33-year RF engineer here. I've designed many Colpitts and other crystal oscillators. This is the best description I've seen of the process in a single RU-vid video. All my engineering co-op students are going to be watching this video.
Man, this is by far the best explanation on Colpitts oscillators that I've encountered in the internet in weeks of research. Thank you very much for taking your time to put all of this together! It was really helpful!
Agree. The way I was taught in college was: Here's the circuit. It oscillates because the open loop gain is equal to or over 1 and the phase shift is zero. Figure it out... good luck!
I've been re-designing the oscillator section of a Ramsey QRP transmitter kit to eliminate chirp from the signal, and this video has been very helpful. A big thank you!
I had a look at Ramsey stuff even though I am not in US and I came to conclusion it was hastily cobbled together pieces of crop with no documentation and over priced to boot. To hear it has fundamental design issues.. what .. me surprised ??
I used this video together with the book "Foundations of Oscillator Circuit Design" and was able to design a 1M Hz oscillator that worked perfectly on its first revision. I tried three different crystals and the worst I got was 999.97K Hz. Thanks!
this outta be the most useful video I have ever seen. I was so done with learning crystal oscillators because I was confused about the gain of the circuit and the reason for the capacitors. Thanks from New Zealand
I like your diagram at 2:47 as it clearly shows what I think/perceive, when I look at this circuit. Just like there is an autotransformer effect with a tapped inductor, which can have a gain of more than 1, well there can be the same effect using the two capacitors (plus the crystal) as you showed them so the " feedback network" has a voltage gain of more than one complete with the phasing required. It is this which makes up for the less than 1 gain of the emitter follower, to full fill the maintenance condition of the oscillator, while the crystal sees to the determination of the frequency. In transfer of knowledge of the working of electronic circuits a slight modification of how to draw the circuit could make a lot of difference is " seeing the explanation" Congratulations for drawing the feedback network in isolation as you did . Normally the Colpitts Oscillator, uses the centre tapped twin capacitor in an earthed/ground mode to phase change the feedback voltage as the feedback network is normally supplied from the collector and so the phasing needs a 180 degree change to meet the condition at the base. Some people seem to think that all oscillators need a phase change of 180 degees but that is not correct as you have shown so well.
Correct! The important phase shift relationship is that the total phase shift through the feedback loop, which must be 360 degrees. I'm glad you liked my depiction of the feedback network, I think that it makes the operation of the feedback network much clearer.
I'm currently going through an class on RF equipment. I'm in the oscillator section of the class right now. I've had a bit of frustration getting my oscillators to...well oscillate. I watched this vid, built this circuit and lo' I've got oscillation at 12Mhz! This a great confirmation for me and I can now take the training wheels off! Thanks for the great vid!
awesome video! Always struggled to understand this circuit in detail. Quick note reference the crystal ESR: This crystal is an old style HC49 @12MHz so 40ohms max is a good assumption, a modern equivalent 12MHz crystal could be a 2.5 x 2.0mm package where the ESR would be spec up to 300ohms max. so some of the assumptions that this is low should be checked. Quick comment regarding tuning the Cload to make an accurate frequency: perhaps use a trimmer capacitor to find the correct value, but don't leave it on the circuit they change value a lot due to atmospheric conditions (humidity, temperature etc) so the frequency would not be stable if this is left in the circuit.
It is instructive to note that since Impedance is a complex value, "j" denotes the square root of -1. "i" is not used here because it is already reserved for AC current value...
Excellent video! One comment: The way you have drawn your schematic implies that there is always DC applied across the crystal, which keeps the crystal constantly stressed, which may not be ideal. You probably want a DC blocking capacitor (short at RF) to prevent this or a Pierce oscillator.
Question about 7:10 to 7:25. On the oscilloscope we can see there that output signal e2 (Green) changes. It is what we expect. Why input e1 (Yellow) changes so much too? I understood that the signal generator has 50 ohms impedance. Therefore the amplitude of the input e1 suppose to change, depending on the load's impedance. However, why it changes so much? Is the load impedance approaches to zero? I've never seen crystal's impedance even at series resonance is less than 30 Ohms, which means it cannot reduce the input signal from 50 ohms generator down to almost zero volts. Video is the best explanation of the Colpitts Oscillator on internet, though. Thanks.
You nailed it sir, the concepts were well explained. Just a little mistake though; the parallel combination of the base resistances should be less than or equal to 0.1*Bmin*Re
Thank you for the informative video. After 10:03 you use the terms impedances as you pointed at X1, X2, etc. X is the term used for reactance, whereas Z is used for impedance. This was a bit confusing -- they're not the same thing. Thanks.
There seem to be a few flaws here. Capacitors C1 (top feedback cap.) and C2 (Bottom capacitor) are solely for feedback purposes. The point of the topology of the C1/ C2 is that it is an impedance match between the low Z of the voltage follower and the required high load Z of the crystal circuit (or it will fail to oscillate ) . This circuit is covered in some detail in analog.intgckts.com/impedance-matching/tapped-capacitor-matching/ and the final formula is Rin/RL = (1+C2/C1)^2 where Rin is the impedance seen looking towards the Emitter resistor by the crystal circuit. If C2 = C1 (which is arrived at with two 150 pF capacitors) then the as seen Z will be four times 4.7k ohms using the as built emitter resistor or a nominal 20K ohms as seen by the crystal circuit. This business of C1 and C2 participating in the tuned circuit of the crystal. No, I don't buy it. If so, those capacitors would move the operating frequency with temperature far more than is the case - the crystal is self contained and includes EFFECTIVELY L and C components, which can operate in series mode or parallel mode. In the case of a free running true L C oscillator then C1 and C2 do participate in the tuned circuit . See en.wikipedia.org/wiki/Crystal_oscillator note that adding capacitance across the crystal will DROP the operating frequency slightly but does not set the frequency, which nearly one hundred percent controlled by the quartz crystal. Derivation of transconductance for a BJT. You have used a generic formula which applies only at audio frequencies. I looked up what is available on the 2n3903 and because Y21 = h21/h 11 (Y21 being the transconductance) you can get a value of 0,025 (mhos) amps (collector) per volt (on the base using Hfe= 100 and Hie = 4 E3 ohms. NB these values too came from audio frequency data - www.onsemi.com/pub/Collateral/2N3903-D.PDF . Finally you make a very interesting point that this oscillator works without a 360 degree phase shift (or 180 degrees plus an input to an inverting amplifier). This would be very unusual for an audio oscillator but is not uncommon with RF oscillators, particularly quartz based ones. Here's a question which I open to any competent person - . your formula for loop gain versus ESR ( Re) would not be applicable if the osc. was in parallel mode, where the Re cannot be seen and the impedance of the crystal appears infinite - does this mean that the Colpitts crystal necessarily runs with the crystal in series mode ONLY when this circuit is used ? My suspicion is that this not always the case.
Barkhausen criterion specify that A x Beta has to be strictly equal to 1. So, there is to be an intrinsic behavior of the system, that lower the loop gain to one to avoid saturation
At the point where you discuss loaded Q, you showed a series resistance of 0.6 ohms. When I did this calculation and included the 4.7k ohm emitter resistor across one of the caps, I got a series resistance contribution from that resistor of 1.65 ohms (1.65-j176) which then added to the 0.6 which comes from the 48k bias divider, for a total series resistance of 2.3 ohms. That seems significant if I'm right (but its late and I'm not sure of anything, lol).
OR maybe the 4.7k ohm emitter resistor is reduced by a factor of beta+1 so it is much less. Now I're really confusing myself. I read one article that says the amplifier presents a negative resistance to the tank.
I’m relatively new to building frequency based circuits, what type of board are you using? How is it connected to your scope ? Is this something that can be done using rg58 coax? Your help is greatly appreciated 🙏🏾
The benefit of your presentation was clearly breaking down the design of the Colpitts Quartz Crystal circuit design into the required components for setup. I was very disappointed that you didn't treat the Crystal with the adequate specifications needed to particularly design a oscillator that requires a crystal anyway. You need to consider Crystal current for ageing and phase noise for starters. Your measurement across the crystal is very inaccurate since the shunt probe capacitance will dampen the actual packaged crystal. The best way is to insert a series capacitor of about 20 to 30 pf in series and measure the drop across the capacitor. That skirts the problem and acquires the accurate Ix data. 1 microwatt is a good operating level for AT and AT strip designs and gives good starting time results. If one doesn't consider these terms the crystal requirements one reaps the results! Other cuts of quartz require more careful considerations as appropriate. Warren Dyckman, K2WD
I don't really understand how the concept of ensuring that the series equivalent resistance being smaller than the internal resistance of the crystal degrades the high Q of the crystal? Why does that condition need to be met??
Nice video! I wonder how it is possible for Colpitts oscillator to achieve automatic loop gain control? Considering the gain needs to be greater than 1 during starting period, and to be 1 during stable period.
@devttys0 … I stumbled across your channel and your videos are outstanding. I see you haven’t posted in a while, hope to see more from you at some point!
Very nice balance between theory and practicality Craig. Before I start digging thru my texts, when you approximated the transconductance you had a divisor of "26" and I don't recall you mentioning where that number came from.
+Dino Papas I didn't; I already felt there was enough math for one video! :) Transconductance (gm) is just the ratio of output current to input voltage. The output current (Ic) for a BJT varies with temperature, and gm can be expressed as Ic/Vt, where Vt is the thermal voltage, which is a characteristic voltage for semiconductors. Thermal voltage for a PN junction is equal to kT/q, which is ~26mV at room temperature.
I liked your thorough design explanation of collpitts oscillator I'm interested in resonance of a load as in hho dry cell. What oscillator would you use and how would you figure out the frequency? Furthermore the resonance needs to be locked in in other words stay resonating with changes in temperature and the voltage/ current without use of crystal as it will vary.
I found fit first time a schematic for such an self resonating circuit although I need help in figuring out what values to use for l1 and l2. Can you shed some light? Thanks.
That's the basic idea I had when trying except I was going to use a mosfet as the pulses need to be sharp and short, about 300ns and have an adjustable dead zone (no pulses). Also the pulses per cell need to peak to about 7 volts. I have 3 neutral plates
The current and voltage will be regulated to certain values like 16 volts and 8 amps. Oh here's ther basic circuit I found ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-rJQGZczJsEM.html Pic iamconnect.com/storage/a/posts/1667218578Screenshot_20221031-072900_RU-vid.jpg iamconnect.com/storage/a/posts/1667219223signal-2022-10-31-082440_002.jpeg My measured cell inductance at room temperature is 2.1 micro Henry and 462 micro farads I tried using a joule thief circuit driving a mosfet. The resonant frequency should be around 2 -5khz using bifilar coil
I have watched your video many times. Very nicely presented. One question. I have experimented with this type of circuit. If I want to use this as a small transmitter can different bias settings increase the output power. I have tried that on a simulator and noticed no change with different bias settings. So if I wanted to increase the output power as a transmitter, what kind of additional stages would you suggest. Thanks
Nice job excellent explanation,congratulations, I would like to know what happen if I replace de crystal for an inductor? how the configuration change?
this circuit wouldnt work for a 32.678 khz crystal right? i have followed all the constraints of the design and i created a circuit in multisim unfortunately it refuses to oscillate
Thanks for such an explicative video! I was wondering... what if i want to use an overtone? I want to design a 100MHz oscillator so I would have to use the third o fifth harmonic of a crystal to aproximately get that 100MHz. How do I do that? Hope you can answer! Thanks!
Hello, Indeed it is the best explanation in the net! I would like please to ask two questions: 1.You said that one of the conditions for oscillation is AB>=1 but according to your explanation AB=(e1/e2)(e2/e1), which is always equal to 1 and can't be greater than 1? 2.At minute 3:30 you present the feedback network as a serial circuit with current I, but the current I (from the emitter) split into two branch: Xc2 and Xc1+Xe? Best Regards, Marco
1. Your assertion that AB will always be equal to 1 assumes that the gain of A and/or B doesn’t affect the original input voltage (e2), but since we have positive feedback, that’s not necessarily the case. To make the numbers simple, let’s say that the gain of A is 0.9, the gain of B is 2, and the initial voltage at the amplifier input (e2) is 1 volt. The loop gain will then be AB = (0.9*2)*1v = 1.8v. That’s a voltage gain, and seems to violate the math..but remember that *this 1.8v is being fed back to the amplifier’s input, e2, so the input e2 voltage is no longer the original value of 1v*. This amplification and feedback process continues until practical circuit limitations prevent further amplification (this is referred to as “limiting”). Since further amplification is not possible due to limiting, at this point the circuit stabilizes to a total loop gain of 1, just as you predicted. 2. Yes, as I mentioned in the video, some current will go down the Xc2 branch as well. But as described, due to the reactances at the resonant frequency, a voltage gain is realized through the Xc1+Xe branch. The current through Xc2 is not terribly important to focus on, but rather Xc2’s affect on the resonant circuit, which is essential to realize a voltage gain through the Xc1+Xe branch of the circuit. Some current also goes through the emitter resistor too!
Beispiel Spannungswandler: Geht auch mit Spule und Rückkopplung auf Ein-Transistor Basis zwischen Kondensatoren sowie kleinem RC- Filter. Einzeltransitor -> NF 50 hz-100 hz Wandler und nicht der Super-IC Kram.-> dafür einfach und locker bis 100 W bei entsprechendem einzelnem Leistungstransistor bei 220V sec. und entsprechendem großem Normal-Trafo und größerem Kondensator (sec und prim vertauscht) zum Wandeln. Der Ringkerntrafo ist gar nicht so toll hierfür. Ok. Eine Bistabile Kippstufe aus zwei Moosfeet Transistoren von 100 A , liefert noch mehr als der Colpit alleine.-> 1000 W. Diesen kann man mit erwähntem Einzelfilter und Rumprobieren aus Rechteck zur Dreieck-Sinus-Mischphase kombinieren und spaart so viel an abgehobener Regeltechnik teurer Geräte mit ewig vielen Einzeltransistoren. Hier benötigt man nur 2 Moosfeet und einen Colpit zum genauen Phasenabgleich. Dazu eine einfache Drucksicherung von Siemens. Ok es gibt viel ausgereiftere Schaltungen, aber das Grundprinzip funktioniert immer noch.
How do you make an amplifier for a this crystal oscillator circuit? I'm trying to increase the voltage to about 6volts and have an oscillation of 32mhz
+dh1ao Phase shift is caused by reactance (capacitors and inductors), which delay signal propagation. Purely resistive components/circuits will not cause a shift in phase. The key to a resonant circuit is that at some frequency (called the "resonant frequency"), the negative reactance of the capacitors and the positive reactance of the inductor (in this case, the inductive crystal) are equal in magnitude. But because they are opposite in sign (capacitance causes a phase shift in one direction, inductance causes a phase shift in the opposite direction), they cancel each other out, and the circuit appears purely resistive, since all circuits have some inherent resistance. So, at the resonant frequency, the feedback network is purely resistive, and has no phase shift; this means that the input and output voltages (e1 and e2) will be in phase.
Hi .. these video are dam super..nice. explanation with real time ckt results...want more videos on Jim William analog IQ testing. problems....thanks for these video's
@35:32 Is it really possible to get two voltage levels at the same point (the Crystal is parallel to the 120k resistor as well as the series of C1 and C2 of which C1 is connected to the base of the transistor)? Now since the base is set to 5.2 V by the voltage divider, how did you get 3.6 V at Crystal unless you are saying that the base voltage would drop to this same level?
Your feedback gain analysis assumes that the xtal only has an inductor, which is totally untrue. The xtal is a self contained resonance unit having a series cap and a parallel shunt cap. Your other assumption is that the xtal is resonating with the cap C1 and C2, which is also untrue.
Well I wouldn’t go so far as to claim that this is *my* feedback analysis. It is effectively a regurgitation of circuit analysis from several respected sources, listed in the references section of the link in the video description (the book by Marvin Frerking in particular). You are correct that there are series and parallel capacitances that are part of the crystal, so the model presented here is approximate. I’ve even made a video on measuring the real-world motional parameters of crystals. However, in this configuration, at the frequency of resonance, the crystal appears *primarily* inductive to the circuit, and thus these capacitances were ignored (I mention this more explicitly in my video on Pierce oscillators). The parallel capacitance of the crystal in particular is typically a few pF, and is effectively just the stray capacitance of the crystal holder and leads. Every reference I’ve ever seen on the Colpitts circuit specifically states that the inductance of the crystal is resonant with capacitors C1 and C2 in this circuit, so I must admit that your statement to the contrary confuses me. In fact, you can replace the crystal with an actual inductor to form a Colpitts LC oscillator. Surely we can agree that the series capacitance of C1 and C2 forms a parallel tank circuit with the inductance of the crystal? So it must be resonant at some frequency. If not resonant at the frequency of oscillation, then what frequency will it be resonant at? And if not resonant at the frequency of oscillation, how is voltage gain realized through the feedback network?
devttys0 first of all thanks for replying! yes everything works out to be true if the xtal is modelled as an inductor. However, the xtal is precision cut for a certain frequency. The xtal resonates at a frequency given by the series inductor and capacitor in the model. This is obviously as precise as a xtal frequency can get. Why would anyone want to load the caps on to the xtal and make the inductor resonate with the loading caps that have 5-10% tolerances on them? That takes away the main advantage of using an xtal. So maybe your analysis may be correct for the way you intend to use the xtal, but the question is whether this is really how an xtal is supposed to be used, given precision requirements. Furthermore, how sure are you that the xtal is resonating with the load capacitors and not with itself? I guess one way is to take away the load caps and see if it still resonates. But then another question is, what if the purpose of the load caps is to complete the feedback loop and not to resonate. I'm genuinely interested to understand this.
All excellent observations and questions! Crystals are sold as either “parallel” or “series” resonant. There is nothing fundamentally different about these two types of crystals. Parallel crystals are simply cut to resonate at the frequency stamped on the outside of the case when a particular external capacitance is placed in parallel with the crystal; the value of this external parallel capacitance is specified by the manufacturer. Series crystals are designed to resonate at the marked frequency with no external capacitance. There are oscillator topologies, such as the Butler oscillator, that are series resonant, while the Colpitts is parallel resonant. Obviously a series crystal will work just fine in a parallel resonant circuit, and vice versa, but if this is done then you are correct in your observation that the actual frequency of oscillation will be slightly different than what is marked on the crystal. If you place any crystal in a series resonant oscillator circuit, you will see its series resonant frequency (the frequency of the series capacitance and inductance inside the crystal); take it out and put it in a parallel resonant oscillator circuit, and it will oscillate at a different frequency (although this frequency difference is usually around 1% or less, and hence is typically measured in parts-per-million). You are also correct that the external capacitances (and inductances!) in *any* oscillator will affect the frequency of oscillation, and that this is generally undesirable. Manufacturing and temperature tolerances in these capacitances will absolutely change the frequency of oscillation. But, there are a few things to keep in mind. First, the crystal itself has manufacturing, temperature, and aging tolerances, so even if you could construct a ideal, perfect oscillator circuit, the frequency of oscillation will never be perfectly exact. Second, there are always unknown and/or poorly defined capacitances in a circuit that will affect the frequency of oscillation. Consider stray capacitances of the circuit board, input capacitances of your oscillator’s amplifier, etc. Many of these will also vary with temperature. Third, the typical alternative to a crystal oscillator is an LC oscillator. These tend to be very unstable in frequency (compared to a crystal oscillator), but this is primarily due to the inductor, not the capacitors. So although capacitors are not perfect, replacing the inductor with a crystal removes the overwhelming source of instability. So, when designing an oscillator circuit, you should never expect it to oscillate at *exactly* a particular frequency. If very specific frequency is required, you typically must include trimmer capacitors to manually tweak the frequency after construction. You also want to choose external capacitances with very good temperature and aging characteristics; this is why very precise oscillators are often ovenized to provide a constant operating temperature, and why they will periodically require re-adjustment! However, the reality is that for most applications such precision is not required, and a run-of-the-mill crystal oscillator will work just fine without any tweaking or fuss. Oscillator topologies such as the Pierce and Colpitts are relatively easy to construct and work “good enough” for most applications, hence their popularity.
Can you elaborate? While the quartz crystal does not have a physical inductor (i.e., a coil of wire), it absolutely does exhibit inductance (as well as capacitance, and resistance). You can, in fact, replace the crystal with a physical inductor to create an LC oscillator, albeit with much less stability than one which uses a quartz crystal.
i will try to test the circuit with out crystal i will put frequency from signal generator to transistor base in same crystal frequency and see if the out become bigger than the input signal that mean the c1,c2 is help the input with the tune frequency as i think . thank you for repay
C1 and C2 (as well as the inductive element, either a crystal or a wire coil), *are* there to help provide frequency selectivity; together they form a tuned circuit. As described in the video, they provide voltage gain due to the opposing capacitive and inductive reactances, but *only* at the desired frequency of oscillation. Frequencies above and below the resonant frequency are phase shifted / attenuated, preventing oscillation at those undesired frequencies.
Yes, you can! It won't be as stable as a crystal oscillator, but using an inductor results in an LC colpitts oscillator. www.electronics-tutorials.ws/oscillator/colpitts.html
thank you very much :) I made a different implementaion with the LC tank as the collector load of a common source configuration with feedback cap from the emitter to the common point of the colpitts capacitors. I like your version better because mine has very high output resistance as I take the output from the collector and I cannot seem to feed the signal to a comparator. Maybe Ill buffer the signal first, because I cannot get your implementation to work :/
Thank you for the video! Here is a little question. I have a book that claims that Colpitts crystal oscillator supposed to have an RFC (>= 10 mH inductance coil) connecting transistor collector and Vcc (example: afiskon.ru/s/9b/04bdefe3e8_3.png ). The role of RFC is to provide stable DC voltage to the amplifier. However some sources including this video don't use RFC. Is it not that needed or it depends?
I'm not aware of that being a requirement, and I can't say I've seen that stated in any book on oscillator design that I've seen. I don't see an advantage of having an RFC as a collector load to the transistor, unless you were planning to take the output from the transistor's collector. 10mH is VERY large, especially for the frequency range specified in the schematic you linked to; I'm sure you could get away with a much smaller value inductor, or even a simple resistor, if your intention was to take the output from the collector. For very low phase noise oscillators providing a clean and stable voltage source is important, but typically I've seen this done using either simple RC filtering, or active filtering, e.g., a "capacitive multiplier" circuit. Wenzel Associates, who specialize in oscillators and frequency sources, has a good write up on the subject of voltage regulator noise: www.wenzel.com/documents/finesse.html. I'd be interested to know what book you saw that in?
I see, thank you. The book is "Practical Electronics for Inventors, Fourth Edition" chapter 10 "Oscillators and Timers" www.amazon.com/Practical-Electronics-Inventors-Fourth-Scherz/dp/1259587541/ Currently I'm trying to figure out how to make an analog AM/FM transmitters/receivers that would use carrier waves around 100-443 MHz. I've read a few books on SDR and DSP some time ago, also I managed to build basic RC filters, Butterworth filters and a Clapp LC-oscillator. As I understand I would probably need an analog multiplier. Anyway could you please recommend something to read on the subject?
15:05 That seems a bit fishy to me. The "this is relatively small, therefore we set it to zero" move shouldn't be applied selectively. Doing that you can get a lot of nonsense results. Of course the result in this case is alright, but I'd argue it differently. I'd split the sum with canceling Re in the left half, giving -j*X2 + X2²/Re and the argue that the right half is way bigger (X2² > X2 and 1/Re both big), ignoring the left half, giving the same result.
Like the videos but where you get all the numbers gets really confusing. Better then some of the other videos but for a beginner like me kind of hard to follow. I understand some of the numbers and where you got them. Maybe put the math along with the diagram would help or more step by step. this point U= X*Y*Z....
+Martin P Thanks for the feedback Martin! Can you give some specific examples of where the numbers got confusing? That way I can be more explicit next time. :)
+devttys0 , at 1 min mark, I see where VCC is the input voltage and ground, but where is the output? Maybe add electricity flow direction. I see where it sort of goes but but beyond where the gate gets charged. I assume the gate opens it now flows to the crystal from e1 to gnd and down to c2 through the crystal to the gate and to C1 C2 ground and crystal @9:42 I guess the output would be from e1
+Martin P Thanks Martin! To answer your question, the output is typically taken from the emitter in a Colpitts oscillator, but it can be taken from any point in the circuit (except Vcc and ground, obviously). You might also want to check out this Colpitts oscillator animation: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-R-bMTlVF0Uk.html.
Oh my god, I've been trying for weeks to understand oscillators, been reading through Grob's Basic Electronics and The Art of Electronics and they completely fail to EXPLAIN how this oscillator works. Most places they just toss around some circuit diagrams with placeholders "C1" "L1" and handwaving explanations like "the network provides feedback" or "the capacitive voltage divider" YEAH YOU FUCK A DIVIDER CAN BE ANYWHERE FROM 1:1 to 1:10'000 SO WHAT DO YOU DO?! Thanks man, really appreciated. I'd like to donate, where I can I donate?
