1. Home
  2. Electronics Design
  3. Phase Inverter for Guitar Amp

Phase Inverter for Guitar Amp

formatting link

I'm trying to understand how the phase inverter works in the amp above(3rd page before the power amp).

I seee it is basically a normal common emitter.

For DC, we simply have the 470 and 15k resistor in series giving the bias for the two tubes which happen to run off the same bias but could have separate resistors. I assume the reason this was done was for accuracy.

Things I don't get or I'm unsure of.

  1. The 470 resistor. The 330K and 1M resistors seem like gate drain resistors so why not just take them directly to ground? In this case the gate of B effects the gate of A and vice versa.

  1. The FB signal. On the secondary of the output transformer there is no ground so the FB seems to be floating and I can't see how it would do anything but potentially add noise. It also only looks to work when the 4Ohm speaker connections are used.

  2. The strange network of resistors. The common point between the 4 resistors(looks like a +) seems to complicated things an extra order of magnitude.

  1. 120 cap on the anode of A. What the heck is it for? Won't it just allow high frequency noise on the output? Also 100k vs 82k.

I was thinking about improving this a bit but since I know so little about it. I was thinking about adding an emitter capacitor for a more stable bias voltage but it doesn't look like it will work because of how the originals ignal and feedback signal are "injected" into the circuit.

I'm guessing that most of the "strange stuff" is to help with the problem of crossover distortion in the power amp?

Anyone mind clarifying whats going on here? I thought the whole idea of the phase inverter was to create an inverted signal for the pull pull power amp section? Inaccuracies in this could easily create unwanted distortion?

Reply to
George Jefferson
Loading thread data ...

"George Jefferson"

** An essential feature of the circuit is that the cathodes of V5A & V5B are linked - that is how signal gets from one to the other half of the tube !!!
** They provide a controlled negative bias to each grid - ie the 1.4 volts dropped across the 470 ohms.
** Drawing error - in fact the bottom end of the secondary goes to chassis.
** Don't be silly.
** Fraid that is the heart of the circuit.

** High frequency stability of the output stage under actual conditions of use.
** Compensates for the - error in output level matching inherent in the circuit.
** Leave it alone - pal. .

** Go on, try it.

** Guess again.

** Doubt you will ever get it.

.... Phil

Reply to
Phil Allison

Looks more like it's a "cathode coupled" phase splitter/differential amplifier with a resistor acting as a pseudo constant current source.

The 470 ohm resistor does provide for correct bias, but there couldn't be two separate resistors or else the circuit wouldn't work properly.

The problem with taking those two resistors directly to ground is that because of the 15k "long tail" resistor the cathodes of the two sections are at positive 47 volts with respect to ground. If the two resistors were grounded it would put the grids at -47 volts wrt the cathodes, putting the tubes well into cutoff. Connecting them to the junction of the 470 ohm resistor and the 15k resistor allows the grids to bias up a few volts negative wrt the cathode, putting the tubes in their linear region of operation.

There'll always be a signal on that leg of the transformer secondary, regardless if the 4 ohm speaker is connected or not. The FB signal is referenced to ground through the 100 ohm resistor in the feedback loop, so there shouldn't be a problem.

Again, it does have to be connected up that way for an AC coupled differential amplifier using this topology to work. :)

f the bottom triode were at AC ground, then the resistors would definitely have to be unequal since the only signal the bottom triode would then be getting would be through the cathode of the top triode. In that situation, if the resistors were equal then the currents through each tube would be equal, and a current rise in the top tube caused by the input signal would cause an equal current drop in the bottom tube, and there wouldn't be any signal to drive the second tube! The cap is probably to equalize the low frequency response of the two outputs given the differing output impedances.

In this case however, where there's also a feedback signal coming in to the bottom tube, I'm not sure that the two resistors have to be unequal for the circuit to work, since now it's acting more like a true differential amplifier. Maybe someone with more tube design experience can comment.

Reply to
Bitrex

^^ bottom tridoe's grid

Reply to
Bitrex

Thanks, it makes a lot more sense seeing it as a difference amplifier and the common node is a virtual ground for it. The only thing that still doesn't make much sense is the FB into the volage divider. If the 15k resistor was grounded then it would make sense.

I guess though that cap will look like a short for audio frequencies and if the FB has no DC then the 15k resisitor is effectively grounded.

Seems like some hocus pocus going on though ;/ How can you be sure the feedback would never drift due to changes in the transformer and/or resistors? Also for low frequencies it would have an effect on the overall bias? I guess not much to really matter though?

What I don't understand is why not just use something as simple as

formatting link

or

formatting link

These don't have any delay which can ruin a amp from what I hear(can't respond fast enough for the transients).

Reply to
George Jefferson

"Bitrex"

** Nonsense.

The drawing has an error.

.... Phil

Reply to
Phil Allison

I'm not entirely sure why they used that arrangement for the feedback, instead of putting the 15k resistor to ground and taking the feedback directly to the grid of the bottom triode and making a straight differential amplifier. It would have more gain than the cathode coupled arrangement, and equal-valued anode resistors could be used.

It seems the designers are taking the feedback from the output stage to the phase splitter instead of the input stage because they want to keep the output operating as linearly as possible, and yet not interfere with the "good" distortion for guitar created by overdriving the input stage. In that case the first type of phase splitter is out, because it has no gain. Because of the impedance transformation by the transformer the power output stage will probably only have a voltage gain of about 2, which isn't much for the feedback to work with!

In addition that design also has the disadvantage that the output impedance from the cathode and anode are wildly different.

The circuits on the audioXpress site may be superior to both designs, but involve pentodes and cascodes and stuff that the amp builders aren't going to get into for reasons of cost and availability of parts. One common dual triode for the phase splitter is all the bean counters have allowed. :(

Reply to
Bitrex

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.