Building a Loftin White PX4 amplifier
The Loftin White design was published by Edward H Loftin and S Young White in the June/July 1930 editions of the US publication 'Radio News'. In fact Loftin and White, who had a laboratory in New York city, had described their circuit in a learned paper presented to the Institute of Radio Engineers in April 1930.
Although the prime motivation of the circuit design appears to have been economy of production, it is now noted for the very high quality of which it is capable. The main feature compared with other single-ended triode designs is that the cascaded valve stages are directly-coupled, that is there are no interstage capacitors or transformers. It is this that gives the design its unique fidelity.
In this article I will describe the amplifiers which I built based on the Loftin White circuit. Right now this is a summary which I will expand on later, as time permits.
I had read about single-ended triode amplifiers, and how good they are, but the industry seems to be split in two over this approach which is diametrically opposite to modern solid-state technology. I wondered if it was a load of bull, and having cut my teeth on a Rogers HG88 and then a Leak Stereo 20, I'd found out that valve amps are good and wanted to have a crack at a single-ended amp.
My interest in this design was sparked by a circuit published in the late-lamented Sound Practices in Spring 1994. The authors, Ciro Marzio and Cristiano Jelasi, based their amp on a 2A3 output valve. As published, I did not succeed in getting this amplifier working satisfactorily, the component values seemed out.
I decided to adapt the circuit to use a PX4 and set about collecting the necessary parts. My schoolboy physics was called upon to calculate the component values required, and I set about the design.
Following electrical design, the next stage was to layout the components, and having decided how to position them, fabricate the case work for both amplifiers. I chose to make a mahogany case with a MDF top panel and 90/10 brass top plate carrying most of the componets.
Having made both sets of casework, I built just one amplifier channel for testing and development.
Getting It Right
When first commissioned, the amplifier was, frankly, not very good. The main problem was hum, and this was traced to mechanical vibration from the massive Parmeko HT transformer and power supply choke.
Once the hum problem had been solved, we (I dragged my wife into this) set about a programme of optimising the component types for the best sound. This was carried out by substitution and subjective listening tests.
From the beginning, the tests were carried out using my Klipschorns (another classic design). It is amazing how loud 4 watts sounds through efficient speakers! As a matter of interest, I tried the Loftin-White driving a little Jim Rogers JR149 speaker (a derivative of the LS3/5a), but this was not a success, there just wasn't enough power...
Finishing the Job
With the first channel right, I built the second channel and installed both in their Brazilian mahogany cases, which I think look rather fetching on their Queen Ann legs.
The top view of the amp shows the Parmeko choke and paper block capacitors as well as the valve lineup of GZ37 rectifier, CV2492 driver, and of course the PX4 output. Also visible is the engraved nameplate giving instructions on how to sequence the low tension and high tension switches.
In the earlier bottom view you can make out the Parmeko HT transformer (ex Royal Navy), AudioNote output transformer, and the filament transformer. I hand-wound the latter, and the secondary windings are connected directly to the valve base.
The complete system has a very up-front presentation, with a wide dynamic range and incredible detail. I had not realised just how much detail had been preserved in the recordings. You can hear the bassonist taking breath, and a cellist knocking the music stand with her bow, for example.
Yet the sound is super-smooth, due I think to the extensive use of paper-in-oil capacitors. Each time a paper-in-oil cap was substitued for a plastic cap, the sound got smoother. I was concerned at one stage that it might be too smooth, but I needn't have worried, detail and attack are preserved and it is all very musical.
The Rest of the System
You may be wondering what the rest of the system is, so here goes:
- Thorens TD124
- SME 3012
- Ortofon SPUG
- Leak Troughline 2
- Croft Enigma pre-amp
The weak point is probably the pre-amp, good though it is, and I know application of the same principles (triode with no feedback) to the pre-amp would reap rewards, but I enjoy the system as it is and this will have to wait until I have more time!.