So, yesterday I was thinking about a good way to hook my radio's output to my car's stereo AUX input. I've got a simple cable (like a iPod -> AUX input jack cable), which works fine, but then I'm juggling the volume knobs on the radio and (left and right VFO) to keep the audio on the car's stereo manageable. My thought process was "hey, if I had some kind of audio limiter circuit, I could drop it in there and the car would get a nice level audio input... I wonder if anyone has done that already?" Fifteen seconds of googling and I found your video. I especially enjoyed your use of the scope and your discussion about what is actually happening along the path through the circuit. You have a good balance of technical discussion and practical demo... very nice. I'm pretty comfortable reading a schematic and turning it into a physical circuit, but I don't have the EE background to design the circuit in the first place. I think I can probably leverage yours into doing what I want. Thank you for sharing!
I tweaked this a bit to use it on the audio output of the micro BITX HF transceiver. It works great! I used a 600:600 ohm isolation transformer between the BITX output and the circuit input. I tweaked the values of R1 and R2 for my headphones, and I eliminated C2. Oscillograph show a reduction of a 4 volt peak-peak signal down to 50 mV peak-peak. Thanks Alan!
Wow, Alan, I should have known you'd have a circuit like this. Someone on TH-cam wasn't sure how to level the output of his bulb controlled wein bridge oscillator. I described leveling to him then after posting I did a Google search and the first TH-cam link brought me right here. I have refereed the gentleman here to your excellent video that doesn't depend on any gain control feedback circuity.
i guess problem with tv commercials might not be in the level, but rather in the compression. the peaks are the same level. also i guess this might kinda distort music and increase noise in quiet moments.
What a beautifully elegant circuit and - as always - a clear and detailed explanation and demonstration of how it works. I love these videos: a little bit of theory and a practical demonstration. I always learn something and enjoy doing it. Thank you.
Afrotechmods ...low power schottky diodes will probably work too, but maybe have a little more variation at the low amplitude end of things. Worth a try...
I built this circuit using 4 BAT54 diodes in MULTISIM. The output limiting voltage is around 10 -15 mV when the input voltage peak to peak is not too low, say 800 mV to 8V. Either too low or too high input will fail.
The timing of this video is incredible. I have been trying to find a way to automatically regulate the output amplitude of a basic function generator I put together. Now that I know the right term for what I'm looking for and how it basically operates, I can press on! Thank you!!
dude, that is awesome! Thank you very much for this. I am working on a graphic spectrum analyzer for my stereo, this will be VERY useful for leveling the audio when i switch from line to mic. (not a precise analyzer, i am more interested in blinking the lights in an interesting manner.)
i dont mean to be offtopic but does someone know a tool to get back into an Instagram account..? I stupidly forgot the account password. I would love any assistance you can offer me
@Brentley Corey thanks so much for your reply. I got to the site through google and im waiting for the hacking stuff now. I see it takes a while so I will get back to you later with my results.
Better performance than the AGC circuits I tried - I like your solution a lot! Now I just have to adapt it to my TV, to stop the annoying commercials to jump out at me :) Or, even better, I could just use the peak detector to automatically mute the set during commercials ;)
Contrary to popular belief the ads are not actually any louder in peak value than the main program content since both the transmit and receive devices have a max volume range before distortion become offensive. Rather what is happening is the average volume increased by boosting the lows and compressing the highs. Further ads generally are recorded with a 'high energy' delivery style similar to a person speaking in an excited voice as opposed to a normal or sombre voice. In essense the producer of the ad is using a circuit similar to this one to maintain that high average volume. So to counter the percived volume change you would need an 'anti' version of this circuit with a slower attack and fast decay response. Personaly my method is much simpler. I don't watch or listen to commercial broadcasters. This is not only kind to my ears but the IQ of the programming is generally raised by a significant order.
Rob B Thanks for the explanation, Rob. I suspect that my cable provider may be a bit wilder than that - I would say that general program is delivered at reduced volume, to make you raise the volume of the set and expose you to the full blast at commercial's time, but I will have a look with a scope, as soon as I have the time and the mood for it. As for your personal method, I wish I was single again to do it too :)
This is awesome!! I remember trying to design something like this in college and was never able to pull it off. Would low voltage schottkys like BAT54 work?
Randy, you'll probably find that you'll need to make the peak detector cap a little larger, depending on the compression in the scanner. I've taken apart my breadboard, but I think I might have to build a soldered prototype of this one too.
K7AGE I built a proper soldered breadboard of this circuit, increased the peak detector cap to 2.2uF, and ran it on my scanner, using an amplified speaker at the output. Works remarkably well.
Wow. As an electronics beginner, this circuit blows my mind. I tried to take a shot at explaining what's going on, let me know if I worked it out wrong: You are essentially using the diode D4 as a "current-controlled resistor", specifically: R = U/I, but the diode is always going to keep its voltage drop at value specific to the diode--let's say it's 0.2V in this case because it's a low-voltage one, so R = 0.2V/I. That leaves R depending on the current flowing through it, and that current is (for the most part) the current coming out of Q2's emitter, programmed by the peak voltage level at its base. And so the current-controlled D4 is taking the place of the second resistor in a resistor divider, the first resistor being the (fixed) resistor R1. The higher the voltage from the peak detector, the higher the current through the diode, the lower its resistance, the more current will flow through it to ground instead of through the output. Does that make sense?
Hi Alan. I bought 10 boards and have built up several based on your component values. The circuit works amazingly well. I'd add that (possibly) the schematic also have a parts list that corresponds to the PCB. I originally used 914 (1N4148) diodes (didn't see the 1N34 on your note sheet in the video)) and although it worked, it didn't kick in until about 1.5 v RMS. That makes sense. I replaced the diodes with1N34A and all is well...extremely well. Beside Ham Radio and SWLing, I'm using one in the studio to level-out old radio broadcasts for remastering...The Shadow, Inner Sanctum, Lights Out, etc. We have a bunch of commercial compressors and limiters available here, but why bother. This thing is a fool proof "one-knob-wonder" for the technician doing the dubbing and editing. I was going to play around with VTD photo resistors and opamps for active gain control or possibly some kind of VCA circuitry. Like I said, "Why bother." Thanks again for all that you do. Bob, N1KPR
Hi Alan, I just had a look at your video, and will be adding this circuit to my QRP-Labs QCX. I found your schematic and even ordered the boards from OSHKosh. You might want to update your schematic in your directory to reflect D1-D6 as 1N34A’s. Thanks for the explanation of the circuit!
Excellent video! I was looking for something like this to help compress air traffic control audio in the aircraft headset, since some stations are very weak while others are loud. (Note: Will only be used in a simulation environment, not in a real aircraft)
Definitely, this is what I must upgrade my TRX with. When listening to HF the audio level is jumping up and down dramatically. Sometimes it makes me jump up and down too. Also, this will help me to take care of my ears.
@@w2aew, of course, I'll use this one as a middle stage before the final amplifier. I'll probably make an external speaker with its own amplifier and this compressor in between. I also think if I'll insert this circuit into microphone preamp in order to make better signal compression on SSB.
Hi, Alan. I did breadboard the circuit. My brother had to catch my mistakes... lol... (I had missed the jumper on Q1, between its Collector and Base.) The circuit worked as advertised. At 1KHz/Sine/3Vpp, the leveled output was 32.4mVpp, measured with the cursors. Of course, as this is breadboarding, the circuit was noisy, with consistent, significant ringing. My scope's math function interpreted it, noise and all, as 46mVpp. I powered the circuit at 12VDC. My component values were fairly close in tolerance to your specifications. Probes were x10/AC Coupled/20M_BWL. DC offset = 0. The output is not centered evenly above and below ground. The positive amplitude is 7mV. The negative amplitude is -25.4mV. I would like for you to tell me more about this phenomenon. See following. Let's consider using the leveled output with an amplifier to output the reduced signal. I guess the mentioned skewing is not actually a DC offset and it should not damage the speaker? The skew would just mean that the speaker would travel farther on one side of center, than the other, but that the natural center point of the speaker travel is not offset with DC, so no damage will occur? I will watch the video again, to consider this, but I do recall that you said that this circuit might not be the best for High Fidelity listening. Thanks so much for all you do for us. It is truly appreciated.
Never saw the basic circuit of a compressor. I first started with a VCA-type compressor where a voltage controlled amplifier is controlled via a peak detector followed by a attack/release envelope, as these components are always part of a modular synthesizer... I´ve build some studio grade compressors for audio processing. Some uses a single ended j-FET as the control element (= UREI clone) some uses LDRs in conjunction with LEDs and/or light bulbs (=Teletronix clone) and, of course tubes. And now a have another topology to play around with as well, many thanx :)
Neat circuit. I do a similar thing using the audio software Audacity to compress the dynamic range of music from a CD or download. If you like listening to Opera or Symphonic music while driving in your car you find you are always fiddling with the volume. Turning it up when the levels dip close to the background noise (car/road) and then down as the levels come back up. So, it makes the highs a bit lower and the lowers a bit higher and keeps everything more-or-less above the background noise. Cheers, Mark
It seems that this circuit would be a nice addition to a signal tracer, so you don't have to ride the volume pot, while you are trying to carefully probe.
Just what I was looking for! No need for JFETs or OTAs or mixers. The use of diodes as variable impedances is really interesting, I suppose that’s similar to how they work on ring diode mixers. Oh and I assume the attack time could be varied by changing the ratio between the two capacitances.
Actually, in diode ring mixers, performance is best when the diodes are fully switched on/off. Yes, attack and hold-delay can be adjusted by playing with the caps and resistor values.
I really like the concept , but i struggle to understand the explanation form 3:55 till 4:10 . Is there some one that can help me get the transistor configuration?
I need this for my tv! Especially with the movies where you can here a door slam, or someone open bag of chips, but you can't make out what they're saying. I would have to make one for each channel (stereo). Thanks.
I'm a live audio engineer. Use compressors/limiters all the time, and never knew they could be this simple! 😳 Would slowing the attack time be as easy as putting resistance on the charging side? (I.E. between C3 and D1/D2) It could make this serviceable even for musical applications.
Correct me if Im wrong, but I think that the clipping produced when changing the volume intensities might be reduced if you increase a little the attack time. Thanks for the video!
Great lesson as always! Can you further discuss the distinctions between audio amplifier gain, input sensitivity, and output power? The simple audio amps I've made, always seem to require a trade off. Is there a way (I'm sure there must be) to construct a universal audio amp which is both sensitive and powerful?
There really isn't a tradeoff per se - it's all a matter of the particular requirements for the application. Sure, amplifiers with a high amount of gain need to be carefully designed to avoid unintended feedback resulting in oscillation. But, other than that - it just comes down to the application. An amplifier designed to take a low-level audio input and amplify it to a speaker - such as a PA (public address) amp, would have too much gain for a higher level signal such as a line-level. If you put a line-level signal into the MIC input, it would likely saturate and distort. It all depends on what voltage level you're starting with, and what you need it increased to. That determines the required gain.
@@w2aew Thanks for your reply! You've highlighted a perfect example; the distinction between a PA system and a line level amplifier... What mechanism would be needed to switch automatically between the two applications? In particular, I'm hoping to build an amp that is both sensitive enough for my small signal experiments, but is still able to handle the larger input signal from a cell phone or tape player when used as a power amp. I've had some success so far building high gain amps that can pick up a 1V inputs or at maximum power amp volume output. Would a AGC circuit be a solution? I tried building one for an AM radio I had made, but it's performance was not optimal and I did lose a bit of gain overall. I'm not sure how to configure one for an audio amp.
Hello, thanks to share with us your knowledge! I have a question: Ok the circuit compensate the input variations, BUT, how I define the desired output level?
Leonardo Takahata For this simple circuit, the output level is not adjustable - it is what it is because of the diode characteristics. The only way to set a desired level is to follow this circuit with an amplifier and adjust its gain accordingly. An amp is necessary anyway, since this output is quite low and the source impedance isn't low enough to drive a speaker.
NICE DESIGN, I'm looking to increase volume slowly, not to wake up startled to my alarm. could you help with a simple circuit that I could put in line with the amp? please
No, the AGC in a RF receiver usually consists of some kind of amplitude detector circuit driving a voltage (or current) controlled variable gain amplifier.
Hi Just an easy question i hope. You mention during the the peak detector explanation that it is a sort of a peak-peak detector. But in the scope plots where you show how the peak detector increase during each cycle, it looks like it is only the positive half cycles and not also the negative cycles where the peak detector increase?
The negative cycles charge up the input cap, the positive cycles transfer the voltage to the output cap. It is not a full wave rectifier, that's why you don't see any "action" on the negative cycle.
@@w2aew thank you for the answer. Yes i see, it behaves like a voltage doubler. I was also fooled by your scope plot where the peakdetector rides on top of the input voltage. Can now see the scsling is different on the two traces 😊
How do I determine if the germanium diode I got is genuine and not a schottky ? Local parts store has 1N60P but I'd like to know if there's a way to tell.
I just didn't understand why use a transistor for it like this - but now I think I got it - it's an identical transistor as Q2, so its voltage drop on base-emitter junction is approximately the same as Q2, so we have nice, precise stabilised bias?
***** You got it! A nice precise bias, that is largely independent of supply voltage. This circuit can operate over at least 5V to 12V supply, and even more likely...
You *could*, but remember that this circuit *does* add distortion, and it needs to be followed by an amplifier - it can't drive a speaker since it has a high output impedance.
Thanks for the reply Alan. I’m going to put one between my tv and my sound bar so i’m thinking it should be ok. only one way to find out and it will give me something to do today as well. Thanks again.. n2cmd
+Dylan Vizcarra Sometimes models don't accurately model reality. For example, SPICE would have no difficulty putting 100 amps through a 1N914A diode - it doesn't know that you'd let the smoke out of the device if you did that! All tools, like SPICE, have their limitations...
+w2aew I am curious now, when you initially decided to design the circuit, since you didn't use SPICE to determine the values of the components, did you analyze the transfer function to decide which values would provide a predetermined output? I'm just curious if/how the analysis techniques my professors are teaching me are actually applied in the real world.
+Dylan Vizcarra This circuit takes advantage of the diode's dynamic impedance changes vs. bias. I am not the original designer of this circuit, this basic design has been around for many, many years. I am surprised that it didn't simulate properly. I remember simulating this many years ago and thought it worked well in simulation (either PSpice or LTSpice - don't remember what i used back then).
+w2aew I see. Well I'll just have to reexamine my simulation and probably just build it to compare the results from LTSPICE to reality. Thank you for your help.
HI, thanks for posting. I'm interested because I'm designing a light control unit. Can I use this circuit for the analog conditioning of a signal coming from a mixer? Of course then I'll need some sort of amplification to feed the 0-3.3V ADC I'm using to get filters' samples.
Just wondering Alan. With 4 channels on the scope do you often wish for more vertical space on the display? What drives the selection of the LCD from the manf. point of view? I am thinking simple availabilty of ready made displays and divers or is there a more fundamental reason? Have you noticed how the screen resolution of computers in particlular portable models has regressed and stagnated at 1080P? Again I suspect, driven by cheap off the shelf 'TV' displays and driver chips.
Not really. I grew up with 4" CRTs and 4 channels, so today's displays are more than adequate. Major manufacturers use industrial quality displays, so don't generally benefit from high volume consumer grade displays.
I am interested in this circuit, IF it can do what I want. So here is my "problem". When listening to my local fire department and dispatch "tones" them out (or pages them out), the tones are very LOUD and I have to turn the scanner down to avoid that excruciating noise. Is there a chance this circuit can attenuate that tone down to the same level as the talking? I don't mind the tone being the same loudness as the people talking, but when that tone goes off, its ear piercing.
Yes, it should do that just fine. However, keep in mind that you can't simply place this circuit between your radio and the speaker. This circuit can't drive a speaker directly. This circuit needs to be inserted before the audio amplifier, or has to have an audio amplifier following it in order to drive the speaker.
@@w2aew awesome!! Thank you!! I'm trying to figure out a very simple way to make a compressor for a guitar pedal. It's only a small part of the design of the pedal, but one I'm stuck on... because I can't have a high parts count for such a small part of the whole. Thanks for the reply!! If you have any more tips, I'd love to hear them ;)
Alan, is it possible to set this circuit up so as to level the volume out for different services received on the scanner, but not so low as to need to Re-amplify them. I scan when mobile and would prefer to just feed from external speaker jack to this type of circuit and the from the circuit to an external speaker? As an option I can also feed from the scanner or this circuit into my cars audio system via a manufacturers provided input jack for ipods etc. Any thoughts? Thanks, Ted
Damn! True experts such as Dave of EEVBlog and Great Scott frequently warn us not to trust nor build just about anything we found on the Internet. But this certainly one of those true experts! Incidentally, why do you not use the Op amp peak detector which you showed in your video #77 so that it can effectively detect very small signal and not "waste" that signal to voltage drop? Does a transistor "follow voltage" such as an op amp does with its own feedback?
There are a lot of ways to design a circuit to do this function. I just thought this circuit was interesting, and something fun to show. Certainly there are ways to do this that will perform better.
@@w2aew Germanium diode is becoming obsolete. Maybe it is really 'the sign' for me to move on. Hahahah! I cannot even find it in Farnell/Newark/Element14 anymore. It is now an audiophile object which fetches high price.
They'll work, but not as well. The larger Vf of these diodes means that there will be more variation of the amplitude as the peak detector is being charged. Low power schottky diodes would be better, if you don't have the Ge diodes available.
Congratulations, very very good. If I want to use 5mV on the output, what changes do I need to make to the circuit? Is there a possibility to regulate this desired amplitude of output with a potentiometer?
@@w2aew I am simulating the circuit on the proteus before assembling the circuit on the protoboard. But it is not working in the simulation, I regulate the amplitude of the input signal and the output signal changes the amplitude. As I do not have the Diode 1N34A in Proteus, I am using 1N4148m is this the reason?
@@lincolngerken The circuit relies on the dynamic impedance variations of the 1N34A. Using a Si diode like the 1N4148 isn't going to work as well, and will require a much higher input signal level before the compression begins to occur.
@@lincolngerken In Proteus you can make and edit your own component (diode 1N34A) with the command "Make a device". Maybe if you share your Proteus simulation file someone can implement the 1N34A for you.
what would you suggest if the purpose would be a music synth? As you maybe know with higher frequencies the amplitude of some waves like saw (ramp) or triangle have lower amplitude due to shorter integration time. Would be very appreciated if the circuit would be explained in detail like you did here. Thank you very much
You re certainly right when it comes to HiFi. But Synths are another world. These harmonics you talk about make the charme of these (analogue) Synths (any why these are often more apreciated than digital synths). But the low noise I hear when the circuit adjusts it self is a issue. If I d filther these out I d cut too much low end from the synth. I have already thought about a compressor on the peak detector to boost high end. I mean there are many circuits of VCOs out there and certainly many of these have this issue and a solution to that. Unfortunately my electrical knowledge is too low to fully understand the location of the solution as well as how it works. Maybe you can pickup one that you find interesting and make a video of how these exactly works. Thank you
Well, that would be one of the things to find out. Oh, or do you mean that I have a pulse in wanted frequency that will empty the charging capacitor? I mean that on higher frequencies (tones) the power is lower would be ok but not double frequecy gives only half voltage. Let me go deeper- A VCO gives me a square. Differtiated (? opposite of integration) I get my pulse. The pulse opens a transistor that opens the cappacitor (charged by 5V as exaple) to the ground. Buffered by a OAmp... See what I mean? Other application would be to eliminate the attack of a guitar singnal. In other words have only the sustain. Starting point of both envelopes should, of course stay "the same" (only few ms difference). In other words a very fast compressor. Thank you!
Take 2 hours to understand this circuit, only one thing still not very sure, the negative audio function of D1, it seem it do not help much in peak detect, as for peak detect it only need D2. for the bias part of D1, I can use 2 of D2 to replace D1. Is that right?
Hay can i use this circuit for an audio amp, some of my friends like to turn up the amp to 11 so i want to put this inside the amp. So i can set an max volume on an hidden pot. The input would be the the aux input of the amp, so driving devices are smartphones, mp3 players etc.
Thank you for this beautiful video! But I don't understand one moment on this schematic - circuit should have a negative feedback to keep a constant amplitude. I don't see any feedback circuit here. How it's work? How we can make attenuation exactly equal to signal rise? Or may be this feedback made by D3, D4 diodes?
It's one of the beauties of this little circuit - no feedback! The diode impedance varies in the right proportion to the detected signal level to make the divider ratio track the signal amplitude.
Ilya Kalistru I haven't played with adjusting R1, but you should be able to make it adjustable over some range. However, as you make it smaller, you may get more distortion due to modulation of the diode voltage.
Hi Allen, I would like to use this circuit not only for "speaker" input, like from my Yaesu HF rig head phone jack, or speaker out put, but also a low level input. Can a "op amp" pre amplifier supply enough voltage to run the input of the circuit, or would I need to add a "LM386" power amp or similar to drive the input? (AJ6O)
I love your videos. They are complete and timeless. Would this be a good circuit to add to the output stage of a XR2206 frequency generator IC ? I put together a cheap kit but the output varies with frequency as expected. I'd like to get the standard 1VP-P at the output, no matter the frequency (audio) input's amplitude. I thought of using your circuit and play with the biasing to get 1VPP out of an LM386. Am I going down the right rabbit-hole?
You can, but keep in mind the following: - This simple circuit does add some distortion to the signal. You have to decide whether it is acceptable or not. - The output amplitude of this circuit is low, with a high impedance, so the use of the external amplifier is necessary. - This simple circuit may have some frequency dependence of its own - I didn't test for that.
@@w2aew Can you recommend to me a circuit that would be more suitable for my application? I've searched everywhere but not too many AGC circuits around that I could understand. I'll breadboard this one and learn from it, but still on the hunt for that sweet sweet AGC that doesn't mess-up the 0.5% distortion (THD) for this chip. Thanks again for a great video. I learned a lot from it.
@@w2aew I found this circuit that I built today, and it works perfect. gives me 1VPP no matter the input voltage for my El-Cheapo DIY Frequency generator. Google search string was "dds buffer amp" and this was the first hit. www.pongrance.com/dds-buffer.html Just a note for anyone looking for the same solution. The wheel has already been invented here too!
Very interesting circuit. Would you say it function the same as with "Tube" based AGC ? Vari mu like circuit ? I know it's done by "rebiasing the grid of the tube but I could never figure out if the tube was cathode biased or grid biased for it to work...Anyway great video as usual !
This is ONE ir my favourites videos coming from you... GORGEOUS my friend... GORGEOUS..!!! Do you know any book for making fx devices like compressors, reverbs...and all that stuff for audio?? THANKS in advance...!
great video! you mentioned speaker level on the input. Did you mean line level? I noticed in your demonstration video you were using an RCA connector off the back of the scanner.(phono output?) I would like to use this circuit on the aux output of a mixing console before it enters a distribution amplifier.
It will likely work with line level as well, as you can see from the beginning of the video. Keep in mind that this circuit does add some harmonic distortion to the signal, so it really isn't suitable for high fidelity audio work.
As mentioned in the video, this is NOT an amplifier. It will not boost the quiet signals, it will only level the loud ones. So, you'd want to turn up the volume going INTO this circuit so that even the quiet signals get limited, then all levels will get limited.
I just ordered the diodes and I'm going to build this. I want to use it to level the stereo output of my cell phone. I listen to music or whatever at night when I go to bed. Do you see any problem applying the peak detector output to two attenuating diode pairs? I'm thinking it will be ok.
Remember, the output of this circuit is high impedance - it will not drive a speakers/headphones directly. You'll have to put an audio amp after this circuit. Also, the clipping is non-linear, so it *does* add distortion.
I was able to reproduce this circuit and I used an op-amp to drive the input since my sig gen wasn't up to the task without distorting. I only monitored the output on a scope. It worked as advertised and did indeed add some distortion and some noise. (no sure why). Onward and upward. Thanks.
@@SuburbanDon Please note that you will likely need an amplifier to buffer the output as well, since it can't drive low impedance loads like speakers or headphones directly.
Very nice little circuit! It fits for my needs :). Thanks for posting this video sir Alan. Thumbs up for you. Anyway,what is the value of R5? I cannot see very well here on my monitor.
very nice video, but I have one question. On your schematic, what are the V+ connect to R3 and T2 mean? They look like ground to me, but you put V+ next to them, are they 12V power?
How much distortion does this circuit add? I would like to use this to level the output from a flat panel TV before it goes to an external amp. Thanks.
From the measurements that I have made, this circuit adds about 10-15% THD (total harmonic distortion). Probably too much for a high fidelity audio system with your flat panel TV.
Alan, great video on a neat circuit. I have a use for this, but in a slightly different application. In the airplane I fly, automated audio alerts are much louder than the radios, so I'd like to put this circuit between the ship and my headset. Problem is, the ship's audio amp expects headset impedance (mine is 270 ohms per transducer, paralleled in mono) and the LM38x amp expects 8 ohms. Would any modification to your circuit be necessary to accommodate these conditions? Also, as a portable device, I'd like to power it from two 9V batteries in series. Is your specified 12V a critical value, or will 18V work just as well? Thanks for all your great content!
I would think that the LM38x devices would work fine with the 270ohm headphones. If there's a problem, simply place a 10ohm resistor in parallel with the phones (appropriately rated for power). The AVC circuit will work over a wide frequency range. I quickly tested it from 5v to 12v, but it should work fine at 18V as well.
G'day Alan I'd like to use this circuit to control the volume on my TV which uses an external amp and speakers. Am I correct in thinking that for stereo use I would need to duplicate this circuit, i.e. one for each channel? Are there any issues to be aware of if I do this?
Yes, you would definitely have to duplicate the circuit - one for each channel. You'll also note that this circuit does add some distortion, so you'll have to decide if it is detrimental to your listening enjoyment. It's fine for narrow bandwidth audio like a scanner, but might be more objectionable for high fidelity audio.
Wow, this is great! I'm wondering a couple things... This looks familiar to an option for those (and by those I also mean me) that are interested in leveraging the power of their PC soundcardsto run as a budget (free) oscilloscope as an intro. With a 1V peak to peak limit can you use this to sending the signal to the mic input? I have seen a few idea circulating around but no real in depth demo video on how to implement it. I realize the limitations of the max 44k limit of the card but I would love to see someone try this is actually see what can be done. Perhaps this would be a time wasting but entertaining project to try...
You may have to attenuate the output a little for the mic input, but it's certainly worth a try. Of course, there would be no vertical calibration on your scope...
To attenuate I think a 10:1 (9M and 1M voltage divider) and 100:1 switch would be a good option. I'm leaning toward an Operational Amplifier with 1M resistors for high impedance and then your circuit as the front end. The tough part is signals below the 20Hz audible range of course and DC measurement. Some suggested the DAC in cheaper china based USB soundcards would work, Adding a bypass jumper and switch on the input cap to prevent AC coupling when wanted would be an option.
Hey I love the circuit, Allen. I wanted to use the circuit with a pair of earbuds so that it would never exceed 85 dBs. So, I was wondering if you had any tips on how to optimize the circuit for a pair of earbuds and still get a decent sound quality from it.
+Daniel Yoon (Yoonyman) Follow this circuit with a small audio amplifier suited to driving earbuds, and adjust the gain/attenuation to get the volume you desire in the buds.
So, yesterday I was thinking about a good way to hook my radio's output to my car's stereo AUX input. I've got a simple cable (like a iPod -> AUX input jack cable), which works fine, but then I'm juggling the volume knobs on the radio and (left and right VFO) to keep the audio on the car's stereo manageable.
My thought process was "hey, if I had some kind of audio limiter circuit, I could drop it in there and the car would get a nice level audio input... I wonder if anyone has done that already?" Fifteen seconds of googling and I found your video. I especially enjoyed your use of the scope and your discussion about what is actually happening along the path through the circuit. You have a good balance of technical discussion and practical demo... very nice.
I'm pretty comfortable reading a schematic and turning it into a physical circuit, but I don't have the EE background to design the circuit in the first place. I think I can probably leverage yours into doing what I want. Thank you for sharing!
I tweaked this a bit to use it on the audio output of the micro BITX HF transceiver. It works great! I used a 600:600 ohm isolation transformer between the BITX output and the circuit input. I tweaked the values of R1 and R2 for my headphones, and I eliminated C2. Oscillograph show a reduction of a 4 volt peak-peak signal down to 50 mV peak-peak. Thanks Alan!
Very nice use of the operating-point-depended impedance of the diode! Will definitly try this out!
Wow. Great circuit. Right to the point and very low distortion. Thank you.
Wow, Alan, I should have known you'd have a circuit like this. Someone on TH-cam wasn't sure how to level the output of his bulb controlled wein bridge oscillator. I described leveling to him then after posting I did a Google search and the first TH-cam link brought me right here. I have refereed the gentleman here to your excellent video that doesn't depend on any gain control feedback circuity.
Brilliant! Have been looking for this. Now to put it together to level down the television commercials. Excellent explanation. Thanks much.
i guess problem with tv commercials might not be in the level, but rather in the compression. the peaks are the same level. also i guess this might kinda distort music and increase noise in quiet moments.
Great job, I have been looking for a circuit like this. You do a great job of explaining the theory, thank you!
Another innovative use of diodes as variable attenuators I never would have considered.
But the quality will not be high.
As seen in cassette recorders ca. 1970.
EXCELLENT video.
What a beautifully elegant circuit and - as always - a clear and detailed explanation and demonstration of how it works. I love these videos: a little bit of theory and a practical demonstration. I always learn something and enjoy doing it. Thank you.
Afrotechmods ...low power schottky diodes will probably work too, but maybe have a little more variation at the low amplitude end of things. Worth a try...
I built this circuit using 4 BAT54 diodes in MULTISIM. The output limiting voltage is around 10 -15 mV when the input voltage peak to peak is not too low, say 800 mV to 8V. Either too low or too high input will fail.
The timing of this video is incredible. I have been trying to find a way to automatically regulate the output amplitude of a basic function generator I put together. Now that I know the right term for what I'm looking for and how it basically operates, I can press on! Thank you!!
dude, that is awesome! Thank you very much for this. I am working on a graphic spectrum analyzer for my stereo, this will be VERY useful for leveling the audio when i switch from line to mic. (not a precise analyzer, i am more interested in blinking the lights in an interesting manner.)
Very nice and handy circuit. Great explanation and demonstration.
i dont mean to be offtopic but does someone know a tool to get back into an Instagram account..?
I stupidly forgot the account password. I would love any assistance you can offer me
@Ares Markus instablaster =)
@Brentley Corey thanks so much for your reply. I got to the site through google and im waiting for the hacking stuff now.
I see it takes a while so I will get back to you later with my results.
@Brentley Corey It worked and I now got access to my account again. I am so happy:D
Thank you so much, you saved my ass !
@Ares Markus you are welcome :)
Better performance than the AGC circuits I tried - I like your solution a lot!
Now I just have to adapt it to my TV, to stop the annoying commercials to jump out at me :)
Or, even better, I could just use the peak detector to automatically mute the set during commercials ;)
Contrary to popular belief the ads are not actually any louder in peak value than the main program content since both the transmit and receive devices have a max volume range before distortion become offensive. Rather what is happening is the average volume increased by boosting the lows and compressing the highs. Further ads generally are recorded with a 'high energy' delivery style similar to a person speaking in an excited voice as opposed to a normal or sombre voice. In essense the producer of the ad is using a circuit similar to this one to maintain that high average volume. So to counter the percived volume change you would need an 'anti' version of this circuit with a slower attack and fast decay response. Personaly my method is much simpler. I don't watch or listen to commercial broadcasters. This is not only kind to my ears but the IQ of the programming is generally raised by a significant order.
Rob B Thanks for the explanation, Rob. I suspect that my cable provider may be a bit wilder than that - I would say that general program is delivered at reduced volume, to make you raise the volume of the set and expose you to the full blast at commercial's time, but I will have a look with a scope, as soon as I have the time and the mood for it. As for your personal method, I wish I was single again to do it too :)
This is *GREAT* Alan. I've been looking for a circuit like this for my scanner's! Thanks!
This is what I've looking for this whole time! Thanks for this, sir..
Just built this circuit for my ubitx kit HF Transceiver and it works a treat,
This is awesome!! I remember trying to design something like this in college and was never able to pull it off. Would low voltage schottkys like BAT54 work?
If afrotechmods tells you "this is awesome" then it is
If dbuezas tells you "this is awesome" then it is
Any new material coming? we need a laugh.
This sort of peak detection is hard to teach in theory but, with a scope in front of you, it’s clear.
This is so cool!!! Thank you for this exceptional tutorial!
Excellent! Thank you! Yeah, I'm going to build it.
This was good for signals >1.5Vpp (diodes fully on?) and
Alan, neat little project. I would like to hear a scanner or radio through it with different volume levels. I might have to build this!
Randy, you'll probably find that you'll need to make the peak detector cap a little larger, depending on the compression in the scanner. I've taken apart my breadboard, but I think I might have to build a soldered prototype of this one too.
K7AGE I built a proper soldered breadboard of this circuit, increased the peak detector cap to 2.2uF, and ran it on my scanner, using an amplified speaker at the output. Works remarkably well.
Alan, how about a short demo video?
K7AGE Sure - here you go:
Automatic Volume Level Control circuit demonstration | scanner
Wow. As an electronics beginner, this circuit blows my mind. I tried to take a shot at explaining what's going on, let me know if I worked it out wrong: You are essentially using the diode D4 as a "current-controlled resistor", specifically: R = U/I, but the diode is always going to keep its voltage drop at value specific to the diode--let's say it's 0.2V in this case because it's a low-voltage one, so R = 0.2V/I. That leaves R depending on the current flowing through it, and that current is (for the most part) the current coming out of Q2's emitter, programmed by the peak voltage level at its base. And so the current-controlled D4 is taking the place of the second resistor in a resistor divider, the first resistor being the (fixed) resistor R1. The higher the voltage from the peak detector, the higher the current through the diode, the lower its resistance, the more current will flow through it to ground instead of through the output. Does that make sense?
That's basically it.
Another cool project. Clever idea. Nice approach. Thanks!!!
Hi Alan. I bought 10 boards and have built up several based on your component values. The circuit works amazingly well. I'd add that (possibly) the schematic also have a parts list that corresponds to the PCB. I originally used 914 (1N4148) diodes (didn't see the 1N34 on your note sheet in the video)) and although it worked, it didn't kick in until about 1.5 v RMS. That makes sense. I replaced the diodes with1N34A and all is well...extremely well. Beside Ham Radio and SWLing, I'm using one in the studio to level-out old radio broadcasts for remastering...The Shadow, Inner Sanctum, Lights Out, etc. We have a bunch of commercial compressors and limiters available here, but why bother. This thing is a fool proof "one-knob-wonder" for the technician doing the dubbing and editing. I was going to play around with VTD photo resistors and opamps for active gain control or possibly some kind of VCA circuitry. Like I said, "Why bother." Thanks again for all that you do.
Bob, N1KPR
I"m glad to hear that this simple design is working so well for you!
That is a very neat circuit! Thanks Alan.
Hi Alan, I just had a look at your video, and will be adding this circuit to my QRP-Labs QCX. I found your schematic and even ordered the boards from OSHKosh. You might want to update your schematic in your directory to reflect D1-D6 as 1N34A’s. Thanks for the explanation of the circuit!
Hi Michael I just built a qcx myself and would like to know how that worked out for you?
Thanks kg5sso
Excellent video! I was looking for something like this to help compress air traffic control audio in the aircraft headset, since some stations are very weak while others are loud.
(Note: Will only be used in a simulation environment, not in a real aircraft)
Definitely, this is what I must upgrade my TRX with. When listening to HF the audio level is jumping up and down dramatically. Sometimes it makes me jump up and down too. Also, this will help me to take care of my ears.
Just beware that you need to follow this circuit with an audio amp - this circuit can't drive a speaker directly.
@@w2aew, of course, I'll use this one as a middle stage before the final amplifier. I'll probably make an external speaker with its own amplifier and this compressor in between. I also think if I'll insert this circuit into microphone preamp in order to make better signal compression on SSB.
Excellent, as always! I had just been thinking about designing exactly this for a scanner.
Hi, Alan. I did breadboard the circuit. My brother had to catch my mistakes... lol... (I had missed the jumper on Q1, between its Collector and Base.)
The circuit worked as advertised. At 1KHz/Sine/3Vpp, the leveled output was 32.4mVpp, measured with the cursors. Of course, as this is breadboarding, the circuit was noisy, with consistent, significant ringing. My scope's math function interpreted it, noise and all, as 46mVpp.
I powered the circuit at 12VDC. My component values were fairly close in tolerance to your specifications. Probes were x10/AC Coupled/20M_BWL. DC offset = 0.
The output is not centered evenly above and below ground. The positive amplitude is 7mV. The negative amplitude is -25.4mV. I would like for you to tell me more about this phenomenon. See following.
Let's consider using the leveled output with an amplifier to output the reduced signal. I guess the mentioned skewing is not actually a DC offset and it should not damage the speaker? The skew would just mean that the speaker would travel farther on one side of center, than the other, but that the natural center point of the speaker travel is not offset with DC, so no damage will occur? I will watch the video again, to consider this, but I do recall that you said that this circuit might not be the best for High Fidelity listening.
Thanks so much for all you do for us. It is truly appreciated.
Edit: I used PN2222A transistors. I do have 2N3904s.
Fantastic explanation, thanks very much!
Very interesting! Thanks for all the great videos!
Never saw the basic circuit of a compressor. I first started with a VCA-type compressor where a voltage controlled amplifier is controlled via a peak detector followed by a attack/release envelope, as these components are always part of a modular synthesizer...
I´ve build some studio grade compressors for audio processing. Some uses a single ended j-FET as the control element (= UREI clone) some uses LDRs in conjunction with LEDs and/or light bulbs (=Teletronix clone) and, of course tubes.
And now a have another topology to play around with as well, many thanx :)
Neat circuit. I do a similar thing using the audio software Audacity to compress the dynamic range of music from a CD or download. If you like listening to Opera or Symphonic music while driving in your car you find you are always fiddling with the volume. Turning it up when the levels dip close to the background noise (car/road) and then down as the levels come back up. So, it makes the highs a bit lower and the lowers a bit higher and keeps everything more-or-less above the background noise. Cheers, Mark
It seems that this circuit would be a nice addition to a signal tracer, so you don't have to ride the volume pot, while you are trying to carefully probe.
Just what I was looking for! No need for JFETs or OTAs or mixers. The use of diodes as variable impedances is really interesting, I suppose that’s similar to how they work on ring diode mixers.
Oh and I assume the attack time could be varied by changing the ratio between the two capacitances.
Actually, in diode ring mixers, performance is best when the diodes are fully switched on/off. Yes, attack and hold-delay can be adjusted by playing with the caps and resistor values.
Nice demo and very well explained!
I really like the concept , but i struggle to understand the explanation form 3:55 till 4:10 . Is there some one that can help me get the transistor configuration?
Extreme educative video!! and extreme usefull and easy/few parts circuit!! THANKS A LOT!!
I need this for my tv! Especially with the movies where you can here a door slam, or someone open bag of chips, but you can't make out what they're saying. I would have to make one for each channel (stereo). Thanks.
So the limiting is done by only diodes? I thought this was much more dificult. Thank you so much for explaining a AGC/ limiter
I'm a live audio engineer. Use compressors/limiters all the time, and never knew they could be this simple! 😳 Would slowing the attack time be as easy as putting resistance on the charging side? (I.E. between C3 and D1/D2) It could make this serviceable even for musical applications.
This simple circuit distorts pretty heavily - OK for communications grade audio, but would likely be ugly for music applications.
Correct me if Im wrong, but I think that the clipping produced when changing the volume intensities might be reduced if you increase a little the attack time. Thanks for the video!
Really useful project! Thanks!
So the integrating circuit’s time period is every time the slope of the signal crosses zero? Well, with hysteresis due to the diode drop?
This seems to be the same as automatic gain control but maybe a slower release time.
Great video ! Thanks
Great video, thanks :-) that's a nifty circuit, I think I'll build one.
We could have completely circumvented the ‘loudness war’ had a circuit like this been incorporated into car stereos!
Great lesson as always! Can you further discuss the distinctions between audio amplifier gain, input sensitivity, and output power? The simple audio amps I've made, always seem to require a trade off. Is there a way (I'm sure there must be) to construct a universal audio amp which is both sensitive and powerful?
There really isn't a tradeoff per se - it's all a matter of the particular requirements for the application. Sure, amplifiers with a high amount of gain need to be carefully designed to avoid unintended feedback resulting in oscillation. But, other than that - it just comes down to the application. An amplifier designed to take a low-level audio input and amplify it to a speaker - such as a PA (public address) amp, would have too much gain for a higher level signal such as a line-level. If you put a line-level signal into the MIC input, it would likely saturate and distort. It all depends on what voltage level you're starting with, and what you need it increased to. That determines the required gain.
@@w2aew
Thanks for your reply! You've highlighted a perfect example; the distinction between a PA system and a line level amplifier... What mechanism would be needed to switch automatically between the two applications? In particular, I'm hoping to build an amp that is both sensitive enough for my small signal experiments, but is still able to handle the larger input signal from a cell phone or tape player when used as a power amp. I've had some success so far building high gain amps that can pick up a 1V inputs or at maximum power amp volume output. Would a AGC circuit be a solution? I tried building one for an AM radio I had made, but it's performance was not optimal and I did lose a bit of gain overall. I'm not sure how to configure one for an audio amp.
Very good development a very clearly explanation.
thanks sr.
Hello, thanks to share with us your knowledge! I have a question: Ok the circuit compensate the input variations, BUT, how I define the desired output level?
Leonardo Takahata For this simple circuit, the output level is not adjustable - it is what it is because of the diode characteristics. The only way to set a desired level is to follow this circuit with an amplifier and adjust its gain accordingly. An amp is necessary anyway, since this output is quite low and the source impedance isn't low enough to drive a speaker.
So i wonder, is it possible to replace some of these diodes by LEDs and have them light up specifically when there is compression going on?
Unfortunately, no. The forward voltage of the LEDs is too high.
Very good, thanks for the explanation!
Awesome!
Trying to put together a list of components needed for building this. Are C1,C2,C3 and C5 non polarized caps?
all non polarized.
NICE DESIGN, I'm looking to increase volume slowly, not to wake up startled to my alarm. could you help with a simple circuit that I could put in line with the amp? please
This might be a silly question, but is this the same circuit topology for an Automatic Gain Control (AGC) block in an RF circuit?
No, the AGC in a RF receiver usually consists of some kind of amplitude detector circuit driving a voltage (or current) controlled variable gain amplifier.
Hi
Just an easy question i hope.
You mention during the the peak detector explanation that it is a sort of a peak-peak detector. But in the scope plots where you show how the peak detector increase during each cycle, it looks like it is only the positive half cycles and not also the negative cycles where the peak detector increase?
The negative cycles charge up the input cap, the positive cycles transfer the voltage to the output cap. It is not a full wave rectifier, that's why you don't see any "action" on the negative cycle.
@@w2aew thank you for the answer.
Yes i see, it behaves like a voltage doubler. I was also fooled by your scope plot where the peakdetector rides on top of the input voltage. Can now see the scsling is different on the two traces 😊
Nice work Alen. Can you share your circuit drawings?
How do I determine if the germanium diode I got is genuine and not a schottky ? Local parts store has 1N60P but I'd like to know if there's a way to tell.
Simple, smart and usefull design, I like it a lot :)
And thank you for detailed presentation of how it works.
One question however - what does Q1 do?
As described in the video, it establishes a DC bias on the peak detector so that Q2 is on the verge of conducting.
I just didn't understand why use a transistor for it like this - but now I think I got it - it's an identical transistor as Q2, so its voltage drop on base-emitter junction is approximately the same as Q2, so we have nice, precise stabilised bias?
***** You got it! A nice precise bias, that is largely independent of supply voltage. This circuit can operate over at least 5V to 12V supply, and even more likely...
Nice, thanks. Looking forward for another circuit fun videos to figure them out :)
Alan great project. Is there any reason I couldn’t build two of these side by side to use on my tv for a stereo input/output?
You *could*, but remember that this circuit *does* add distortion, and it needs to be followed by an amplifier - it can't drive a speaker since it has a high output impedance.
Thanks for the reply Alan. I’m going to put one between my tv and my sound bar so i’m thinking it should be ok.
only one way to find out and it will give me something to do today as well. Thanks again.. n2cmd
I simulated the circuit in SPICE and was unable to duplicate your results. I find that to be very interesting.
+Dylan Vizcarra Sometimes models don't accurately model reality. For example, SPICE would have no difficulty putting 100 amps through a 1N914A diode - it doesn't know that you'd let the smoke out of the device if you did that! All tools, like SPICE, have their limitations...
+w2aew I can see that being the problem. I suppose I'll just have to build it to be sure. Thank you for the response.
+w2aew I am curious now, when you initially decided to design the circuit, since you didn't use SPICE to determine the values of the components, did you analyze the transfer function to decide which values would provide a predetermined output? I'm just curious if/how the analysis techniques my professors are teaching me are actually applied in the real world.
+Dylan Vizcarra This circuit takes advantage of the diode's dynamic impedance changes vs. bias. I am not the original designer of this circuit, this basic design has been around for many, many years. I am surprised that it didn't simulate properly. I remember simulating this many years ago and thought it worked well in simulation (either PSpice or LTSpice - don't remember what i used back then).
+w2aew I see. Well I'll just have to reexamine my simulation and probably just build it to compare the results from LTSPICE to reality. Thank you for your help.
HI, thanks for posting. I'm interested because I'm designing a light control unit. Can I use this circuit for the analog conditioning of a signal coming from a mixer? Of course then I'll need some sort of amplification to feed the 0-3.3V ADC I'm using to get filters' samples.
Good video and thanks. Can this be made to amplify lower signals as well as attenuate the smaller ones?
You'd have to include an amplifier at the input or output, or both, depending on the desired signal levels.
@@w2aew I'm glad that you understood what I meant. I had already sent when I realized what I had done. Thanks for the reply 👍
Just wondering Alan. With 4 channels on the scope do you often wish for more vertical space on the display? What drives the selection of the LCD from the manf. point of view? I am thinking simple availabilty of ready made displays and divers or is there a more fundamental reason? Have you noticed how the screen resolution of computers in particlular portable models has regressed and stagnated at 1080P? Again I suspect, driven by cheap off the shelf 'TV' displays and driver chips.
Not really. I grew up with 4" CRTs and 4 channels, so today's displays are more than adequate. Major manufacturers use industrial quality displays, so don't generally benefit from high volume consumer grade displays.
I am interested in this circuit, IF it can do what I want. So here is my "problem". When listening to my local fire department and dispatch "tones" them out (or pages them out), the tones are very LOUD and I have to turn the scanner down to avoid that excruciating noise. Is there a chance this circuit can attenuate that tone down to the same level as the talking? I don't mind the tone being the same loudness as the people talking, but when that tone goes off, its ear piercing.
Yes, it should do that just fine. However, keep in mind that you can't simply place this circuit between your radio and the speaker. This circuit can't drive a speaker directly. This circuit needs to be inserted before the audio amplifier, or has to have an audio amplifier following it in order to drive the speaker.
Thanks for the great video. How about distortion level, is it suitable for HiFi?
OK for voice, not suitable for HiFi.
Would this work in line with head phones? All mp3 wav and flac files have diffrent volumes
Only for old high-z crystal headphones. For modern low impedance headphone you would still need a buffer/amplifier after this circuit.
This is so cool!! Thank you for this. So I understand better, what components need to change if my input is 500mV pp?
This circuit will work as-is with 500mVpp at the input.
@@w2aew awesome!! Thank you!! I'm trying to figure out a very simple way to make a compressor for a guitar pedal. It's only a small part of the design of the pedal, but one I'm stuck on... because I can't have a high parts count for such a small part of the whole. Thanks for the reply!! If you have any more tips, I'd love to hear them ;)
@@VaansWorld Just remember that the output of this circuit can't drive a low impedance load - so you'll need to buffer it to drive your next stage.
@@w2aew ah, right. Thank you!
Would using a 9V power supply cripple the circuit? Sorry for all the questions...I just can't seem to get it going
Alan, is it possible to set this circuit up so as to level the volume out for different services received on the scanner, but not so low as to need to Re-amplify them. I scan when mobile and would prefer to just feed from external speaker jack to this type of circuit and the from the circuit to an external speaker? As an option I can also feed from the scanner or this circuit into my cars audio system via a manufacturers provided input jack for ipods etc. Any thoughts? Thanks, Ted
Damn! True experts such as Dave of EEVBlog and Great Scott frequently warn us not to trust nor build just about anything we found on the Internet. But this certainly one of those true experts!
Incidentally, why do you not use the Op amp peak detector which you showed in your video #77 so that it can effectively detect very small signal and not "waste" that signal to voltage drop? Does a transistor "follow voltage" such as an op amp does with its own feedback?
There are a lot of ways to design a circuit to do this function. I just thought this circuit was interesting, and something fun to show. Certainly there are ways to do this that will perform better.
@@w2aew Germanium diode is becoming obsolete. Maybe it is really 'the sign' for me to move on. Hahahah! I cannot even find it in Farnell/Newark/Element14 anymore. It is now an audiophile object which fetches high price.
Would 1N914 or 1N4148 work as well, or will the germanium only work?
BTW- Great project!
They'll work, but not as well. The larger Vf of these diodes means that there will be more variation of the amplitude as the peak detector is being charged. Low power schottky diodes would be better, if you don't have the Ge diodes available.
Congratulations, very very good.
If I want to use 5mV on the output, what changes do I need to make to the circuit? Is there a possibility to regulate this desired amplitude of output with a potentiometer?
You can put a resistor divider or potentiometer across the output to give you a lower output voltage.
@@w2aew I am simulating the circuit on the proteus before assembling the circuit on the protoboard. But it is not working in the simulation, I regulate the amplitude of the input signal and the output signal changes the amplitude. As I do not have the Diode 1N34A in Proteus, I am using 1N4148m is this the reason?
@@lincolngerken The circuit relies on the dynamic impedance variations of the 1N34A. Using a Si diode like the 1N4148 isn't going to work as well, and will require a much higher input signal level before the compression begins to occur.
@@w2aew OK thank you. I'm going to buy the 1N34A.
@@lincolngerken In Proteus you can make and edit your own component (diode 1N34A) with the command "Make a device".
Maybe if you share your Proteus simulation file someone can implement the 1N34A for you.
what would you suggest if the purpose would be a music synth? As you maybe know with higher frequencies the amplitude of some waves like saw (ramp) or triangle have lower amplitude due to shorter integration time. Would be very appreciated if the circuit would be explained in detail like you did here. Thank you very much
I wouldn't recommend this circuit for any type of music / high-fidelity audio. This circuit adds a lot of harmonic distortion to the signals.
You re certainly right when it comes to HiFi. But Synths are another world. These harmonics you talk about make the charme of these (analogue) Synths (any why these are often more apreciated than digital synths). But the low noise I hear when the circuit adjusts it self is a issue. If I d filther these out I d cut too much low end from the synth. I have already thought about a compressor on the peak detector to boost high end. I mean there are many circuits of VCOs out there and certainly many of these have this issue and a solution to that. Unfortunately my electrical knowledge is too low to fully understand the location of the solution as well as how it works. Maybe you can pickup one that you find interesting and make a video of how these exactly works. Thank you
What is the source and mechanism of the distortion here? Diode D4?
Well, that would be one of the things to find out.
Oh, or do you mean that I have a pulse in wanted frequency that will empty the charging capacitor?
I mean that on higher frequencies (tones) the power is lower would be ok but not double frequecy gives only half voltage. Let me go deeper- A VCO gives me a square. Differtiated (? opposite of integration) I get my pulse. The pulse opens a transistor that opens the cappacitor (charged by 5V as exaple) to the ground. Buffered by a OAmp... See what I mean?
Other application would be to eliminate the attack of a guitar singnal. In other words have only the sustain. Starting point of both envelopes should, of course stay "the same" (only few ms difference). In other words a very fast compressor.
Thank you!
Another great video Alan, really enjoyed watching it. Quick question about the circuit: Is D3 superfluous? There is already a PN junction on Q2 (BE).
The circuit will work with D3 shorted also.
Take 2 hours to understand this circuit, only one thing still not very sure, the negative audio function of D1, it seem it do not help much in peak detect, as for peak detect it only need D2. for the bias part of D1, I can use 2 of D2 to replace D1. Is that right?
What happen if I remove D1, use 2 of D2 for bias
is this how guitar compressors pedals work too?
Hay can i use this circuit for an audio amp, some of my friends like to turn up the amp to 11 so i want to put this inside the amp. So i can set an max volume on an hidden pot. The input would be the the aux input of the amp, so driving devices are smartphones, mp3 players etc.
Thank you for this beautiful video! But I don't understand one moment on this schematic - circuit should have a negative feedback to keep a constant amplitude. I don't see any feedback circuit here. How it's work? How we can make attenuation exactly equal to signal rise? Or may be this feedback made by D3, D4 diodes?
It's one of the beauties of this little circuit - no feedback! The diode impedance varies in the right proportion to the detected signal level to make the divider ratio track the signal amplitude.
***** diode is only one part of voltage divider. Other part - R1. Is it adjustable?
Ilya Kalistru I haven't played with adjusting R1, but you should be able to make it adjustable over some range. However, as you make it smaller, you may get more distortion due to modulation of the diode voltage.
Hi Allen, I would like to use this circuit not only for "speaker" input, like from my Yaesu HF rig head phone jack, or speaker out put, but also a low level input. Can a "op amp" pre amplifier supply enough voltage to run the input of the circuit, or would I need to add a "LM386" power amp or similar to drive the input? (AJ6O)
+Curtis Wilson An op amp based amplifier can certainly provide enough drive for this.
I love your videos. They are complete and timeless.
Would this be a good circuit to add to the output stage of a XR2206 frequency generator IC ?
I put together a cheap kit but the output varies with frequency as expected.
I'd like to get the standard 1VP-P at the output, no matter the frequency (audio) input's amplitude.
I thought of using your circuit and play with the biasing to get 1VPP out of an LM386.
Am I going down the right rabbit-hole?
You can, but keep in mind the following:
- This simple circuit does add some distortion to the signal. You have to decide whether it is acceptable or not.
- The output amplitude of this circuit is low, with a high impedance, so the use of the external amplifier is necessary.
- This simple circuit may have some frequency dependence of its own - I didn't test for that.
@@w2aew Can you recommend to me a circuit that would be more suitable for my application?
I've searched everywhere but not too many AGC circuits around that I could understand.
I'll breadboard this one and learn from it, but still on the hunt for that sweet sweet AGC that doesn't mess-up the 0.5% distortion (THD) for this chip.
Thanks again for a great video. I learned a lot from it.
@@w2aew I found this circuit that I built today, and it works perfect.
gives me 1VPP no matter the input voltage for my El-Cheapo DIY Frequency generator.
Google search string was "dds buffer amp" and this was the first hit.
www.pongrance.com/dds-buffer.html
Just a note for anyone looking for the same solution. The wheel has already been invented here too!
Very interesting circuit. Would you say it function the same as with "Tube" based AGC ? Vari mu like circuit ? I know it's done by "rebiasing the grid of the tube but I could never figure out if the tube was cathode biased or grid biased for it to work...Anyway great video as usual !
Basically the same function (reduce gain / increase attenuation) when input signal is high - but different operating principals.
This is ONE ir my favourites videos coming from you...
GORGEOUS my friend... GORGEOUS..!!!
Do you know any book for making fx devices like compressors, reverbs...and all that stuff for audio??
THANKS in advance...!
Glad you liked the video. Unfortunately, I don't have a reference or book that has such circuits.
i like this book 'Small Signal Audio Design'
great video! you mentioned speaker level on the input. Did you mean line level? I noticed in your demonstration video you were using an RCA connector off the back of the scanner.(phono output?) I would like to use this circuit on the aux output of a mixing console before it enters a distribution amplifier.
It will likely work with line level as well, as you can see from the beginning of the video. Keep in mind that this circuit does add some harmonic distortion to the signal, so it really isn't suitable for high fidelity audio work.
just tried this @ 9v. seems to limit the loud signals but not boost the quiet, but could just be me. excellent tutorial nonetheless!
As mentioned in the video, this is NOT an amplifier. It will not boost the quiet signals, it will only level the loud ones. So, you'd want to turn up the volume going INTO this circuit so that even the quiet signals get limited, then all levels will get limited.
Do you have a BOM posted? I would like to build one up, thank you.
I did something similar for a school project. What is the bandwidth of this circuit. Thanks David Lee
The circuit is usable well into the HF range.
I just ordered the diodes and I'm going to build this. I want to use it to level the stereo output of my cell phone. I listen to music or whatever at night when I go to bed. Do you see any problem applying the peak detector output to two attenuating diode pairs? I'm thinking it will be ok.
Remember, the output of this circuit is high impedance - it will not drive a speakers/headphones directly. You'll have to put an audio amp after this circuit. Also, the clipping is non-linear, so it *does* add distortion.
@@w2aew ok thanks for the info. I'll still five it a try.
I was able to reproduce this circuit and I used an op-amp to drive the input since my sig gen wasn't up to the task without distorting. I only monitored the output on a scope. It worked as advertised and did indeed add some distortion and some noise. (no sure why). Onward and upward. Thanks.
@@SuburbanDon Please note that you will likely need an amplifier to buffer the output as well, since it can't drive low impedance loads like speakers or headphones directly.
@@w2aew I understand, thanks.
Very nice little circuit! It fits for my needs :). Thanks for posting this video sir Alan. Thumbs up for you.
Anyway,what is the value of R5? I cannot see very well here on my monitor.
R5 is 100 ohms - value not very critical.
very nice video, but I have one question. On your schematic, what are the V+ connect to R3 and T2 mean? They look like ground to me, but you put V+ next to them, are they 12V power?
Yes, those are +12V connections.
Awesome circuit! I just don't understand why the diode D3 is necessary. Would it work without it?
Thanks
+Ricardo Nunes It's just used to help set the appropriate bias current through D4.
How much distortion does this circuit add? I would like to use this to level the output from a flat panel TV before it goes to an external amp. Thanks.
From the measurements that I have made, this circuit adds about 10-15% THD (total harmonic distortion). Probably too much for a high fidelity audio system with your flat panel TV.
@@w2aew Thank you for the info.
Alan, great video on a neat circuit. I have a use for this, but in a slightly different application. In the airplane I fly, automated audio alerts are much louder than the radios, so I'd like to put this circuit between the ship and my headset. Problem is, the ship's audio amp expects headset impedance (mine is 270 ohms per transducer, paralleled in mono) and the LM38x amp expects 8 ohms. Would any modification to your circuit be necessary to accommodate these conditions?
Also, as a portable device, I'd like to power it from two 9V batteries in series. Is your specified 12V a critical value, or will 18V work just as well?
Thanks for all your great content!
I would think that the LM38x devices would work fine with the 270ohm headphones. If there's a problem, simply place a 10ohm resistor in parallel with the phones (appropriately rated for power). The AVC circuit will work over a wide frequency range. I quickly tested it from 5v to 12v, but it should work fine at 18V as well.
Thank you; reply much appreciated!
G'day Alan I'd like to use this circuit to control the volume on my TV which uses an external amp and speakers. Am I correct in thinking that for stereo use I would need to duplicate this circuit, i.e. one for each channel? Are there any issues to be aware of if I do this?
Yes, you would definitely have to duplicate the circuit - one for each channel. You'll also note that this circuit does add some distortion, so you'll have to decide if it is detrimental to your listening enjoyment. It's fine for narrow bandwidth audio like a scanner, but might be more objectionable for high fidelity audio.
Wow, this is great! I'm wondering a couple things... This looks familiar to an option for those (and by those I also mean me) that are interested in leveraging the power of their PC soundcardsto run as a budget (free) oscilloscope as an intro. With a 1V peak to peak limit can you use this to sending the signal to the mic input? I have seen a few idea circulating around but no real in depth demo video on how to implement it. I realize the limitations of the max 44k limit of the card but I would love to see someone try this is actually see what can be done. Perhaps this would be a time wasting but entertaining project to try...
You may have to attenuate the output a little for the mic input, but it's certainly worth a try. Of course, there would be no vertical calibration on your scope...
I think the Zeitnitz soundcard oscilloscope program will do that. It's actually a nice one.
www.zeitnitz.eu/scope_en
To attenuate I think a 10:1 (9M and 1M voltage divider) and 100:1 switch would be a good option. I'm leaning toward an Operational Amplifier with 1M resistors for high impedance and then your circuit as the front end. The tough part is signals below the 20Hz audible range of course and DC measurement. Some suggested the DAC in cheaper china based USB soundcards would work, Adding a bypass jumper and switch on the input cap to prevent AC coupling when wanted would be an option.
Hey I love the circuit, Allen. I wanted to use the circuit with a pair of earbuds so that it would never exceed 85 dBs. So, I was wondering if you had any tips on how to optimize the circuit for a pair of earbuds and still get a decent sound quality from it.
+Daniel Yoon (Yoonyman) Follow this circuit with a small audio amplifier suited to driving earbuds, and adjust the gain/attenuation to get the volume you desire in the buds.