Highest headroom 9 volt gain stage???

[Bill Mountain]

As you may have heard me bemoan on this very forum I have an issue with two basses that have very different outputs.  My not so scientific experiments tell me my one bass is about 10x louder than the other (they measure somewhere around .2v and 2v).

I have been playing around with clean boosts in front of my pedal experiments as a way to even out the volume differences.  Me being me, I want the gain stage to be able to boost the high output bass if need be.  A simple opamp boost is fine for the weaker bass but the more powerful bass can clip most circuits with the tiniest bit of gain (even rail-to-rail opamps).  The most "clean" gain I've been able to achieve is with a CMOS inverter boost.  I know they are rail-to-rail but the gain is supposedly not very accurate so I don't even know how much I actually got.

Before I start planning 30v circuits (which is how much I need on a Micro Amp to keep my high output bass from clipping it), what circuits have you built that produced the most amount of clean gain on 9 volts?  I know MerlinB has the Glass Blower but that is a lot of parts for a simple input stage.

I'm willing to stick with the CMOS Inverter but that's also a lot of space for a little bit of clean gain.

Thoughts?

[Gurner]

I have been playing around with clean boosts in front of my pedal experiments as a way to even out the volume differences.  
Thoughts?

My thought is that if this boost is intended to be the first stage feeding into other pedals, then  it's pointless considering a 30V circuit.....each follow on pedal will be running at 9V, so it'll simply clip those follow on pedals.

I would be thinking about a signal of 6V peak to peak max, which means your boost should be no more than 3x for the loudest bass....nothing to stop you having a switchiable gain depending on which bass you're using though.

[Mark Hammer]

The headroom will be given by the voltage swing possible within the supply voltage.

Do consider that , even within the same "design", device-to-device difference can determine headroom.  So a simple op-amp stage might use a device that lets the signal swing - unchanged - within a half-volt of the rails, or perhaps no more than within 1.5V of the rails.

If complete avoidance of clipping is your goal, you probably want to go with a charge pump or other higher-than-9 voltage source.

[Bill Mountain]

Thanks for the tips guys.  I'm sort of a one pedal-at-a-time-guy.  I would basically be building a dirt box with a clean boost/input gain/sensitivy control followed by other normal dirt box controls.  If I ran it at 30 volts but the distortion stage uses diode clippers then any thing behind the pedal won't see too much output.  I've run a Micro Amp into a Ross Distortion and it's great for level matching my basses but I can also get some meaty crunch if I use the boost on the high output bass.  But with so much clipping in the first stage it also gets quite dark sounding.  I use a variable bench top power supply so I have had a 30v circuit like this on the bread board and the clarity was astounding.  I've been reading about different ways to get high voltage to my pedal and they all seemed troublesome.  I figured, if I can squeeze the most amount of clean gain out of the first stage then I could probably live with it at 9 volts.

Also,  I know it shouldn't make a difference but I have noticed (even though I didn't want to believe it) that the distortion stage did sound clearer at 30 volts.  I may try a rail-to-rail opamp that can handle huge input voltages for my 9 volt experiments (any ideas - maybe a CA3130?) but it might be worth just experimenting with a 30v power supply.

[Jordan A.]

It sounds like you have messed with this a bit already, but if the goal is more headroom for your high output bass you will probably "need a bigger boat".

If the 2 volt signal you quoted is an RMS measurement, that's about 2.8 volts peak, or 5.6 volts from top to bottom.  That's most of your 9 volt supply eaten up already without any gain added yet.  With a gain of 1.5 you get an 8.4 volt peak to peak signal, which is going to clip (or be right on the edge of clipping) pretty much any circuit with a 9 volt power supply.

I would think the high output bass needs a pad more than a boost.  5.6 volts PP I would think is plenty for most amps you might plug into, and probably too big for some. A typical 12AX7 input stage might be biased at around 1.5 volts DC or so, for instance, meaning it will start to distort (nicely) with about a 3 volt PP signal.

It sounds like you either need to pad down the high output bass or mess with bigger power supplies, maybe go bipolar?

If you have a scope and ideally a signal generator this will be a lot easier, as you can breadboard up some circuits and see *exactly* how much headroom you have and how much gain you can get away with.

cheers,

jordan

:edit: posted at the same time as Gurner and Mark, whoops

[R.G.]

The two comments are correct.

Unless you're using something like a transformer output to step up the output voltage, power converters to make more voltage, etc, a simple gain stage can't go outside the voltage of the power supply. If your power supply for the gain stage is 9V, then 9V is all you're going to get out of it. Notice that this is a statement about the output voltage available, not the gain.

So, the maximum headroom of a 9V powered gain circuit can't be more than 9V for a theoretically ideal amplifier. Some rail to rail output opamps approach this within millivolts. But it's still 9V.

From there, you pick your gain. The maximum gain you can use for an undistorted output is just the 9V pk-pk output divided by the maximum input signal you demand to be undistorted. If that's typical-guitar 100mVrms, that's 283mV (about) peak to peak and the max gain that can put that out undistorted is 31.8. If the max signal is 2V rms, that's 5.65V peak to peak, and the max gain for undistorted output is 1.59.  You have to adapt the gain to the input signal if you have a limit on the output voltage.

And the observation that there is a limit in the biggest input the following stage can accept is valid too. If you have 9V peak to peak of signal going into an amplifier with a 9V supply it can't have any gain larger than unity if it's to be undistorted, and even then it too has to have rail to rail outputs (and internal processing!) to be undistorted.

Headroom is one of those words that imply that there's some room for pushing things a little more. There isn't. There's a maximum output a maximum gain, and a maximum input for all circuits. And once you fix the power supply voltage to work from, the others rapidly get nailed down.

[Gus]

Besides clipping the gain stage there is the interaction of the passive bass with the cable and input resistance of the preamp.

search for my posts I often add a guitar or bass sim to the effect  sim screenshots.

here is a "pickup soak" idea with a guitar and cable sim.  Maybe try this on the hotter bass

[Bill Mountain]

True you can't go over 9 volts!  I was just looking for a circuit that would produce the most clean gain.

I've thought about pads but I haven't thought of the best way yet.  Buffer -> Pad or Slight boost Boost -> Pad?

I could have a simple opamp stage with a switch for -6dB, 0dB, or +15dB.  I would need a resistor divider after the stage to get less than unity for th -6dB.

I tried to build this but it was ultra noisy:

http://www.diystompboxes.com/smfforum/index.php?topic=97560.0

Gus,  I like the idea of conditioning the signal.  I need to read your post a couple of times to understand it though

[Gus]

The sim I posted is an emitter follower(EF).  Input is at the X to the left is the cable and guitar sim.

This circuit is bootstrapped for higher input resistance and to have lower value bias string for more stable bias.

The volume control is R5/R13 it is drawn as two resistors in the sim.  Make it one 10K volume control think of it like the volume control on a rangemaster but in the emitter leg of the circuit and not the collector leg like the rangemaster.

C4(cable capacitance) interacts with the input of the EF and the guitar or bass.

C1 adjusts the bottom bump if you want it.  .47uf is the big bump and 10uf is the blue flatter bottom curve.  The different curves are from different C1 values.

I posted this as an idea for you it has a high input resistance because of the bootstrapping and R5/R12(10k volume control as an adjustable pad).  So it is a buffer with a pad.

You could make a opamp buffer with a 10K to 100K volume control.  Fet input opamp like a TL071 wired as a buffer with an input resistance of 1meg with a volume control after the gain stage.

[Mark Hammer]

One often sees peak limiters or compressors built into bass amps, and you have indicated this stage is for bass.

Part of the reason why yu see such peak limiters on board bass amps is because there is a real risk of bass signals exceeding either the headroom specs of the amp, or the speaker cone launching into outer space (bass speakers usually having more forward/backward motion).

Are you aiming for "gain" or for output?  If one's aim is gain, with all dynamics preserved, then you need to be rather conservative in applying it to bass, because of the real possibility of those huge peaks exceeding headroom of the system.  If one's aim is output, however, then a judicious application of some peak limiting, to allow for max gain application to everything else except the peaks, is a smart idea.

[Bill Mountain]

The sim I posted is an emitter follower(EF).  Input is at the X to the left is the cable and guitar sim.

This circuit is bootstrapped for higher input resistance and to have lower value bias string for more stable bias.

The volume control is R5/R13 it is drawn as two resistors in the sim.  Make it one 10K volume control think of it like the volume control on a rangemaster but in the emitter leg of the circuit and not the collector leg like the rangemaster.

C4(cable capacitance) interacts with the input of the EF and the guitar or bass.

C1 adjusts the bottom bump if you want it.  .47uf is the big bump and 10uf is the blue flatter bottom curve.  The different curves are from different C1 values.

I posted this as an idea for you it has a high input resistance because of the bootstrapping and R5/R12(10k volume control as an adjustable pad).  So it is a buffer with a pad.

You could make a opamp buffer with a 10K to 100K volume control.  Fet input opamp like a TL071 wired as a buffer with an input resistance of 1meg with a volume control after the gain stage.

I really like this.  I enjoy mixing elements in my builds.  Transitor buffer -> opamp boost.  I don't need a lot of output.  I just want to handle a lot of input.  I was afraid that if I cut then boosted I would create noise issues.  I remember one of the opamp design books I have shows a circuit with different input gain ranges.  I need to dig that out and see what it was.

I guess throwing away excess gain at the input would be easier than designing a circuit and power supply to handle the excess gain.

[Bill Mountain]

One often sees peak limiters or compressors built into bass amps, and you have indicated this stage is for bass.

Part of the reason why yu see such peak limiters on board bass amps is because there is a real risk of bass signals exceeding either the headroom specs of the amp, or the speaker cone launching into outer space (bass speakers usually having more forward/backward motion).

Are you aiming for "gain" or for output?  If one's aim is gain, with all dynamics preserved, then you need to be rather conservative in applying it to bass, because of the real possibility of those huge peaks exceeding headroom of the system.  If one's aim is output, however, then a judicious application of some peak limiting, to allow for max gain application to everything else except the peaks, is a smart idea.

I was just reading about parallel compression earlier today.  I wonder how it would work if I split the signal and hard limited one side and put a mix control to bring back some missing dynamics at the input of the next stage.  The dirt stage is going to compress again anyway so it may not be as noticeable.

[Mark Hammer]

The nice thing about peak limiters is that they don't deprive you of much in the way of dynamics.  A compressor will, because it is both lowering the ceiling and raising the floor, but a peak limiter simply keeps the peaks in check without compromising the dynamics of the rest of it.

[Rob Strand]

I'm with Gurner, the signal chain is normally set such that 9V effects won't clip.  Any attempt go beyond this brings up endless clipping problems.

Why think boost, boost, boost?  The best solution is to attenuate the high output bass, after that everything is equal and there is are no follow-on problems with effects in the chain!    You won't have a noise problem because the signal is already hot (by a significant factor).

I've found that basses with excessively hot outputs often end-up clipping the on board preamps.  Some people then go for 18V rails on the preamp but it only cause clipping problems up the chain.

(Many bass amps have a high-gain/low-gain input because the bass output vary so much and the front end will clip.)

[head_spaz]

I was just looking for a circuit that would produce the most clean gain.

And that would be Merlin's GlassBlower Mk II, operating in class G.
No charge pump required to get almost 9VAC sinewave.
Quite brilliant!

[Gurner]

I was just looking for a circuit that would produce the most clean gain.

And that would be Merlin's GlassBlower Mk II, operating in class G.
No charge pump required to get almost 9VAC sinewave.
Quite brilliant!

In the accompanying diagram on that page, the output signal  looks to have a lot of zero crossover distortion (which figures as that relates to the 0.6V 'deadband' 'off' state of the transistors) and such crossover distortion really isn't pleasant to the ear. There's a degree of irony re using that method to get a larger AC 'clip free' signal, because you'd normally want more rail headroom thereby avoiding clipping (distortion) at the signal peaks, but with that circuit you get a larger AC signal out, but unfortunately introduce distortion in the zero cross region ....I agree it's innovative & clever, but if you want a larger distortion free signal with plenty of headroom, a charge pump (or transformer) is the way to go.

[merlinb]

In the accompanying diagram on that page, the output signal  looks to have a lot of zero crossover distortion

That is the signal appearing on the opamp power rails, it is NOT the output signal.
The audio output signal is distortion free.

[Gurner]

In the accompanying diagram on that page, the output signal  looks to have a lot of zero crossover distortion

That is the signal appearing on the opamp power rails, it is NOT the output signal.
The audio output signal is distortion free.

My Apologies...should have paid closer attention...my eyes are so used to 'left input'.... "right output" type flows & I missed that!

Cool circuit.