rob
Just follow this with a single npn transistor stage and you have it.
There will be a small bias current flowing out of the gate of the n-type JFET. The bias resistor has to be sized to keep the resulting voltage within a few tens of mV of ground...
True, but as it's only of the order of 1 or 2nA at room temperature, that is still only 10 or 20mV with a 10M resistor. There are probably more compelling reasons for keeping the gate resistor as low as is needed to set the input impedance, such as the problems of picking up stray noise at a high impedance point.
@@RexxSchneider Yes, 1 or 2 nA at room temperature, but increasing exponentially at higher temps. In sonar applications with small piezoelectric sensors (~10nF PZT ceramic capacitors), the bias resister is a significant low frequency noise source and reducing the bias resistance increases this noise, so values between 20Meg and 50Meg are most common. I wouldn't be surprised if this value was lower in other applications.
Ive messed around with making a jfet hifi preamp, never could get it lower than .05 percent distortion without it loosing its little bit of tube vibe, in the end its just easier to get the tube sound with tubes 😁
how about the following amplifier/dc-power-source: transformer (volts up) -> capacitive voltage divider (tunable amplifier output level, dc-remover) -> full bridge rectifier + smoother capacitor (dc-output stage), should be very tunable and very frequency independent, linear amplification, dont know what class amplifier that is, but no transistors required
is there something wrong with the sufficiently large cored transformers that they are not directly used as amplifiers
something to do with power multiplication, not just signal level amplification
well at least if functions as an easy variable dc voltage power supply, variable with both the transformer and the capacitive voltage divider
Howdy.
In my mind there was an error about impedances.
When the impedances match half the power is consumed in the driver transistor + driver resistor and half in the load. The voltage drop is 1/sqrt2, not ½ which is -3dB.
Regards.
No, it is 1/2 as stated (not exactly 1/2 in a real circuit because the transistor is not a perfect device). Remember the transistor is functioning as a current source (or sink in this case) with very high drain impedance, therefore Rd is the output impedance. If you need confirmation, build the circuit and measure it.
@@JohnAudioTech Howdy again.
Donnerwetter ! You are correct. Thanks for straighten me.
In high regards.
What year did you buy that MPF102, and do you have any recommendations for current production jFETs? All the old standbys are obsolesced and getting rarer than hens' teeth. Most of the available through-hole jFETs are dirty-grade choppers.
Miller effect on g-d capacitance at 12V is negligible at audio frequencies. It's significant in triodes, which have higher g-a capacitance and operate with anode voltages an order of magnitude higher; in fact some designers like to add an external capacitor from gate to drain on jFET stages to approximate the Miller capacitance of triodes.
linearsystems (LSK389, LSK489, LSK170, LSJ689, LSJ689). canned dual jfet around 30us$ a piece.
interfet
The MPF102 hasn't been available from Mouser for some time. You can still get the similar 2N3819 from them, but I tend to baulk at paying around £1 for a single small-signal transistor. On the other hand, it's near-impossible to design a sensible bias point for these JFETs as their Vgs @0.2mA is specified as somewhere between -0.5V and -7.5V. Try selecting Rs to cater for that without applying negative feedback from drain to gate (which is actually the only sensible solution to get a repeatable quiescent drain voltage).
I've been experimenting lately with some of those dirty-grade chopper JFETs, specifically the J111/J112, which are cheaply available, and I've been impressed with the tightness of parameters across the samples I've tried so far. I still doubt that I'll ever find an application for them as amplifiers, though, as modern JFET opamps are far easier to design with, and give as good performance with high-impedance sources as any discrete design.
@@RexxSchneider Only reason I can think of for using a discrete jFET instead of a FET-input op amp is to introduce some of "that sweet, sweet harmonic distortion" by operating them in their triode region (the part of the response curve described in the video) to simulate a triode gain stage. In the case of the J11x, this is typically less than 2V, so you'd need really low system noise to get anything out of it. There are other ways of introducing this distortion (e.g. a diode and resistor from output to ground) that are more tuneable and less dependent on the transistor's individual characteristics and can be used with op amps, so there's another nail in the jFET gain stage coffin.
Still, it would be nice if anyone was still making the MPF-102 or J201, if only for the people who like these legacy designs.
@@AnalogDude_ I have some 2N4085 from many years ago when I needed a high impedance input with minimal temperature drift. That was before we had JFET opamps, and they were all that was available to do the job. I see you can still get them for $20+ apiece, but they don't really make sense anymore when an OPA2134 is half the price.
@@petersage5157 The higher the Vp, the lower the maximum stage gain for a given supply voltage, so using JFETs with higher Vp doesn't actually offer any improvement in noise margin.
The 2N3819 is a drop-in replacement for the MPF102 with near identical characteristics and it's still currently manufactured by Central Semiconductor Corp.
Hi John. Could you please do the same video with a MOSFET? Sorry if you did allready. Could you please point me to it in that case?
By the way: it is "Siemens" not seimens.
You have a problem again with the variability of the threshold voltage across samples of a MOSFET. I suggest using negative feedback from drain to gate and a significant source resistor to stabilise the voltage at the drain, which means you'll need a fairly high supply voltage. Try using a 15V supply, with Rd=4K7 and aim for Id around 1mA. Then connect 100K from drain to gate and 100K from gate to ground. That will set roughly 10V at the drain and 5V at the gate, so the source will be a couple of volts below that for a small-signal MOSFET like the 2N7000. You'll then need a source resistor around 3K, but you can split it and bypass the larger part to give some ac gain while retaining dc stability. Perhaps try 150R + 2K7, with the 2K7 bypassed with 220μF. Make the 150R bigger to decrease gain and improve linearity.
S like Z
@@pliedtka er... what? 🤔
@@RexxSchneider thanks for the detailed answer.
TH-cam suggested me his video dealing with a MOSFET in the meantime also.
@@thomasr.5443 You're welcome. If you mean his video using a power MOSFET as a class A amplifier driving a speaker ( th-cam.com/video/D4tHgHfljyY/w-d-xo.html ) then you can see the problems of setting a bias point which he tackles by using a trimmer resistor. Obviously that's a suitable solution for the hobbyist, but doesn't scale well for production. You can of course apply the same principle of trimming to a low-power pre-amp design using something like the 2N7000, and you should be able to get x10 gain with low noise and decent linearity even from a high impedance source. For low impedance sources, I still recommend using a BJT like the BC549C.
Creative video, thanks :)
At 13:46 in video I think you meant 1.4 millisiemens ?
8:05 Would you be so kind to explain, how the Ohm's formula is used to calculate voltage drop over a resistors, i forgot how it's done.
12V - 3K3 = 6 volt?
12÷3300 = 0,003636364 Amps.
12÷3690 = 0,003252033 Amps (if the fet wasn't there).
Roland uses something similar in their System 100M VCA (yusynth). Rg = 33K, no Rd.
Voltage drop across a resistor is a function of current, I*R=V. In the video, I determined what I needed the resistor value (3k3) and the voltage at the output node (6v). From that, I needed the current, V/R=I
@@JohnAudioTech Thank you very much for your answer, but ...
12V÷3300R = 0,003636364 Amp × 3300R = 12V again.
using a "voltage divider calculator" assuming the FET aint there:
3K3 and 390R = 1.268 Volts, so the voltage drop would be 10,732 V or am i missing something?
i tried your circuit in the Falstad simulator: tinyurl/2bg4httk, normally it works perfect, accept i don't calculated the beta for the JFET you're using.
@@AnalogDude_ From the supply (12v) to the output node (source) I needed 6v. 12-6=6v, therefore 6v across Rs (3k3) which results in a current of 1.8ma. Now that I know the idle current the transistor needs to conduct at, I can look at the graph to find the gate voltage needed for 1.8ma. It was around -1.2v for the real device, so I need 1.2v drop across the source resistor (Rs). 1.2/.0018=666 (680 is closest common value).
Beta is not the correct term for JFETs. Were you calculating the voltage gain of the circuit?
@@JohnAudioTech the Fallstad simulator has a calculator for that of you click on the JFET part. I'm looking at the graph of the 2SK30A at the same graph you used in the video, it has multiple plots, would it be the same -1.2V?
@@AnalogDude_ John did explain that JEFTs tend to have quite a wide spread of Vgs values, which is why you can only get a ballpark value using the nominal datasheet values. For a real circuit where biasing is important it's likely you will have to modify the resistor values to correct the biasing.
Newby, pertnear 70, gave up on electronics in HS because the study of logarithms didn't harmonize with puberty :-). A mental block for me now is the concept of impedance (type of resistance or "thirst" or combination?). It seems directly related to the resistance at an input(?) but what the hell is "output" impedance? Seems a contradiction. Oh well, maybe it'll come to me if I live long enough :-) since I'm not very studious. Flashing lights, sirens and motors are kinda fun. I don't think I'll be making any breakthroughs :-)
Output impedance determines how much current the circuit can deliver into a load., imagine a resistor in series with a perfect voltage source. If a circuit has a low output impedance it can drive low impedance load e.g. an audio power amplifier needs a very low (fractions of 1 ohm) output impedance to deliver useful power into a 4 ohm speaker. A high impedance output can only deliver small currents, so is useful only for driving a high impendence load.
Impedance is synonymous with resistance in DC circuits and simple AC circuits with only pure resistive loads, but in AC circuits you can have complex loads (i.e. capacitive and inductive) which causes a phase shift between current and voltage, so impedance becomes a vector (magnitude and phase).
Sir single supply amplifier (class D ) tda8932 amplifier clip limiter circuit please 🙂
I can't watch Eddie. Every time I do, I end up buying a new piece of a test gear!
2 the moon, alice in chains? I found Boba-Fet-Fin!! & his little dog Shin-Fet-Bet! good luck! lots of potential !!
Frist view
WERE IS YOUR BYPASS CAPACITOR?
I have been electronic engineer for 35 yrs and can honestly say never bothered with JFETs, they really are useless and have been obsolete since the 80s, seriously an OPAMP is simpler and easier to get hold of, as for input impedance there are plenty of high input impedances opamps. One job they might be really good at is a voltage controlled resistor but again these days plenty of other options.
_"seriously an OPAMP is simpler and easier to get hold of"_
An op-amp with a JFET input stage for instance? 😀
It's harder to match J-fets than bipolars in case of power amp input stage. Today all of the popular Toshiba J-fets are out production. But for some preamp, headphone amp I would try some of the new chips. Other than that it depends on application.
I started working as an electronic design engineer 50 years ago, and I've also never found an application where a single discrete JFET was the solution. I suppose that a high impedance source requiring low noise amplification at low cost might be one possible application, but it would require bias trimming and/or lots of dc feedback to give a repeatable operating point.
chi può fornirmi questo MOSFET con questa sigla 835A LM340T15 7815