Formulas and Components for noobs

[PRR]

> How do they pull that off?

I only mean that each gallon drawn from the mains is filtered and re-used dozens or hundreds of times. Not that a car-wash is a net water producer.

But for purposes of *this* discussion, pretend it is the old-fashioned car-wash which draws a huge load of fresh water for every car washed. You could not feed such an operation from a household supply line. Perhaps a 1.5" line, and fed from a 12" main or the short-end of an 8" main, so steady car-wash load would not drop mains pressure if a fire had to be fought.

[bluzeyonecat]

Lol! Well PRR, I can tell your a troubleshooter and a marksman at it. After looking at it today I realized how close together those breadboard holes are! Lol! Yep, stray connection. After fixing it, worked like a charm.
So while my mojo was working I thought I'd try the sine wave schem Gibson shared with me. That one didn't have a good ending.
Two quick questions about it. First, does the second transistor on the right have 9v going into its collector? Seems like alot of juice. Second, what is that "100" and "22" going to ground at the bottom? I assumed they were resistors just not sure of the value. Thanks guys. I'll post the link.

[bluzeyonecat]

Here's the link to the schem in question. I was working on the top one first but hope to eventually build the bottom.
http://www.geofex.com/FX_images/q&dosc.pdf

[antonis]

First, does the second transistor on the right have 9v going into its collector?

Yes, it does...
Q2 doesn't mind a lot because it's collector  current is set by it's Emitter resistors and Q1 collector voltage...

Second, what is that "100" and "22" going to ground at the bottom? I assumed they were resistors just not sure of the value.

Yes, they are...
Τheir values are in Ω (Ohm) and you may also meet them as 100R & 22R or 100E & 22E - generaly R, E or nothing --> Ohms..

[GibsonGM]

You can simply put a 10K resistor there, from the emitter to ground, to get 'full voltage' output from it.    I put a 10k pot on mine.

[bluzeyonecat]

Ok. Thanks for clearing that up a bit for me. That 22R value threw me a curve. Thanks a bunch!

[suncrush]

Are we ready to talk about tubes yet?  I really want to talk about tubes!

The most common type of tube used in stompboxes is probably the 12A_7 family--the 12AX7, 12AT7, and 12AU7 are probably the most common varieties, in order from most gain to least. Others exist.

These are dual triode tubes, meaning they contain two seperate triodes in one device.  That means you get 2 gain stages per tube. They work similarly to inverting op amps. Gain is controlled by putting a resistor between the positive terminal of the power supply and the plate (anode) of the triode.  Signal is applied to the grid, and the tube amplifies it, outputting at the plate.  One typically connects the output of the first triode to the grid of the second triode, using both sides for amplification.

Unlike op amps, tubes require a separate heater circuit, as high temperature is required to "boil off" the electrons that make the tube work.  In 12A_7 tubes, you can use a 12-V potential that goes in pin 4 and out pin 5 (or vice-versa; the heater is agnostic about polarity.)  Or, you can run the heaters in parallel at 6 V, in pins 4 and 5, and out pin 9.

In conventional use, tubes are high voltage devices.  The plate is spec'ed to run at 100-250 V.  In stompboxes, it's common to use "starved-plate" designs, with the plate at 9-18 V.  The tube still works, and still provides amplification, but the signal breaks up sooner, giving more distortion.  That said, it's pretty easy to build a decent clean boost, especially at 12-18 V.

People say tubes are hard to work with, but I don't buy it.  They're not any harder than transistors.

[bluzeyonecat]

Cool! Are those "starved plates" also referred to as "hot plates" ? If so, I played in a band and the guitarplayer had one rigged. I remember him talking about it.

[GibsonGM]

I think the guitarist might have meant a Hot Plate attenuator, which you put between the amp & speaker, Bluze.   "Soaks up some signal" to let you push the amp harder at a lower volume.   AKA  "power soak". 

"Starved plate" just means operating a tube at too low a voltage for proper electron flow from cathode to anode, squishing the device's curves and making it act "non-linear".     

The major 'difficulty' I find with tubes, which IS notable, is that they're not compact...you can't just pop 'em on a PCB (not easily...).  You can put 4 BJT's in the space one tube socket occupies.   And to do them 'full scale', you have to deal with the power supplies, of course.

I do like the sound, however  

[suncrush]

I think the guitarist might have meant a Hot Plate attenuator, which you put between the amp & speaker, Bluze.   "Soaks up some signal" to let you push the amp harder at a lower volume.   AKA  "power soak". 

"Starved plate" just means operating a tube at too low a voltage for proper electron flow from cathode to anode, squishing the device's curves and making it act "non-linear".     

The major 'difficulty' I find with tubes, which IS notable, is that they're not compact...you can't just pop 'em on a PCB (not easily...).  You can put 4 BJT's in the space one tube socket occupies.   And to do them 'full scale', you have to deal with the power supplies, of course.

I do like the sound, however  

on the other hand, you can do a lot without PCB's, like wiring the filter cap from the plate of diode 1 to the grid of diode 2. You can solder up a simple signal boost circuit by wiring everything directly to the tube socket, no PCB at all.

[GibsonGM]

Sure, point to point!     They also have a lot of 'stage appeal', in that you can show them off!

[bluzeyonecat]

Yea.. Still trying to wrap my mellon around these transistors. They're a breed of their own for sure.
I don't think this falls into the "proper" category but I've taken some of my test pots and soldered a 330R between pin lug 2 and 3, giving them "built in limiting resistors". Lol!
Aside from that, since the transistors can function as a switch, is there a way to ramp them up to be a tone generator? Maybe thats whats going on in the "tank" section of the schem. I also read that it is possible to turn an op amp into an oscillator. That would be a fun build if possible.

[GibsonGM]

Like I'd mentioned, Bluze, the 2 caps and 2 R's, each with R or C to ground, make up the "Twin T" oscillator tank.   So yup, you certainly use them for tone generation, switching and amplifying... The transistor is driving the oscillator...the combination of R and C are resonant at whatever frequency you set, and the amplifying device gets it going/replaces energy lost to resistance.    The 2nd transistor buffers the signal to it doesn't distort when you put a load on the rig.  It's like pedaling a bike, in a way...a flywheel.

Easy way to think about transistors, along with looking up a lot of Youtube videos on basic transistor circuits which are very very helpful, is to picture the collector and emitter as a pipe, whereby water flows from top to bottom ('conventional current', not electron flow).    The base is controlled by a much smaller current that you make flow from base thru emitter....it's like base current opens a valve or floodgate, and lets more/less collector current flow along with it.   The base current is multiplied by the transistor's gain to give collector current.      

It is essentially a variable resistance..."TRANSferred resISTOR".   It operates on current, not voltage, so there is 1 great reason to keep on beating on Ohm's Law.  You need it to set the current.   Gotta be careful about going too far out there with BJTs and opamps until you can do Mr. Ohm well, or IMO you may get really frustrated not knowing what's going on w/circuits.    But by all means build things and have fun with it!  lol

[ashcat_lt]

They work similarly to inverting op amps.

A tube by itself is absolutely nothing like an opamp!  A single triode is much more (almost exactly) like a transistor - basically a resistor whose value is controlled by the grid voltage.  You can string them together to make an opamp the same way you can do with transistors, but that's a different story.

[suncrush]

True.  I meant more in the way you control gain.  I wasn't being very specific.

[ashcat_lt]

IDK, I think you control gain in a tube about the same way you do in a transistor.  Basically, you figure out how much its resistance will swing with a given control voltage and then decide how much of what part of the voltage divider you want that to account for.  Just like a transistor, you can take your output from either the top or the bottom, put a resistor above or below or both, and it divides down the supply voltage.

Opamps are completely different.  Ultimately, they do divide down the supply voltage to create their output, but we usually ignore the ways that that happens.  It's just a triangle. 

We use voltage division to control gain in non-inverting opamp stages, but things are a bit funny.  I call it Bizarro World, because everything is kind of upside down and backwards.  In fact, the gain equation for a non-inverting opamp is exactly the inverse of the voltage divider equation for the two resistors in the feedback loop.  We divide down the feedback, and the opamp applies enough gain so that it still gets to the inverting input at the same level as what's coming in the other. 

I'm not sure how I can apply what I call PRR's rule to the inverting opamp setup, though.  I can almost see how it's got to work, but I sure as heck am not ready to explain it to anybody.  It's easier to switch over to current here.  The current through Rf must always be the same as Rin (until the opamp hits its limit) in magnitude but opposite polarity.  They have to add to 0.

[PRR]

> voltage divider
> voltage division to control gain in non-inverting opamp
> inverting opamp setup

Voltage dividers are like levers. (Actually a lever is more like a transformer, resistor voltage-dividers are very limp/lossy levers.)

I stuck a pry-bar in my tractor today. 1 inch from cab to frame, 30 inches from frame to hand. 1:30. A 10-pound hand force caused 300 pounds of cab-lift force.

Recall from dummy-science that levers come in several "classes" depending where the hand, fulcrum, and load are. I never remember the classes, but I know where to put the three points to pry with good ratio.


The passive voltage divider has the hand at the top, fulcrum at bottom, and load in the center. Load moves less than hand but in the same direction.

The NI active (op-amp) connection also has fulcrum at bottom, but hand (input transfered through opamp + and - inputs) is in the middle and load is at the top. Load moves more than hand.

The inverting op-amp gain-set network has the "fulcrum" in the *center*. Like a see-saw. One end goes up and the other end goes down, the distance (voltage) in proportion to the arm-lengths (resistances/impedances).

[bluzeyonecat]

That mechanical analogy really helped me "see" it better. Thank you.
So, skipping over to a LPF for a sec. (got a feeling this thread could jump all over the place ) what is the formula for determining...hmm..determining how to set it up for your needs? Resistor value with cap value Im sure somehow. Thanks for your time.

[GibsonGM]

A passive LPF, first order, is a series resistor followed by a cap to ground.   Output is taken at their junction.

The cutoff frequency is found by:  Fco = 1 / 2pi (R)(C)    F in Hz, R in MEGS*, C in uF
*R is in MEGS because this allows us to not have to convert the cumbersome uF to something we can work with!   So, 100k = .1  in the formula


You can re-arrange this, like with Ohm's Law, to find what R you need for a chosen C, etc.  <   Good to do this, as you won't find as many values for C as we commonly have for R!

To see how this works, it's really good to sim a simple RC LPF/HPF with LT Spice or another simulation program....the curve you get is not quite as radical as you may expect, but for our uses it's pretty good.  

Try finding a Youtube vid about this, I bet there's one out there where you can SEE someone set one up
------------------------------
Ex: want to pass frequencies below 200Hz.  Pick a cap... .1u
R = 1 / 6.28*200*.1  = 1/125.6 = .00796  Multiply by 1,000,000 since we used MEGS and uF, so we get 7962 ohms, or 8K. 

To be honest, I don't do this manually, I wrote a VB program that does    If you're not good at moving unit around, you can easily make errors. I do. So I check it:

200 = 1/6.28 (.007MEGS) (.1u)    ~ 200Hz.    So I know I did it right.

 

[amptramp]

One formula that some pedal builders may understand if their other half objects to the time / money / attention this hobby takes: