Class AB amp designs in pedal format

[suryabeep]

Are there commercial pedals out there that use a class AB amplifier to produce a clean boost instead of the traditional class A amp design?
Also, in order to make a class B amp with V+ and V-, could I theoretically use a voltage doubler (IC 7660S) and connect +18V to V+, 0V to V-, and 9V to "Ground"?

[rezzonics]

A power amplifier at 18V needs to provide enough current, a voltage doubler does not provide that current.

[suryabeep]

so how, if at all, could one build a class B amp in pedal format? assuming 9v daisy chain supply and no transformers?

[rezzonics]

EHX magnum 44 uses 24V input, but given the power, probably is a class D. I don't think a class AB on a small pedal enclosure could provide more than 5W or so.

Sent from my SM-J320FN using Tapatalk

[rezzonics]

Tubecake TC3 is 3W with 18V 500mA DC power input. Not sure if it's class AB.

[R.G.]

?  why ?

[suryabeep]

I'm sorry, I worded my question poorly. What I meant to ask is: is it possible to use a class AB topology in a pedal to make a clean booster pedal? just for the sake of variety?
My second question should have been: I've been reading up on the different types of amplifier topologies and Class B requires both positive and negative voltages. Can I use a voltage doubler circuit to provide the necessary voltages to make a class B clean boost?
Is any of this stuff even possible?

[Rob Strand]

possible to use a class AB topology in a pedal

Sure.  Opamps are generally class AB.

Class B requires both positive and negative voltages. Can I use a voltage doubler circuit to provide the necessary voltages to make a class B clean boost?

Technically you can have class B with a single rail.    The easiest example to understand is to have a pull-pull arrangement with a transformer.  The second is simple to have a vcc/2 bias idea with a class-B amp;  here the output stages will be class B but the drivers will be A or AB.

Anyway, you can use a power converter to generate new rails.

[printer2]

How much voltage you have has nothing to do with the class of amplifier. If you boost the voltage or not the opamp will not know if the power supply is done right.

The lowly 741 has a Class AB output. Something a little more hifi,

https://www.electrosmash.com/jrc4558-analysis

[PRR]

> Class B requires both positive and negative voltages.

(As Rob says) No.

If you want "clean", simple class A circuits are very clean.

As said, nearly all modern chip opamps are AB output. They do run very clean. They can run single supply.

If the designer has done his job right, you can not tell if it runs A or AB (clean B is tougher). If he has not done his job right, then the sound sucks, and you do not care if it sucks A or AB or B or even C ("gated").

If you want POWER, class A runs too hot. Class B (or AB) runs cooler. Class B distorts horribly. (There are pedals which use this effect, though without going to a full B power amp.) AB distorts less, and with very careful design as clean as any A stage. Doug Self's books show AB may potentially be cleaner than basic A, but the difference is generally undetectable, if the design is good.

If you want POWER, then you also have to consider the sizes of your power and output capacitors. For *stereo* amplifiers it is generally *cheaper* to use +/- power rails so you do not need output caps. There's other smaller advantages so you often see +/- power on mono amps. But it is NOT a requirement. The first Dynaco transistor amps, and a series of fine Yamaha bass amps, were single-supply AB. The real rush to dual supplies might be traced to Crown DC 300, which was also sold for lab experiments for high-power signals far below the audio band. (The DC300 could be rigged as a DC power supply, either polarity.)

[Steben]

Isnt the 386 based range of pedals the ultimate expression of class ab in a pedal?

[teemuk]

If you want "clean", simple class A circuits are very clean.

Not neccessarily. They just don't introduce crossover distortion.

That, of course, doesn't mean they couldn't introduce wealth of other forms of distortions, like generic clipping, or harmonic distortion and intermodulation distortion due to non-linear amplification at other regions than "crossing over".

Given that class-A efficiency is ridiculously poor I would expect that they actually introduce much more distortion than a class-AB amp powered by equal power supply.

Plus in a proper class-AB design crossover distortion is minimized to the point where you have to have precision instruments to actually even measure it. Take OpAmps for example; In signal processing chain from recording audio, to the point of mastering a recorded album, the signal has probably passed through hundreds of opAmps. Yet, people rarely even notice any crossover distortion in the finished media.

[teemuk]

Class B requires both positive and negative voltages.

In any operational active electronic circuit we always have positive and negative voltages. Let's consider a basic "unipolar" power supply, say, with +9VDC and common 0V supply rails:

+9V is positive in reference to 0V, 0V is negative in reference to +9V.

True "bipolar" rails (with positive and negative voltage in reference to 0V) are needed only if direct (or "DC") coupling is mandatory and the signal must have zero crossing point at 0V and swing around it in both negative and positive voltage. If direct coupling is not a mandatory requirement one can always  accomplish that with basic capacitive or transformer coupling that "isolates" DC voltage.

Basically +30VDC, +15VDC, and 0VDC -rail setup is equal to +15VDC, 0VDC, -15VDC -rail setup. The only difference is that signal zero crosses at +15V in the first and at 0V in the latter. Because of that DC offset, the latter must be AC coupled to stages that have a different DC offset level (such as GND / 0V / common).

Perhaps that also also answers your question about +18V, +9V and 0V -rail setup... Yes, it's otherwise equal to +9V, 0V, -9V circuit but the signal will swing around +9V and therefore such circuit needs to be capacitively coupled (or transformer coupled) to stages that happen to have different DC offset level, such as 0V.

[teemuk]

A power amplifier at 18V needs to provide enough current, a voltage doubler does not provide that current.

We don't really know that...

In general "voltage doubling" power supply can supply only half of the current of a supply without voltage doubling. Because a power supply can only provide a certain amount of power one loses in current in order to double the voltage (P=UI). If voltage is "pumped" even higher, the current output capability likewise decreases in same proportion.

So a decent approximation is that a voltage-doubled power supply can supply only half of the current a "non-doubled" supply could. Whether this is enough or not to run an amplifier depends on that amplifier. If it doesn't have to generate a lot of current to loads then doubling is fine. If it's short on current already without voltage doubling it most certainly will perform even worse in that regard with voltage doubling.

[R.G.]

An even bigger consideration for one-IC charge pumps is that the switched-capacitor mechanism has its own internal equivalent impedance. For the usual suspects, this is generally in the range of 50 to 100 ohms. That means that even if the bulk supply to the charge pump can supply hundreds of amps, the charge pump itself acts  like a 50-100 ohm resistor in series with the charge pump output. Switched capacitors are used to replace resistors in many active filter ICs.

[PRR]

> Class B requires both positive and negative voltages.

I may see your confusion.

A Class A amplifier can be built with one (or more) active devices.

A Class B audio amplifier MUST use TWO active devices.

This is device-count, not supply count. B audio is often done single-supply.

(There are non-audio amplifiers. A tuned radio amplifier can be worked class B with a single device. Half the wave is missing, but the tuned circuits round it out to restore the whole wave.)

[suryabeep]

so are there any pedals that use class B design to achieve (stompbox-type) distortion? Could someone share a schematic with me? I understood the relative voltages, but I'm still somewhat confused about the implementation.

[printer2]

I think you might have a misunderstanding on Class B amplification and are attributing something worthwhile in it. As an example, 18 Watt amps are known to sound fizzy but not in a good way. The reason is the output section move into Class B territory. Doubt you will find much in as far as pedals go but you can learn about it yourself easy enough. Get an opamp and put a 2N3904 and 2N3906 behind it but not in the feedback loop. The transistor pair will be in Class B.



http://www.electronics-tutorials.ws/amplifier/amp_6.html

[suryabeep]

I see, thanks.

[amptramp]

With a Class B amplifier, even a slight variation in bias will put you into either Class AB or a bad case of crossover distortion.  This is the situation where the bias is too far into cutoff for 180 degrees of the signal to be carried by each active device and near the zero point, you get a form of distortion where additional harmonics are added that sound harsh:



There are some crossover distortion pedals which give the sound of a cheap transistor radio operated in cold weather.  Brimstone Audio has a two-band crossover distortion pedal that allows you to set characteristics separately for high and low frequencies.  It works, if you want that sound.