I miss Snickers, and i think you do to. I think it is time, for a new cat in your life. It is not a betrayal against Snickers, oh the very contrary. Snickers is watching and approving , as he knows he will se you again. Don’t know when, don’t know where. A new chapter is open, fill it with enjoyable moments. Bear hug from a unknown friend :)
Thanks for your time in doing these videos, it's been 30 years since I had a small signal course in college and most of this information has been lost along the way. Your explanation of common mode rejection and the use of a constant current source is great.
Thanks, so much for the theory and practice - it really helped me understand how my LTP works (it has two different plate resistors and a single shared cathode resistor) and how I could improve it to help balance the output (equal plate resistance plus CCS)!
When testing, rather than using a sine wave, I use a triangle wave. That makes it much easier to see the linearity of the output of an amplifier. In this application, I wouldn't call it a constant current source, but rather, a constant current sink. (RIP LED😪😪) You don't know how many times I've had a stray wire, or even a test lead touch the wrong thing and cause (sometimes very expensive) failures! I've used a similar circuit to this in tubes. I used the jpentode half of a 6AN8 tube, with the cathode tied to the negative bias supply, and the anode tied to the cathodes of a 12AX7, to create a phase splitter to drive the output tubes.(6L6GT). Having the high impedance tail was important, to drive the grids of the output tubes equally, but even more importantly, one input was for the signal source, the other, was used for feedback. With that, setting the gain of the output stage was almost as easy as determining the gain of an OP-AMP. (A+B/A) The same results can be had, even using a low, or medium MU double triode. (12AT7-12BH7) Think of triode tubes as 'depletion mode" field effect transistors, except, very high impedance, hot running, high voltage, microphonic, and fragile. They are however, extremely fault tolerant, and heat is no issue.
Always fascinating. When I burn out components I just chop their heads off. Or maybe just their legs. I don't want them finding their way into other circuits.
Super tut, video, circuit and performance (on a breadboard, to boot.)! Let's talk about building a PCB... I take it that you used 2N3904/6s. It might help to use a (thermally-tied) paired transistor package (might the CC amp negate the need?) And, we might need higher power. So, what transistors would you suggest? I know that a breadboarded circuit may use just the bare minimum operational components and leave off additional components that improve performance. What, if any, might those be (caps) and where would they be placed? For power, for temporary use, a battery would clearly be best. But, how to power it, long term? A wall wart (Linear/with Transformer, for lower noise?) would separate the power supply noise and heat from the board/case. Add a Ferrite Bead, near the PCB end of the wire? Surely it needs to be in a metal case. But, what about other shielding, or board isolation zones, copper pours, etc? If you wanted to flesh out the circuit, hand draw a complete schematic and give us a link to it, that would be just wonderful. I would be glad to draw the schematic and PCB layout in KiCad and share it with everyone. I know that I, for one, would likely build it out, too. Yes, I know this is just a prototype. I'm headed for the bench to dig out my breadboard... Woot!
A little confusing; are you measuring the common mode ripple at the top of the tail resistor, and then (switch in the other position) : at the collector of the ccs (junction of the two 220 resistors)?
This circuit actually does that. In fact, you can use one input for the signal, and use the other for feedback. With that, you can set the gain just like an OP-AMP. (gain=A+B/A)
@@vincentrobinette1507 I'd imagine it is not as effective as using two op-amps to make the conversion? Correct if I'm wrong. I'm willing to learn more!
will you ever come out of your comfort zone of low distortion nad build something more crazy, like intended for guitar use amplifier? or even better j-fet based transistortube distortion simulating circuit? i know it is a bit strange question , considering you mostly build hi-fi equipment even from voltage regulators, but people say that some charmonics are good for you, and tube amps are simply too scary for beginers to play with... well there are low voltafe tubes but that not the same thing i am mostly interested in a way hifi amplifiers clip when they run out of headroom vs tube amps and sometimes jfets mpf102, bf245a and j201 are all going the way of the dinosaurs just like your favorite amps but they don't really make working copycat transistors like those as with the power amps and soon-ish there might be too late to play with them
I would very much like to know what low-cost preamp you would recommend especially one designed for subwoofers in an automotive application. A step-up buck power supply would be handy too. I really liked the report you did on the TPA3116 power amp. I installed 2 of these to power the outboard dash drivers powered by a buck supply set to a voltage of 26 v. results were pretty good for powering the American Bass 3-1/2 " full range speakers. Still need to install protection circuit to eliminate the thump when powering on.
Aside from testing these cheap amplifier boards every now and then, I rarely review audio products and have no idea what brand of preamp to use. I would likely just build my own.
I want to make a pre-amp for a Sennheiser e835 Dynamic mic. If I used a single op amp (recommendation, please,) to cancel the cable noise, would I just put the constant current source on the op amp's output pin? If so, how might the CCS be made = same as in the video, or other very specific details, please. Thank you, in advance. I plan to use this with your LM386 amplifier.
IIUC the equivelent circuit diagram of the LM386, it has a current mirror on the emitters of the two pairs of input Darlington transistors. Therefore, an additional, external CCS would not be needed. Do I understand this correctly?
HI, i see you and others using dummy loads to measure the performance of amps. I wonder if you could explain your methods? LAst time i tried to test an amp with a dummy load i made i got sparks. The load was built correctly. I was trying to compare the wave shape going into the amp and what it looked like at the output. What is the correct way to connect an oscilloscope or other test equipment to amp under test that is connected to the mains. I heard this in passing in some videos but not come across an explicit video yet. Cheers for the great videos.
The scope probe ground is tied to earth ground of the electrical supply. If the amplifier's circuit ground is also tied to earth, it may be a problem because the output of the amplifier that is marked negative may or may not be tied to earth depending on the design. The amplifier should be powered using an isolated power source. Also be careful that your signal source to the amplifier is not tied to earth ground either because that will cause the amplifier to become earth grounded too. You can get an isolated probe or use differential probing methods if you can't isolate the amp.
@@JohnAudioTech HI, thank you very much for the quick reply. I will try what you have suggested. I have just checked, there is a ground lift switch on the back, i hadnt noticed before. When it is off the negative is tied to mains earth and not if on. I will be extremely careful in future and follow your advice. Thank you so much for taking the time to reply, really appreciate it. Take care. Colin
Hi, John. I have laid out your circuit in KiCad in two versions, through hole and SMD. I still have a couple of tweaks, but I will be ready to share it, soon. One thing that I need to confirm is that the input decoupling capacitors on the transistor bases have a value of 1uF/Non-Polarized; correct? Also, I have to admit to not really being able to follow along with the video, all that well. I am not faulting you, in any way. However, without seeing the connections that you were making, I had difficulty. I guess I am more of a visual learner? I would love for you to make an extended version of the video, which shows where the all the signals are being sourced from and talks us through things, again. Also, I would benefit from knowing where this circuit would be used, in real-world applications. The "black magic" of this circuit is tremendously intriguing and I am pretty excited about it. I don't want to miss out on understanding it, just because I was afraid to speak up. Please and thank you. I so greatly appreciate your work and efforts!
@@steveswan5714I have these 2 way red and green 5mm 2pin leds and they explode if I bias them the wrong way. Literally explode. I gave them away, with a warning. They were pricey too haha.
how about just analog-to-digital conversion, then dsp, then output, independent from the input, so you have only one output valve to control, op amp huh
The higher the current gain of the transistors, the lower the need to use a matched pair. And there are also emitter resistors for degeneration. So you have internal feedback. This is also the reason, why the voltage gain is only around 20. If you don't use emitter resistors you can get considerably higher voltage gain, but you get also more sensitive for transistor parameter differences. But you can improve the voltage gain by use of a current mirror instead the collector resistors. And you can also implement emitter degeneration resistors to the current mirror. But in this case your voltage gain is allready so high, that you cannot operate the differential amplifier in open loop anymore.
Accidents happen and this time it happened sadly to this LED .let us look at it from the bright side, he didn't ceased to be a component but he evolved to a resistor.
I am sorry for your loss. It lived the very best life an LED can have - to provide an awesome boost in linearity.
You gotta be *tough* to be a sparky.
Nice video John!
Sorry for your loss.
Very educational, I have learnt a lot here. 👍
Nice stuff. Like it. Hope others have a “common” experience!
I saw it in the comment section. But it hits when you say 'It doesn't LED anymore'.
I miss Snickers, and i think you do to. I think it is time, for a new cat in your life. It is not a betrayal against Snickers, oh the very contrary. Snickers is watching and approving , as he knows he will se you again. Don’t know when, don’t know where. A new chapter is open, fill it with enjoyable moments. Bear hug from a unknown friend :)
Thanks for your time in doing these videos, it's been 30 years since I had a small signal course in college and most of this information has been lost along the way.
Your explanation of common mode rejection and the use of a constant current source is great.
Damn... bro paid and didn't even get a response.
Thanks, so much for the theory and practice - it really helped me understand how my LTP works (it has two different plate resistors and a single shared cathode resistor) and how I could improve it to help balance the output (equal plate resistance plus CCS)!
To the ex LED thank you for your service, a light that once shone brightly, extinguished far too soon. You will be remembered.
That ending was the best!!!
😂😆😂
Very informative thankyou John, great stuff
Interesting and very thoughtful !...cheers.
Thank you so much. I am trying to learn the difference of using just a resistor vs CCS and now I know.. loud and clear.
Very interesting. Would have been good to have and explanation of how the constant current source works though. A separate video maybe?
Already is.
See John's video from 7 years ago: th-cam.com/video/5M3aHJyMIzM/w-d-xo.html
Check w2aew's channel for good content on CCS, LTP and current mirrors.
What a beautiful theatrical ending!
Appreciate your work.
When testing, rather than using a sine wave, I use a triangle wave. That makes it much easier to see the linearity of the output of an amplifier. In this application, I wouldn't call it a constant current source, but rather, a constant current sink. (RIP LED😪😪) You don't know how many times I've had a stray wire, or even a test lead touch the wrong thing and cause (sometimes very expensive) failures! I've used a similar circuit to this in tubes. I used the jpentode half of a 6AN8 tube, with the cathode tied to the negative bias supply, and the anode tied to the cathodes of a 12AX7, to create a phase splitter to drive the output tubes.(6L6GT). Having the high impedance tail was important, to drive the grids of the output tubes equally, but even more importantly, one input was for the signal source, the other, was used for feedback. With that, setting the gain of the output stage was almost as easy as determining the gain of an OP-AMP. (A+B/A) The same results can be had, even using a low, or medium MU double triode. (12AT7-12BH7) Think of triode tubes as 'depletion mode" field effect transistors, except, very high impedance, hot running, high voltage, microphonic, and fragile. They are however, extremely fault tolerant, and heat is no issue.
R.I.P. little 5 mm LED.
I was struck a few days ago by a similar accident when a power supply clip lead slipped and took down 5 LEDs at once.
'Been there, done that'!!
Great explanation, thanks
LOL! great explanation, first off, but got several laughs this time, too.
The signal's sexy and it knows it!
Always fascinating. When I burn out components I just chop their heads off. Or maybe just their legs. I don't want them finding their way into other circuits.
Wait, you don't bury them in the back yard?
I dont think that poor LED is ever going to find its way into another circuit.
@@JohnAudioTech most of the components I have buried in my backyard actually work.
Super tut, video, circuit and performance (on a breadboard, to boot.)! Let's talk about building a PCB... I take it that you used 2N3904/6s. It might help to use a (thermally-tied) paired transistor package (might the CC amp negate the need?) And, we might need higher power. So, what transistors would you suggest? I know that a breadboarded circuit may use just the bare minimum operational components and leave off additional components that improve performance. What, if any, might those be (caps) and where would they be placed? For power, for temporary use, a battery would clearly be best. But, how to power it, long term? A wall wart (Linear/with Transformer, for lower noise?) would separate the power supply noise and heat from the board/case. Add a Ferrite Bead, near the PCB end of the wire? Surely it needs to be in a metal case. But, what about other shielding, or board isolation zones, copper pours, etc? If you wanted to flesh out the circuit, hand draw a complete schematic and give us a link to it, that would be just wonderful. I would be glad to draw the schematic and PCB layout in KiCad and share it with everyone. I know that I, for one, would likely build it out, too. Yes, I know this is just a prototype. I'm headed for the bench to dig out my breadboard... Woot!
Are you going to add a few more stages and build a discrete op-amp?
Yes, my next power amp.
What's the use of current mirror in input
Stage
It increases the gain of the stage.
DED: dark emitting diode.
A little confusing; are you measuring the common mode ripple at the top of the tail resistor, and then (switch in the other position) : at the collector of the ccs (junction of the two 220 resistors)?
Love your videos ❤
It was a good LED, never hurt nobody.
which result would be better if the led was replaced with a zener diode for ccs.. thanks sir
Hi, Can you please explain the purpose of LED in the circuit please?
I'm feel sorry to mr LED😔
👍😂😂😂
You can change Re from 4kOhm to 400kOhm that would give good CMRR…. And then increase V- to 1kVolt….. 💥
Have you ever worked with single ended to differential signal conversion?
This circuit actually does that. In fact, you can use one input for the signal, and use the other for feedback. With that, you can set the gain just like an OP-AMP. (gain=A+B/A)
@@vincentrobinette1507 I'd imagine it is not as effective as using two op-amps to make the conversion? Correct if I'm wrong. I'm willing to learn more!
All op amp using cc in diff. Amp
Like lm4558 ne5532 lm4562
That's exactly right.
will you ever come out of your comfort zone of low distortion nad build something more crazy, like intended for guitar use amplifier? or even better j-fet based transistortube distortion simulating circuit? i know it is a bit strange question , considering you mostly build hi-fi equipment even from voltage regulators, but people say that some charmonics are good for you, and tube amps are simply too scary for beginers to play with... well there are low voltafe tubes but that not the same thing
i am mostly interested in a way hifi amplifiers clip when they run out of headroom vs tube amps and sometimes jfets
mpf102, bf245a and j201 are all going the way of the dinosaurs just like your favorite amps but they don't really make working copycat transistors like those as with the power amps and soon-ish there might be too late to play with them
I would very much like to know what low-cost preamp you would recommend especially one designed for subwoofers in an automotive application. A step-up buck power supply would be handy too. I really liked the report you did on the TPA3116 power amp. I installed 2 of these to power the outboard dash drivers powered by a buck supply set to a voltage of 26 v. results were pretty good for powering the American Bass 3-1/2 " full range speakers. Still need to install protection circuit to eliminate the thump when powering on.
Aside from testing these cheap amplifier boards every now and then, I rarely review audio products and have no idea what brand of preamp to use. I would likely just build my own.
C'mon John, you shouldn't have buried it - it was still kicking out a couple of ohms which we could use elsewhere :)
I want to make a pre-amp for a Sennheiser e835 Dynamic mic. If I used a single op amp (recommendation, please,) to cancel the cable noise, would I just put the constant current source on the op amp's output pin? If so, how might the CCS be made = same as in the video, or other very specific details, please. Thank you, in advance. I plan to use this with your LM386 amplifier.
IIUC the equivelent circuit diagram of the LM386, it has a current mirror on the emitters of the two pairs of input Darlington transistors. Therefore, an additional, external CCS would not be needed. Do I understand this correctly?
HI, i see you and others using dummy loads to measure the performance of amps. I wonder if you could explain your methods? LAst time i tried to test an amp with a dummy load i made i got sparks. The load was built correctly. I was trying to compare the wave shape going into the amp and what it looked like at the output. What is the correct way to connect an oscilloscope or other test equipment to amp under test that is connected to the mains. I heard this in passing in some videos but not come across an explicit video yet. Cheers for the great videos.
The scope probe ground is tied to earth ground of the electrical supply. If the amplifier's circuit ground is also tied to earth, it may be a problem because the output of the amplifier that is marked negative may or may not be tied to earth depending on the design. The amplifier should be powered using an isolated power source. Also be careful that your signal source to the amplifier is not tied to earth ground either because that will cause the amplifier to become earth grounded too. You can get an isolated probe or use differential probing methods if you can't isolate the amp.
@@JohnAudioTech HI, thank you very much for the quick reply. I will try what you have suggested. I have just checked, there is a ground lift switch on the back, i hadnt noticed before. When it is off the negative is tied to mains earth and not if on. I will be extremely careful in future and follow your advice. Thank you so much for taking the time to reply, really appreciate it. Take care. Colin
Hi, John. I have laid out your circuit in KiCad in two versions, through hole and SMD. I still have a couple of tweaks, but I will be ready to share it, soon. One thing that I need to confirm is that the input decoupling capacitors on the transistor bases have a value of 1uF/Non-Polarized; correct? Also, I have to admit to not really being able to follow along with the video, all that well. I am not faulting you, in any way. However, without seeing the connections that you were making, I had difficulty. I guess I am more of a visual learner? I would love for you to make an extended version of the video, which shows where the all the signals are being sourced from and talks us through things, again. Also, I would benefit from knowing where this circuit would be used, in real-world applications. The "black magic" of this circuit is tremendously intriguing and I am pretty excited about it. I don't want to miss out on understanding it, just because I was afraid to speak up. Please and thank you. I so greatly appreciate your work and efforts!
I am watching the video repeatedly and I am starting to pick up on some things. Just wanted you to know that I am doing my homework.
This was helpful. There is more to study out, of course... th-cam.com/video/B33Vdd7cWg0/w-d-xo.html
write your name on your paper notes, I take screenshots for future reference, nice to give credit where credit is due. TY. Always useful info.
LOL missed your chance to include a "in memory of" at the end of the video. :-)
When i pop an led i don't bury them they get cremated 😂👍
I like it when they "go out with a BANG!". (470 uF @200 volts of capacitive discharge aught to do it)
@@vincentrobinette1507 😂
@@steveswan5714I have these 2 way red and green 5mm 2pin leds and they explode if I bias them the wrong way. Literally explode. I gave them away, with a warning. They were pricey too haha.
I'd hate to think of the state of my lawn if I buried all the components I've killed of the years.
My lawn would be a land fill!😁😁
When a capacitor dies, do you pour a bottle of Gatorade over its grave?
to be found in 10000 years???
Plastic contamination to the ground (earth)! Man! Take care!
how about just analog-to-digital conversion, then dsp, then output, independent from the input, so you have only one output valve to control, op amp huh
Hi John, quick question: the transistors no need to be a matching pair ? To run this circuit, just curious 🧐
The higher the current gain of the transistors, the lower the need to use a matched pair. And there are also emitter resistors for degeneration. So you have internal feedback. This is also the reason, why the voltage gain is only around 20. If you don't use emitter resistors you can get considerably higher voltage gain, but you get also more sensitive for transistor parameter differences.
But you can improve the voltage gain by use of a current mirror instead the collector resistors. And you can also implement emitter degeneration resistors to the current mirror. But in this case your voltage gain is allready so high, that you cannot operate the differential amplifier in open loop anymore.
@@gkdresden thank you for your response, is really appreciated 👍
It would be better if it's a matched pair. But it's not necessarily required. It all depends on how much precision you want.
Accidents happen and this time it happened sadly to this LED .let us look at it from the bright side, he didn't ceased to be a component but he evolved to a resistor.
Resistance is futile.
LED CEMETERY. (also WOW you have such impossible to dig ground)
Sorry, couldn't get past the "wiggle wiggle yer"