This one is a departure from the norm for me. I usually don’t dabble with solid state, but when I found this early Magnavox Transistor amp, I was intrigued. The transition from vacuum tubes to transistors didn’t happen overnight. Many of the first solid-state amplifiers were designed and built by engineers who had spent their career designing with tubes, so many of the early SS designs shared a lot of DNA with their hollow-state progenitors. This specimen is pretty spiffy – the design implements an entirely conventional single ended RIAA preamp, feeding a transistor driven transformer phase splitter for the big TO-3 push-pull output transistors. If you look closely, you can see that the factory assembly technician mixed up the installation of some of the capacitors. Both speaker coupling caps should have been 250 mfd and are installed in close proximity to a 100 mfd power supply filter capacitor. Unfortunately, the tech installed one of the 250 mfd caps in the power supply, and the smaller 100 mfd cap on one of the speaker outputs. Hard to say how much this impacted frequency response. Given the speakers that were originally fitted to the console, it probably wasn’t very noticeable. Listening to the amp on the bench was pretty un-impressive to start. Driven with ‘flat’ line level inputs, the RIAA equalization circuit enforced a frequency response curve that looked more like the bell curve from my 1st semester calculus class than anything else. Balance control implementation was the same wonky ganged feedback to ground adjustment that Magnavox was so enamored with on their tube amps.
My first act as rehabilitator in chief was to rip out the accursed RIAA preamp and tone and balance controls. In place of the balance potentiometer, I installed a single 200 ohm resistor between the feedback loopback and ground for each channel. Directly driving the phase-splitter/preamp gave a much-improved tonal result. Convinced I had a solid starting point, I went ahead and tore out the AC feed, all the power supply rectification and filter caps, along with the remaining circuit capacitors.
Building it back better, I spent 3 million dollars on a handful of new components and promptly lost most of them in my workshop after burning the invoice so I couldn’t return what I had left. Wait – that’s another story. Sorry…
What I really did do was install a new IEC power entry point, massively increased the B+ filter capacitance while retaining the original cool selenium rectifiers, now wired in parallel with modern diodes.
I also replaced all the circuit capacitors with top-shelf modern upgrades. The original carbon comp resistors checked out okay, but I did go ahead and replace the fixed feedback and balance resistors with trimpots to allow me to tune and balance the channels against each other. Finishing touches include a clear lexan baseplate, Bolivian Rosewood faceplate and uber-cheesy and ratty looking LED power on indicator lamp. Speaker connectors are still the stock spade connectors, but I did build up a pair of pigtails that convert to banana jacks for my testing. The proud new owner will receive those as a ‘special free gift’.
How does it sound, you ask? Coming from me, an ardent poo-pooer of all things solid state; pretty good. Keep in mind that a lot of the mojo folks associate with tubes really comes from the transformers. The soft inductive coupling is really where the magic is at, but let’s keep that to ourselves for right now before we stir up a controversy.
The amp delivers a clean output up to and beyond 5 watts into an 8 ohm load. Total harmonic content @ 1 kHz / 1 Watt is a smidge over 0.5%, raising only slightly to 0.58% @ 5 watts. As shown below, in the linear operating range even order harmonics dominate, and the frequency decay is a nice exponential. Soft clipping onset occurs at around 7 watts where the third order odd harmonic starts to dominate, driving THD up to 2.6%.
Frequency response is well matched to just about any full-range driver that you would be tempted to use. Minus 3dB points referenced to 1 kHz @ 5 watts are 18 Hz / 18kHz respectively. I contemplated adding a low pass filter in the feedback circuit to shore up the HF response, but selective feedback can tend to just homogenize and dull overall presentation. Remember, the goal here was not to build a zero distortion, infinite bandwidth lab quality signal amplifier. And, yes, in my younger days I did smuggle in speakers and a source to my employer’s lab to listen to music on a hi-zoot linear amplifier. It sounded absolutely hateful. That experience was one of a handful that eventually put me off solid state almost entirely.
After listening to this one as it is, I am convinced the current setup is the way to go. Sound is nice and warm, with a good tubey-vibe. Reminds me of a really clean SE EL84 but with more power.
There are couple of items you should keep in mind when considering this amp. One is the fact that the input sensitivity is a bit low. You are going to need a preamp that delivers close to professional line level outputs (+4 dBu) to get the full power output of 5 watts. The other is that there is no power switch – plug it in, and the amp is ON. I have been on a bit of an automation bent lately and had both the amp and its preamp connected to an Alexa enabled plug. When I wanted music, it was easy. I just had to say, “Alexa, turn on the Magnavox”, followed by “Alexa, play Johnny Cash”. If only everything in life could be this easy…
Foxtrot 9 Audio 