Anybody done anything interesting with a common base configuration?

[brett]

Hey the common base is part of the differential pair, it is used a lot 

Would I be right in saying that common base and differential pair configurations work very well up into the RF spectrum?  I've heard it mentioned before that you can get a response "hump" way way out there in radio land.  Some clever filtering/suppression might be called for.
cheers

[gez]

I can't find the link, but there's a regular contributor to EPE mag who has a number of projects on line and one is a 'speaker mike'.  If you plumb a speaker into a common base amp it acts as a Mike.  Although you can plumb it in the normal way, the CB connection is less noisy IIR.

I think intercom speakers worked both ways in some studios (in the old days), so perhaps this method was used?

[Sir H C]

Hey the common base is part of the differential pair, it is used a lot 

Would I be right in saying that common base and differential pair configurations work very well up into the RF spectrum?  I've heard it mentioned before that you can get a response "hump" way way out there in radio land.  Some clever filtering/suppression might be called for.
cheers

Common base is used into RF because you can get voltage gain without the cursed multiplication of the Miller capacitance from the base to the collector (gate to drain) that you have in common emitter gain stages.  Often the common base is driven by a common emitter stage, but the loading of the common base stage kills the common emitter stage's gain so you don't get the miller cap acting up there.

Also in RF, diff pairs with resistive loads are often used for gain stages without negative feedback.  These work pretty well, though again, often you put that common base stage in there to keep the frequency response up.

[Eb7+9]

another intresting use is the common grid config found in some Music Man amps - those that feature both power transistors and power tubes in the output stage, where the power tubes are acting as current buffers ... the cascoding increasing the drive impedance that the output TX sees while shileding the power transistors from the high B+ ... not sure if it's also a way to extend the power headroom but otherwise running the drive devices into a lower load makes them transfer signal with more linearity, and therefore less harmonic distortion ... I've seen some texts treat common-base as a voltage amplifier - this is quite misleading, could be why it's not used more ... I don't want to stir the pot too much but there's quite a number of situations that are better analysed/designed when looking at current as the primary signal variable ...

Pls breadboard or simulate the common base stage as the input stage of the guitar preamp\OD\fuzz. Also pls note that adding various capacitors in parallel with pickup will make absolutely different frequency response compared with the traditional pickup response. Surely you'll find it very interesting to play with that CB BJT stage (I did it a long ago and have been surprised a lot with differnt sounds).
The same effect can be achieved with the "OpAmp inverting stage without the input resistor". To be correct the pickup impedance plays a role of the "input resistor" here.

I missed this one ...

yep, there's a popular misconception that you can "load" a PU by driving it into too low an impedance - in truth that's only relative to the type of sensing circuit that's used ... in fact the reverse is true, you end up with more bandwidth when extracting a signal current instead of a signal voltage ... that's because the parasitic capacitance of the PU is now shorted out and plays no role - that's what makes zero ohm (current-mode) useful for high speed work ... to us though it also means that the characteristic "hump" of the voltage mode characteristics is gone in current-mode - at least in the sensing transfer ... this raises a different set of noise issues though

the Fuzz-Face is a modded current-mode input amplifier and it's low Z (somewhere around 1k to 3k depending on devices) moves the PU's hump way up ... you would think that such a low-Z input would kill all the top end of the guitar signal (according to loading theory) but it doesn't - that's because the circuit is amplifying a current and not voltage ...

the trans-resistance amplifier DDD mentions will give you a virtual short to drive the PU into - and depending on what you stick in the feedback path of the op-amp you end up with a voltage output that can have several types of spectral response - quite unexplored stuff, or at least infrequently mentioned ... when experimenting with mic pres this winter I went through the same process you describe and started out with "why not go common base with interstage transformers ?!" just simple power gain through impedance buffering and increased current ratios ...

eventually I ended up using an op-amp based, first time I went directly from a Ribbon motor (and got a lot of noise), second time I went through the Ribbon Mic transformer and got better S/N (which makes sense because of the introduced series inductances) ... in current mode the noise current of the source dumps current directly into the short - there's nothing to curtail it, unlike in voltage mode where the inductance of the PU and the series resistance both act to minimize what the (voltage) sensing circuit is seeing ... still, it works and gives an unfamilar sound ... I've also done something similar with single coil pickups which then allows them to operate in a zero-ohm environment - what I'm trying to do there is make them "noise-less" - increased voltage-noise immunity is another desirable feature of zero ohm current-mode - while leaving the overall response as close to I(x) by re-introduction of the hump in the volatge transfer ... which an be done passively by introducing a coil in series with the op-amp's output as been mentioned before ...

speaking of humps ... there's much to be said about where that hump lies exactly, especially in relationship to the mid-dip function of the tone stack in the amplifier ... if they are too far off, you can get a weird souding response - as many of us who have experimented with custom PU's over-wound by reduced-bandwidth techs as Brett aptly points out ... lining up the PU peak with the tone stack is very possibly something the early Fender people were aware of ... I'm guessing

now, throw in capacitive loading by the cable and that peak drops infrequency makes the guitar sound warmer, which is not the same effect when sticking a much larger valued cap and cutting highs significantly - by which point the hump is gone - but aside from the warmer feel it might even make the whole tone-stack/peak matching line up better (conjecture) ... maybe that's why Jimi had "that" sound on BOG

~jc

[gez]

http://www.zen22142.zen.co.uk/Circuits/Misc/Sp-mic.htm