Transistor clipping

Started by nonoxxx, March 23, 2016, 11:27:47 AM

Previous topic - Next topic

nonoxxx

Hello,
I am a long time reader of this forum but first time poster :) , and it makes me want to try making some stomps. I read a lot of theory and circuit analysis (I had some basic electronic knowledge) so I understand well the clipping theory for opamps with diodes.

But recently I take a look of the schematic of the KRANK Distortus Maximus (really nice )  and it seems that the distortion is only done with 1 transistor (2n5088) and no diodes?  How just a transistor can have such gain??
I have searched some infos or documentation on transistor clipping and all the schematics where with diodes or leds.

nocentelli

The transistor boosts the guitar signal, but it is fed into a 386 chip amplifier running at maximum gain (x200) which easily distorts with even light strums.
Quote from: kayceesqueeze on the back and never open it up again

nonoxxx

I understand ,so it's the transitor's signal boost to the circuit that make the opamp distort .
Thanks :)

induction

Quote from: nonoxxx on March 23, 2016, 12:12:25 PM
I understand ,so it's the transitor's signal boost to the circuit that make the opamp distort .
Thanks :)

Just for clarification, the 386 is a power amp, not an op-amp. It has a different pinout and will distort differently from an op-amp. Trying this topology with an op-amp will generally produce a very different, and generally less pleasant type of distortion.

antonis

#4
Quote from: nonoxxx on March 23, 2016, 11:27:47 AM
How just a transistor can have such gain??
Allthough this isn't the point here.. :icon_wink:

A transistor can have a maximum voltage gain of 20 times the power supply..
(w/o emitter degenaration & assuming a collector bias point at half of PS)
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

nonoxxx

Thanks , I was just reading the LM386 documentation and realize it's a power amp , and just to be sure the gain on the LM386 is set at full by joining pin 1 and 8 together but on the documentation the schematic for this use a 10uf cap,
I think krank don't use it to not alter  frequencies? am I right?

induction

#6
Quote from: nonoxxx on March 23, 2016, 12:12:25 PM
I understand ,so it's the transitor's signal boost to the circuit that make the opamp distort .
Thanks :)

One more thing I forgot to mention. The transistor boost does definitely increases the distortion from the 386, but the gain from the 386 will provide plenty of distortion on its own. Take a look at the Lovepedal Purple Plexi for example.

Edit:
Quote
Thanks , I was just reading the LM386 documentation and realize it's a power amp , and just to be sure the gain on the LM386 is set at full by joining pin 1 and 8 together but on the documentation the schematic for this use a 10uf cap,
I think krank don't use it to not alter  frequencies? am I right?

A smaller cap would limit the amplification of lower frequencies, but the 10u cap is big enough to amplify the whole guitar frequency range. I suspect the cap is used in the datasheet to cut out frequencies below the audible range that might interfere with the circuit or cause noise from moving input cables, etc., but I'm not sure. Some folks have reported that they prefer the sound with a resistor or jumper between pins 1 an d 8 to the sound with a cap, but that's probably a personal preference thing.

nonoxxx

Thank's dudes your answers are awesome

GibsonGM

Yeah, your 10u is probably going to pass most of the audio spectrum, noise included.  Now, it's not as big a deal on something this small, but sometimes you WILL want to use a network that will cut bass more in order to make the distortion "not suck".   The other thing to think of is dumping all audio freq's below anything useful (80-100Hz for sure, maybe even higher).  This helps take care of power supply noise.

Between stages of tube preamps offers a good place to think about that...too much bass can encourage blocking distortion (farting....) and can also make your OD/distortion sound 'flubby'.   Don't need much low end to make a guitar sound good!   Sometimes if your OD is not quite as nice as you would like, taking some bass out might make a pretty big difference...

Just some extra info for nonoxxx...welcome to the forum! 
  • SUPPORTER
MXR Dist +, TS9/808, Easyvibe, Big Muff Pi, Blues Breaker, Guv'nor.  MOSFace, MOS Boost,  BJT boosts - LPB-2, buffers, Phuncgnosis, FF, Orange Sunshine & others, Bazz Fuss, Tonemender, Little Gem, Orange Squeezer, Ruby Tuby, filters, octaves, trems...

PRR

> on the documentation the schematic for this use a 10uf cap

The output must be biased to _half_ supply to meet the Power Output numbers on the Spec Sheet.

There is an uncertain +/-10mV DC input offset. (Probably less, maybe more.)

In the Low Gain (20X) connection, this 10mV input DC error is multiplied to 200mV or 0.2V output DC error. On 6V supply, you want 3.0V DC output for full power. You might get 2.8V DC. This is close-enough for practical purpose.

In the HIGH Gain (200X) connection, with a short, the 10mV input error could be 2 Volts output error. "3.0V" with 2V error would be practically slammed to one side. Output power would be WAY down. Buyers who test clean power output could be unhappy.

High gain with a cap (high gain for audio but low gain for subaudio/DC), the DC error is back to like 0.2V, everybody happy.

Most modern '386s won't have 10mV input error (chip cooking has got a lot better than when the '386 was born). Also we generally run the higher 9V supply, so the hi-gain error still leaves a little swing. And for most guitar '386 projects we WANT distortion, and can be perfectly happy if one side of the wave is bashed more than the other.
  • SUPPORTER

nonoxxx

Thanks again,
Your helped me to solve the mysteries of my favorite distortion :). Generally for High gain stuffs,  I prefer a screamer or clone in front of an amp's lead channel but this Krank's or clones (body rot)  seems to sound really interesting (tight , clear , lot of balls).

Another noob question, which cap's brand in this list do you recommand ? http://www.banzaimusic.com/Film-Foil/
Yesterday I wanted to order part's (good cap's and resistors) but there is too much choices I am a little lost :)

GibsonGM

Mallory, Vishay, Xicon are well-known manufacturers...sometimes in expensive amp builds, I will use Mallory's just *because* - more to do with a "standard" than performance...but being honest with you, there really isn't much audible difference!  (based on manufacturer, I mean)  (Bring on the flames, I can take it, ha ha....)

IF a cap is built correctly - and most major manufacturers will do a decent job - then you are left with TYPE of cap...electrolytic (noisy, workhorse...), tantalum (not much better)...ceramic (blah)...POLY - desirable for most audio path applications...yes, we can debate that there is SOME distortion of your signal in many of these caps, and maybe the $3 unit doesn't show this the way the 10 cent one does...but as for audible difference?  For guitar?? 

True audiophiles might enjoy sniffing different cap brands, but if the manuf. is reputable, my opinion is that you are fine choosing them by TYPE...and poly types are great for our purposes...ceramics for pF range stuff are fine, too.   Many opinions on this, ha ha....check out Smallbear, see what they're offering...millions of those caps are all over, from Mouser to Tayda and in between...
  • SUPPORTER
MXR Dist +, TS9/808, Easyvibe, Big Muff Pi, Blues Breaker, Guv'nor.  MOSFace, MOS Boost,  BJT boosts - LPB-2, buffers, Phuncgnosis, FF, Orange Sunshine & others, Bazz Fuss, Tonemender, Little Gem, Orange Squeezer, Ruby Tuby, filters, octaves, trems...

nonoxxx

Ok so for price , size and available capacity I think I will go for Xicon caps (greenies) or panasonic metal film, any other suggestion in this list?
http://www.banzaimusic.com/Film-Foil/ ( I live in France so it's seems that this german shop is the more convenient considering their price, stock and shippingcost.