Biasing for unity gain inversion stage (single transistor)

[PenPen]

I'm still a ways to go from posting the final result of a design I've been messing with, but I have a potential problem I think I may run into and I expect I will need a simple inverter to resolve it. Basically, I'm making a booster design that uses a mosfet at the input, and splitting the signal from there using both the drain and source as a signal line. The source stage runs into a rangemaster style treble boost. The drain will go into something darker, I haven't yet completed that portion. Each output will then go into opposing legs of a pot for blending. The problem that I forsee is that the signal from the source is not inverted until it hits the treble boost side. The other side is inverted from the mosfet but then flipped back to normal in the planned jfet boost stage. When these two are then mixed back together, I expect the signals will then be out of phase and will cancel each other out to some degree. Which I would like to leave as a feature, actually, but make it switchable between out of phase and in phase. So, I planned to implement a unity gain inverter that can be bypassed by switch. Problem is, the math involved with biasing doesn't yet make sense to me and usually ends up with me testing different resistor values until I find the right combo, and consulting other schems for reference. I get some of it but not too well. Making something unity gain is over my head completely. Can anyone explain how to achieve unity from the collector of a single bipolar?

[PenPen]

I'll just bump this one time, it dropped to the second page without any responses. Does anyone have any ideas? Is this just a plain dumb idea? Even a response to tell me I suck would be fine.

[petemoore]

  Yupp I got one.
Bigger emitter resistor without bypass cap and larger collector resistor, bound to reduce current or voltage.
  Going larger on the emitter would reduce the available voltage swing...as would moving the bias at which the collector is sitting...
  ...k then what...an opamps inverting input used as buffer...I gotta go read data sheets...
  Smaller base bias resistors [from base to V+ and or V-, or...] would load down the input, alot of the time reduceing high end content first.
 
 

[davebungo]

I think you may already be staring at the answer(s).  Think about it... you mentioned a MOSFET input stage which will be used to generate an inverted and non-inverted signal?  If the signals are to be the same (only inverted), you must make the source and drain resistors the same value (and use no source bypass cap).  The same current flows through both resistors and so the same voltage albeit inverted is generated across each.  So, why not simply switch which one you feed into your next stage e.g. put a switch in to allow the treble booster to be driven off the either the source or the drain i.e. non-inverted/inverted.  Bear in mind though that the "phase splitter" if I may call it that, will provide no gain. so if you want some boost you will have to do it elsewhere.  Another possible problem may be the load presented to the MOSFET stage by the subsequent  stages - you may have to buffer them or accept some signal loss - shouldn't be too much I would have thought if you use source and drain resistors in the order of 10K, and your input impedance of subsequent stages is 10X this.

[PenPen]

  Yupp I got one.
Bigger emitter resistor without bypass cap and larger collector resistor, bound to reduce current or voltage.
  Going larger on the emitter would reduce the available voltage swing...as would moving the bias at which the collector is sitting...
  ...k then what...an opamps inverting input used as buffer...I gotta go read data sheets...
  Smaller base bias resistors [from base to V+ and or V-, or...] would load down the input, alot of the time reduceing high end content first.
 
 

Right, I figured it would be larger resistors, but what values am I looking for at my collector and emitter? I remember reading a post or article by Jack Orman that detailed the formula for choosing resistor values with a target gain, but I can't find it now.

I want to avoid using an opamp. After the split signal is blended back together I have a transistor buffer stage planned.

Now the more I think about it, I'm not certain its going to be out of phase. I still want an inverter, as an optional 'out of phase' switch. But I'm thinking about how jfets operate, which is going to be the 'dark' stage from the drain of the mosfet. The mosfet drain line will be inverted, I think, could have that backwards. Then it goes to a J201 fet. These are normally on, and positive (?) voltage causes the gate to close, thus creating more voltage on the drain of it, so therefore it does not invert the signal. So the signal will be inverted from the mosfet drain stage (again, I think it is. I could have that backwards). The mosfet source stage is non-inverted, but gets inverted by the bipolar treble boost stage, so it is then inverted when it gets mixed back in. I think this is all correct, I could be horribly wrong though. In any case, I still need a single transistor inverter to flip one of the signals for the out of phase feature. I understand I can just bias a bipolar for unity gain and take the signal from the collector, and it will be inverted in respect to the signal at the base.

Am I thinking this through correctly? Anyone have the formula for choosing bias resistors to achieve a target gain?

[PenPen]

I think you may already be staring at the answer(s).  Think about it... you mentioned a MOSFET input stage which will be used to generate an inverted and non-inverted signal?  If the signals are to be the same (only inverted), you must make the source and drain resistors the same value (and use no source bypass cap).  The same current flows through both resistors and so the same voltage albeit inverted is generated across each.  So, why not simply switch which one you feed into your next stage e.g. put a switch in to allow the treble booster to be driven off the either the source or the drain i.e. non-inverted/inverted.  Bear in mind though that the "phase splitter" if I may call it that, will provide no gain. so if you want some boost you will have to do it elsewhere.  Another possible problem may be the load presented to the MOSFET stage by the subsequent  stages - you may have to buffer them or accept some signal loss - shouldn't be too much I would have thought if you use source and drain resistors in the order of 10K, and your input impedance of subsequent stages is 10X this.

I know I should be posting a quick schem to illustrate the idea. I had considered the set up you mention here, but I kind of wanted to have both a treble boost and a darker jfet boost blended together, so I'd be using both lines from the mosfet simultaneously. I thought about flipping the drain and source of the mosfet by switch, but since one stage is non-inverting (taking the output signal from the drain of a j201 FET is non-inverting, right?) and the other is inverting, I think it would be in-phase no matter which signal was presented to both stages. Achieving the out of phase signal by switch is what I'm after, a basic bipolar with unity gain so there isn't a big difference between in/out of phase.

Man I really bit of more than I can chew with this one.

[davebungo]

I know I should be posting a quick schem to illustrate the idea. I had considered the set up you mention here, but I kind of wanted to have both a treble boost and a darker jfet boost blended together, so I'd be using both lines from the mosfet simultaneously.

That's as I understood it

...taking the output signal from the drain of a j201 FET is non-inverting, right?) and the other is inverting,

It's the other way around.

What I'm suggesting is not that you flip the drain and the source but that you feed one stage, say the jfet boost, permanently off the drain, and feed the treble boost off either the source or the drain - that way the resulting end mix will contain either in-phase or out-of-phase signals depending on the switch position.  If I'm still missing the point then may be a schematic would be good.

[PenPen]

...taking the output signal from the drain of a j201 FET is non-inverting, right?) and the other is inverting,

It's the other way around.

What I'm suggesting is not that you flip the drain and the source but that you feed one stage, say the jfet boost, permanently off the drain, and feed the treble boost off either the source or the drain - that way the resulting end mix will contain either in-phase or out-of-phase signals depending on the switch position.  If I'm still missing the point then may be a schematic would be good.

Ah, ok I see what you are saying. I like this idea. I'm going to have to do some experimenting to see if this is going to load down the mosfet too much, but I like the idea and it sounds like one way to go. I can make the output buffer gain a smallish gain stage to account for signal loss at the mosfet.

So, I was wrong in my understanding, a J201 FET has the signal inverted on its drain, in respect to the gate signal? I thought positive voltage on the gate causes it to turn off, thus increasing the voltage on the drain. Do the J201's turn off from the negative voltages then?