The Audio Pages

Background (Extract from The Book)

In May 1998, Dr.John Nicholson, Programme Director of the University of East Anglia's Science Starter Programme, gave the author go-ahead to design and teach a course on high-end hi-fi.  It was agreed that it would be a five-week course, starting with a general introduction to hi-fi and electronics, and progressing through semiconductor amplifiers, valve amplifiers and loudspeaker design.  Blending theory and practical work throughout, the overall design of the course soon took shape, but it lacked a real conclusion.

A chance reading of an advert in Everyday Practical Electronics led to the eventual purchase of a 5J/180E transmitter valve, which it was decided would form the centrepiece of week five of the course, entitled "The super-fi amplifier: Nick's flagship workshop".  Course leaflets were prepared, and in September 1998 the first presentation of DIY Super-fi, as it became known, was given in the Elizabeth Fry building at the University of East Anglia.  The first four workshops went very smoothly, week three being especially well received, during which the students built working, single-ended amplifiers based on the PCL86 valve. Week five, however, was to cause serious problems.

Soon after the purchase of the 5J/180E it emerged that, although electrically almost indestructible, mechanically it was completely otherwise.  Transmitter valves have very brittle filaments, and with only one valve in the planned circuit, the slightest impact could have caused instant, catastrophic, total failure of the amplifier, and the sudden and premature end of the fifth workshop!  Late on the night before the fifth session, the decision was taken to hurriedly substitute a multiply parallel push-pull class B design, based on the already used PCL86 which thankfully had been purchased in a very large quantity.

Work proceeded through the night, and by morning a working push-pull amplifier was ready, complete with variable bias and a separate, variable HT power supply.  This was successfully demonstrated, first at an HT of 340 volts with the grid bias varied from near zero volts to around -18v, thus demonstrating class A, class AB and class B.  Although interesting, this was not what had been advertised, so, in order to at least attempt to fulfill our stated aim of producing a very powerful valve amplifier in week 5, the HT was slowly raised while maintaining the PCL86s cut off, this being achieved by steadily increasing the grid bias negative value.  Spectacular failure occurred at an HT of 1730v - this has to be compared with the valve's rated anode voltage of 230V !  The peak output of this circuit was calculated to be in excess of 1000 watts, albeit briefly !

For the next presentation of DIY Super-fi, it was decided to use the 5J/180E, but in parallel with another identical valve.  This would have two advantages: firstly, if one valve failed it would not be catastrophic; secondly, the potential output power would be colossal.  The second 5J/180E was duly ordered.

When it arrived, there was a major surprise: whereas the first valve had arrived in a wooden case the size of a tea-chest, with an internal box containing the 5J/180E that was isolated from the outer casing to stop vibration, the new valve had come in a small cardboard box, with minimal packing.  Horrified, the author hastily checked the filament, which to the general astonishment of the course technicians was intact !  Obviously, these valves were more mechanically robust than we had thought.

Encouraged by this discovery, a test circuit for the first 5J/180E was built in July 1999, and the valve was proven as an audio amplifier while running at a reduced EHT of 1500v, and without forced air-cooling.  As the filament alone was consuming 300 watts, the tests had to be conducted quickly, as the valve quickly heated up to above 100 degrees Celsius. The next step was to be a full power run, this time with a powerful industrial air-blast unit.

While the cooling unit was under construction, the decision was taken to raise the anode voltage of the 5J/180E, the idea being that this would further increase our chances of success, and increase the power to the point where the international hi-fi press would take an interest (see Appendix 1).  Nothing succeeds like excess!  The author readily admits that over-engineering is not a practice he is used to!

Facilities for testing Ko'lok were kindly provided by Norwich Community Workshop, as UEA's Science Starter Programme funding ran out before Ko'lok could be demonstrated.

Details of the three circuits used in DIY Super-fi, that is the multiply parallel push-pull PCL86 circuit, the single-valve 5J/180E circuit, and the final, over-run two-valve 5J/180E circuit which formed Ko'lok, are given in the next three sections.  The majority of the data on the output valve is contained within section four.  All the support modules (LT PSUs, HT PSUs, air-blast units, control units, transformer units, driver amplifiers etc.) are also described in later sections.

While the responsibility for course and hardware design has rested solely with the author, the author owes a tremendous debt to the team at and beyond the University of East Anglia who made this project possible.  They are listed below: