Clipping diodes after jfet booster goes silent.

[allmektig]

Earlier tonight, I did some experiments with diodes as a passive clipping device straight from the output on the guitar.

I thought that boosting the signal before clipping it would result in a stronger effect, so I went to look for a booster I could build with the parts I had.
I decided to go for this circuit:
www.beavisaudio.com/techpages/blocksfragments/images/JFET_Booster_Schematic.gif

Then I plugged in the little device I built earlier tonight, wich in effect just put two diodes in parralell, pointing each way, from signal to ground. Then it all went silent.

I'm trying to learn a bit more about pedals so I can start to build some of my own, but this has me stumped.

Can anyone explain why this happens?

Edit: It doesn't actually go completely silent. The volume is substancially reduced, I get REALLY short sustain and the only sound I get from it is a kind of crackling sound.

[JohnH]

Does the booster work properly on its own?

Such JFET circuits can make a very good overdrive on their own - try two end to end. Heres one I built:

guitarnuts2.proboards.com/index.cgi?board=fxschemes&action=display&thread=4987

JFETs are very inconsistent however, and the best component values vary for individual JFETs. And with a J201, I would think that R4 should best be a higher value. You should select it so that the voltage at the drain D, is just over 1/2 the battery voltage. also, try a higher value for R6, and try without C4, or with a resistor in series with it, cut reduce gain.

John

[allmektig]

Yeah, it works on it's own.
I'm just trying to figure out why adding the diodes would cause it to crackle and behave strangely instead of just clipping the signal.

What's the purpose of r6 in this circuit?
I thought it was just to reduce volume or something related to that, since a volume pot would be in the same place, so I took it out to see what happened, and it seemed to work pretty well with no r6 at all.

[allmektig]

I did as you suggested, and removed c4, increased the value of r4.
With no diodes, it works good.
When I add the diode box I built, wich places two diodes with different polarity in parrallel with r6 it sounds like this:
soundcloud.com/a-melodyc-shivering-form/wtf

Edit: I tried a couple of diodes directly on the breadboard, in parralell with r4.
When they faced towards ground, I just got that spluttering crackling sound, when they faced the other way it didn't seem to have any impact on the sound at all.

[JohnH]

Putting diodes across R4 will definately mess with it, making the biasing of the JFET wrong

R6 seems like unnecessarily too small a value as a pull down resistor for the output cap, but if it was a volume pot it would make sense at about 50k to 100k.

Another trick with this type of circuit, try small caps in place of c4, for a treble boost in place ogf an overall gain boost.

J

[col]

I'm curious: what happens if you reduce the gain of the booster?

[col]

And, just to check: the diodes are definitely wired between hot and ground, and are not in series with hot!?

[JohnH]

Jan 20, 2013 2:20:22 GMT -5 col said:

And, just to check: the diodes are definitely wired between hot and ground, and are not in series with hot!?

Thats right. But, ii suspect t may sound better with a series resistor on the output, then the diodes from there to ground.

[allmektig]

There was an error in my last post.
I didn't put the diode in parralel with r4, but with r6, the 51k resistor at the output.

I tried to use an output transformer, and put the clipping diodes after that, and that worked with no problem.

What resistor would I replace with a pot to make the gain adjustable in this circuit? And could I drive it so hard as to make it clip? Perhaps adding a second jfet, and have the first drive the second?

Just trying to get a grip on how these things work, instead of just soldering other peoples designs.

[JohnH]

Jan 20, 2013 15:27:00 GMT -5 allmektig said:

There was an error in my last post.
I didn't put the diode in parralel with r4, but with r6, the 51k resistor at the output.

I tried to use an output transformer, and put the clipping diodes after that, and that worked with no problem.

What resistor would I replace with a pot to make the gain adjustable in this circuit? And could I drive it so hard as to make it clip? Perhaps adding a second jfet, and have the first drive the second?

Just trying to get a grip on how these things work, instead of just soldering other peoples designs.

the 51 k can be replaced by a pot to give variable output volume - or put a pot in series with C4 to control the gain. Drive these hard and they crunch nicely, and sound much better than diodes - see the circuit I linked to before, which has two such stages, followed by a buffer - which is not esential.

[allmektig]

Could a busted output cap let dc trough, wich would then go to ground via a diode from output to ground?

I'll have to test that when I get home.
How would changing the value of that cap change the sound?

What's the role of r5?

[JohnH]

Jan 20, 2013 16:17:11 GMT -5 allmektig said:

Could a busted output cap let dc trough, wich would then go to ground via a diode from output to ground?

I'll have to test that when I get home.
How would changing the value of that cap change the sound?

What's the role of r5?

If c3 is leaking dc, then the diodes would mess with the voltages on the JFET, stopping them being biased properly, and explain a failure to work.

c3 can be smaller, particularly if R6 is larger. If it is too small, it will start to cut down on bass. Again, if you study my circuit you will see very similar arrangements.

R5 and R4 set the bias on the JFET, so the steady state voltage at the drain D is just above mid supply voltage, allowing it to swing up and down with a good range. For a given JFET it is the ratio of these two that is important. You can double each one and it will still work and use half the power. But, the output impedance of the circuit is equal (almost) to R4, so as shown, it can drive fairly low input resistances on a mixer for example, But for a guitar amp input, it can have a much higher impedance.

There's a sonic effect too. I find that the lower you have r4 and r5, the softer the clipping is when the JFET gets over-driven. That is worth experimenting with. Also, you can see on my circuit I use a different JFET, which I find sounds better in a guitar circuit.

How to JFETs work?

oversimplified answer:
Analogous to a valve, the voltage at the gate G controls the current between drain D and the source S.

The gate has a very high input impedance.

For this type of JFET, current is cut off when the gate is about 0.8V below the source, and as the gate voltage rises, more current flows. As current flows through R5 it makes the source voltage rise, while the gate is still tied to ground by R3, so it approaches this point where the gate is 0.8V below the source, and it reaches equilibrium, we have a steady state flow of current. The same current is flowing through R4 and R5.


Assume c4 does not exist:

Now a signal voltage gets superimposed, the gate voltage varies up and down, and the source voltage roughly tracks it. Current varies up and down.

Since that varying current is also passing through R4, which is larger than R5, the voltage swing at D is greater than at S, or at G - the input. Hence we have a voltage gain

Now add C4, assuming it is large. It does not affect the steady dc bias state, but it stops the source voltage varying with the signal. This means that the change in voltage between gate and source is much greater now, because the source cannot track it. The JFET now changes its conductance much more on each swing, there is a greater variation in current and much more gain.

is that any use?

John

[allmektig]

Thank you SO much for that answer. It was very helpfull.

Now, if I can just understand exactly what "impendance" is.
I think I got roughly an idea.
I'm not 100% sure about "bias" either. It's one of those terms I kinda roughly ounderstand but I'm not sure about the details.

Could you perhaps point me in the direction of any resource that explains those terms in such a way that even a porridge brain like me can understand it?

[ashcatlt]

Impedance is frequency dependent resistance. Resistors and transistors have the same resistance no matter what frequency is going through. Capacitors have more resistance at lower frequencies (infinite at DC), and inductors have more at higher frequencies.

Bias usually is a voltage that we add to the signal in order to get it to swing in a range we can use in our circuit. Coming into the circuit usually the guitar signal swings around 0V. Given a single-sided supply (like a battery, or DC wallwart), the active portion of the circuit can not supply anything less than 0V, so the "bottom half" of the signal must be cut off. We add a DC voltage to make the guitar signal wiggle around the middle of the power supply so we can amplify the whole thing. It is sometimes also called a "reference voltage" (Vref), "virtual ground", or even "audio ground".

[allmektig]

But how can a output have a resistance?
I kind of thougth of impendance as being a result of the interaction between resistors and capacitors in the output and input creating a filter and/or voltage divider.

Another newbie question:
Is it possible to saturate a transformer and achieve clipping that way?

[JohnH]

Jan 20, 2013 19:04:36 GMT -5 allmektig said:

But how can a output have a resistance?
I kind of thougth of impendance as being a result of the interaction between resistors and capacitors in the output and input creating a filter and/or voltage divider.

Another newbie question:
Is it possible to saturate a transformer and achieve clipping that way?

Pieces of active audio circuitry have input impedances and output impedanaces. Its usually fair enough to think of them as 'resistances' (not a good medel for a guitars pickup however)

The input to a piece of gear will draw some current from whatever is plugged into it. If it has low input impedance (Z) it will try to draw more than if it has a high input Z.

Can that output supply the required current without being brought too far down in level by having too much current sucked out of it. If it has a low enough output Z then yes. In audio preamp and mixer stages. we usually want to have higher input impedances being supplied from lower output impedances.

That booster circuit has a high input Z, basicly equal to the 1M resistor since the JFET has almost infinite input Z. It is ideal to plug a guitar into, since it draws very little current. Its output Z is about 6.8k, which is medium low - good to drive another similar amp stage, but no good for headphones etc, which have a relatively lower Z.

Another analogy: Two circuits, each giving an output swing of 12V. One is relatively high output Z, and is like a tiny 12V battery that drives a garage door opener - it can provide 12V but not much current. The other is low output Z, like a 12V car battery, which can supply a lot of current and still maintain 12V output.

I dont know about the transformer question. That was my cue to drop electrical engineering and change to structural design at college.

J

[ashcatlt]

Most of the active components - transistors, ICs, etc - can be considered purely resistive. Their reactive components are usually outside the scope of small signal, audio frequency circuitry.

A filter is a voltage divider. Consider the high-pass filter at the end of your circuit - C3 and R6. Picture C3 as though it were a variable resistor, only rather than turning it manually it automatically turns depending on the frequency it's fed. Higher frequencies turn the resistance way down. Lower frequencies turn it way up until you get to 0Hz - DC voltage - where it becomes an open circuit.

JohnH's suggestion that the out-Z is ~ 6.8K is only true for frequencies for which the cap is negligible. It's infinite at DC.

The connection between a source and a load (an output to an input) is also a voltage divider. Think of the out-Z as the "top" resistor, and the in-Z as the bottom and remember that both of these are "variable" by frequency.

Somebody on another forum once said "Everything useful is a voltage divider", and I'm still finding ways that this is true!

[allmektig]

Replaced the output cap, and now it's working.
Thank you SO much for your explenations.
It's really really helpfull, and I've learned a ton.

[candyflipper7]

Thanks for the post. I have been learning a lot from everyone. Great reading.

[allmektig]

And this is the result of this thread:



The gizmodrive.
I replaced r5 in the oricinal schematic with a 8k resistor, and added a 5k potmeter in series with c4.
Then I added a DPDT on/off/on switch with two selections of diodes, one germanium on one side and one germanium and one schottky on the other side, and a 100k audio taper volume pot in the end.


I don't have a drill, so I have't been able to drill the holes yet.