My understanding of the LPB1 (correct me) & help request

[chromesphere]

Hi everyone,

I'm a self-proclaimed theory dumbass.  I've learnt ALOT with building guitar pedals, but theory is still the one thing I struggle with. 

Now that that's out of the way, I wanted to through some things, that I think I understand.  I'm not really worried about 100% accuracy here, I'm just trying to get my head around the workings of this simple circuit.

Beavis layout, scroll down the page: http://www.beavisaudio.com/techpages/blocksfragments/

So from what I understand, R1 and R3 will set the base (bias) voltage to 0.9v's (DC).   So you add a guitar signal to that, lets say (probably wrong but again, not worried bout the accuracy its just an example) 100mv's.  So you have a signal at the base that will vary from 0.9v to 1v. 

On the other side of the transistor you have a 390r and 10k resistor.  That tells me that when the transistor is off its basically an open circuit to ground.  The output voltage will be 9v's.  When the transistor is on, the output will be 351mv.   So the output signal will sway from 9v to 351mv?

Assuming im correct with this stuff, I have a few questions.

1) Why do we need the output cap, if the output signal is going to be .351 to 9volts?  Is it there to attenuate the AC signal?  (I understand how the output pot works btw)

2) Where does the transistors HFE come into this?  A 2n5088 could have a hfe of 500.  I guess if its too high it will clip, but in this situation, if you times 100mv * 500 your already up to 5volts.  Confused... 

I think there are some serious holes in my understanding of how this circuit works.  I will be getting an oscilloscope for my bday, looking for to learning more!

Your help is very much appreciated as always

Paul

[phizone]

This may help with part of the questions: http://www.beavisaudio.com/techpages/HIW/hiw1.gif

[Resynthesis]

Oddly enough, Hfe is not the most important factor, the gain of the circuit is R2/R4 which is about 25 in the Beavis circuit

[chromesphere]

This may help with part of the questions: http://www.beavisaudio.com/techpages/HIW/hiw1.gif

Thanks phizone, read it about 10 times this week.  It helped, but im still not sure about the above 2 questions.

Oddly enough, Hfe is not the most important factor, the gain of the circuit is R2/R4 which is about 25 in the Beavis circuit

That's what I don't understand.  If the gain of a transistor is 500, but you can change the gain with r4.

I think all of these questions will be answered with an oscilloscope.  Ill just wait till that shows up.

Cheers,
Paul

[Kipper4]

Paul your understanding is already greater than mine.
I have an aweful lot to learn.
I cant wait to see how you get on with a scope and not much theretical knowledge.
I want to see if it boosts your understanding of circuits or not.
Do let us know how you get on
Rich

[EATyourGuitar]

lets begin. the LPB1 is a text book case of an NPN transistor amplifier operating in the linear region. this is single ended amplification (single supply, single transistor). the voltage divider on the base sets the bias voltage for the base. this bias voltage must always be greater than the BE forward voltage (0.7 for silicon). you also need room to swing up and down with the guitar string so 0.9v is a good guess. that gives you +/-200mv headroom or 0.4v total peak to peak. the emitter resistor limits the gain but it also makes your guitar pedals sound pretty much the same regardless of temperature and Hfe. this is important for manufacturing but it is also important to reduce distortion from the gain being too high. then you have the collector resistor. this value is selected so that the collector will be right at a voltage that is half the supply when the device is only consuming quiescent current. quiescent current is the total power in amps that the device will consume with no signal at the input. this is like saying that a 10k collector resistor has the exact same voltage drop and current going through as the transistor does from C to E. so resistor current = Ice while resistor voltage drop = transistor CE voltage drop. the output coupling cap is there so it does not send any POWER (DC) down your guitar cables. guitar cables are for audio, not power. we don't want to damage the next pedal in the chain and we do not depend on it having its own cap. you can't control the other guys pedal but you can build your pedal right.

[EATyourGuitar]

I forgot to mention that the collector resistor current does not exactly = Ice since Ie = Ice + Ibe but Ib is so very very small that we assume Ice to be %99.9 of Ie

[PRR]

Do you mean this?



> R1 and R3 will set the base (bias) voltage to 0.9v's (DC).

0.818V. (Yes, 0.9V is a fine pencil estimate; just want to hear that you know the difference.)

And probably less due to base current (which is one place hFE comes in).

I'll pencil +0.8V DC for now.

> So you add a guitar signal to that, lets say ... 100mv's.  So you have a signal at the base that will vary from 0.9v to 1v.

There are several ways to measure an AC (Audio) signal. But a key feature is: audio swings BOTH ways.

So let's say your test signal swings +0.1V and -0.1V.

Then the 0.8V swings up to +0.9V and down to +0.7V and back to +0.8V every cycle.

> Why do we need the output cap

The Collector sits at perhaps +5V at idle. With signal it swings *both ways* around that point, say up to +8V and down to +2V.

We want that 3V both-ways swing.

The 5V DC idle level is meaningless as far as audio, and can be disturbing to the next amplifier stage.

When signals stay inside the box, all under one Designer, there may be just one cap between stages.

When signals leave the box, it is common courtesy to cap-couple ALL external connections. Don't let DC out of the box. Don't trust other boxes not to leak DC into your stages.

[chromesphere]

Excellent explanation PRR (as always)!  Forgot about the -0.1v on the guitar signal swing!  Thanks for pointing that out.  Ok thats another piece of the puzzle that i get, thanks for your time PRR.

(and eatyourguitar thank you too)

Paul

[duck_arse]

my theory is also minimal, but .....

the Hfe is a measure of how LITTLE base current you need for a particular collector current. low Hfe means the base bias resistors will usually be lower values (to supply more base current undisturbed) than with a high Hfe transistor, and the high value resistors mean a higher input impedance for that transistor.

I think. and probably little is the wrong term.

[chromesphere]

Hah, current...havent even begin to try and grasp how the current interacts in any electronics at all yet... lol

Paul

[EATyourGuitar]

the Hfe is a measure of how LITTLE base current you need for a particular collector current.

correct

low Hfe means the base bias resistors will usually be lower values (to supply more base current undisturbed) than with a high Hfe transistor, and the high value resistors mean a higher input impedance for that transistor.

getting warmer! input impedance is a tricky bitch when the source is a giant coil of wire in your guitar. before we factor in the guitar inductor thing, you can find a formula in a book for calculating input impedance at any given frequency. this takes into consideration the input cap, the base shunt, and the transistor impedance. there is more to it than just Hfe. the BE PN junction has its own magic for every current, voltage, frequency. usually we can make reasonable assumptions about the junction and lump all silicon junctions together with the same simplified formula. all this math just spits out a number we call Z. the frequency response and the way it responds to different guitar pickups is another layer of the onion on top of that. its good to know and good to forget. after all that I still use my ears.

[ECistheBest]

TBH, i find BJTs a lot trickier than FETs. input (base) current and its relationship with other currents is hard to understand, at least for me.

[EATyourGuitar]

BJT is easy if you only care about DC signals. this is the easy math for a BJT. doing math for AC signals depends on the caps and the devices before and after the transistor. so understanding the basic function of a transistor being a current amplifier is very very simple.

http://www.electronics-tutorials.ws/transistor/tran_2.html

scroll down to "Output Characteristics Curves of a Typical Bipolar Transistor"

that picture shows you exactly how much current the device will pass as (Ice) for any given current flowing into the base (Ibe). this is for dc only but you can guess pretty close the ac gain if you know the dc gain. every transistor datasheet has this information. this is just one example so refer to your datasheet for the transistor you are using to get the real curves. Hfe, alpha, beta are all related and they all describe dc current gain.

[PRR]

> dc only but you can guess pretty close the ac gain

Transistor does not know AC from DC.

Transistor properties are different Large Signal versus Small Signal. We normally design the DC as a LS problem, the audio as a SS problem. But mostly the differences are small in audio amplification.

With modern (post 1970) transistors, AC is DC up to some frequency above the audio band.

Designing with say 2N3906, you can assume hFE is 100-300 from DC to about 1MHz, then falls linearly with frequency down to unity at 250MHz.

We here don't care what happens above 1MHz.

When designing exotic "audio" amplifiers you sometimes have to account for action to 10MHz or above. In pedals this is super-rare. (Once in a great while you do a 3-transistor buffer which tends to howl from loop-delay around 10MHz. Usually an alternate design or different transistor cures the problem without tedious understanding.)

[EATyourGuitar]

maybe you can help me PRR, I have a circuit that has problems with high frequency amplification in phase from input to output. the transistors I'm using are PN2222 rated for 300mhz datasheet.

I built this thing as shown in this schematic without C8 and I still had some kind of high frequency oscillation even with board mounted pots, 2 layer board with decent size traces and a ground plane. I have been building them every way possible and just not having luck solving the noise issue completely. notice I added an RF filter on the input and the output that does something but still having some noise even on battery.


so I loaded it into the sim and it looks good so I can't see why the filters do not work to remove ALL the high frequency noise on the more recent pedals I built.





could it be that I have a rare case of guitar pedal 300mhz syndrome? this is driving me nuts trying to sort it out with my limited knowledge of RF. these are in metal boxes and they act funny when you hold them close to mains in the wall.

[PRR]

How do you think it id high-freq oscillation?

Forward gain is over 1,000. Amplify anything by a thousand, you are sure to hear hiss.

[mayoayox]

I want to see if it *boosts* your understanding of circuits or not.

nice one