- The power amp phase splitter (as opposed to the Channel One preamp phase splitter) is an unusual design for a guitar amp. It is composed of a 12AX7 drive stage capacitively coupled to a 12AX7 concertina splitter. This architecture was used widely in the 1950’s for guitar amps, including many Fender tweed classics.
- However, the Lectrolab circuit includes positive feedback from the cathode of the splitter to the cathode of the previous driver triode. This is NOT typical of classic guitar designs. When I first saw it, I thought I had made a mistake in tracing or documenting the circuit, but after making sure this was not the case I began to investigate precedents.
- Lectrolab S950 splitter circuit showing 180 degree phase signal relationships:
- Compounding my initial confusion was the fact that the 47K resistor feeding into the bottom half of the concertina splitter was unconnected at the opposite end. The resistor was located away from the rest of the circuit; it was near the power amplifier output. Based on the circuit design it is virtually certain that this resistor should connect to the output transformer secondary, serving as the source for the negative feedback loop. The value of this resistor, 47K, seems high, and would appear to result in an unusually low level of feedback to the driver cathode (note 1 below). Since this resistor is located on the power supply circuit card which appears to have been extensively modified, It is possible that this is not the original value. A 4.7K value is very consistent with what we would expect to see. Is it possible it was replaced in error? We will see.
- I found hi-fi circuits that use positive feedback in this section of the amp, such as the Fisher 400, Dynaco ST-35, Eico, and others. Here is one that has similarities, but a different, more refined, approach to balancing the splitter output: The potentiometer in this circuit balances the two outputs to achieve minimal IM distortion.
- I found one splitter circuit that is very close the Lectrolab. It is in a hi-fi amplifier built by Tadaatsu Atarashi (note 2 below) who published an article about it in a Japanese magazine called “High-Fidelity Audio: Radio Technology” in October 2003. I cannot find that article, nor any references to that magazine. According to the website, Mr. Atarashi is well-known among some hi-fi enthusiasts for his “new” reproductions of very old amplifiers. The same website says Mr. Atarashi used a Rogers amplifier as his model for this one.
- A post in a hi fi forum (note 3 below)states that Mr. Atarashi’s amp was modeled on a Rogers Junior monoblock, manufactured in the mid-fifties. Rogers was a English company in or near London. This is as far back as I can trace this particular circuit. I have not found a Rogers Junior schematic.
- The differences between the Lectroab and the Rogers/Atarashi splitter are threefold:
- Resistor and capacitor values – Although there are value differences, each component in the Atarashi amp splitter performs the same function as its counterpart in the Lectrolab.
- The addition of bypass capacitors in the Rogers/Atarashi circuit around the anode and cathode resistors of V1, and around the 3.9K NFB resistor. These are highlighted in yellow, above.
- The anode load bypass would reduce instability at high frequencies.
- The NFB bypass is to increase NFB at high frequencies, thus reducing high frequency gain in the output stage. Similar to turning down a presence control.
- The cathode bypass increases gain at all audible frequencies.
- The addition of the 220 ohm resistor to ground at the “tail” of the V1 cathode. This is also the tail of the V2 cathode. The entire circuit is sitting on top of this resistor. The Fisher 400 schematic above also uses this tail resistor. It serves as a voltage divider for both the positive and negative feedback to V1. In the Lectrolab, the singe V1 cathode resistor serves this same purpose, at least partially because it is unbypassed by a capacitor which, if there, would bypass much or most of the feedback to ground.
1. See Eli Duttman’s post at http://www.diyaudio.com/forums/tubes-valves/167688-concertina-splitter-question-need-help.html