The SV series of valves produced by Svetlana are some of the very last valves designed, and are specifically aimed for use in the audio field. Designed in the mid 1990s they were only produced for a short, 4 or 5 year period, and as such are becoming scarcer as time goes by.
This SV572 amplifier is single ended with a 9 pin Aikido driver circuit. The output stage is configured slightly compromised so that the valve is swappable between SV572 and SV811. The differences between the 2 valves have been documented in some obscure corners of the interweb but curiosity dictated another look.
The final version involved the introduction of: a solid state rectifier, a new choke-input power supply, a new bias scheme, dc heating, an HT delay and double voltmeter to assist in biasing. The driver circuit is still the original aikido circuit using ecc82 and ecc83, designed by John Broskie.
Basic circuits
The heaters were powered by Rod Coleman regulators, supplied by 12v 12A smps power supplies.
The chassis
- A Hammomd 17 x 12 x 3
- Small holes less than 5mm have been centre punched and drilled directly with a block of wood beneath
- Larger holes, 5-12mm have been drilled at 5mm and then enlarged with a stepper drill
- Largest holes are made with a hole punch, i.e valve socket holes.
- Square holes have been made with a 6mm drill and jig saw.
- Immaculate paint job once again by Dave Tuplin of T&G Motors in Alford, Lincs
Fixed bias
The fixed bias scheme was originally run from the HT transformer. With the introduction of the delayed start on the HT transformer this method was no longer possible, so a new circuit was constructed. In the circuit shown the bias is adjustable from approximately -65v to -120v(mains at 250v), which is entirely adequate for swapping between SV811 and SV572. Locating small 72v transformers proved difficult so 2 transformers with 2 x 0-18v secondaries were used, with all secondaries wired in series. The circuit board was constructed so that the screw adjusters were accessible from the top of the chassis, eliminating the need to get inside to re-bias. Monitoring the bias is achieved via the voltmeters shown on the LCD.
DC Heating
Previous versions had presented a degree of hum, especially when partnered with efficient loudspeakers so it seemed like a good opportunity to introduce some dc heating. Having made the decision it was decided to go all the way, and this resulted in the inclusion of Rod Coleman’s regulator boards. These are available in kit form from Rod directly(search for Lyrima Ltd on the internet for contact details). The regulator boards still require a power supply, which with these valves is not trivial, no gain without pain. The heatsinks are from Conrads. The transformer used was a frame type which happened to be in the bit box. Without one to hand this may well have ended up as a toroid, as research has shown they are slightly less bulky. The power supply for the regulators has since been changed to smps.
HT delay and bias measurement
Originally this amp was going to have a basic delay circuit based on the 555 timer chip. Thinking about how to implement a voltmeter at the same time resulted in research of the AVR 8 bit microcontrollers. The voltmeter is required to bias the sv572 output valves. The AVR chip, ATTINY461, can do both jobs without breaking into a sweat. For more detail on the AVR chip see the LCD paragraph below. This is an indulgence, having not played with embedded chips since the 1970s this became very tempting. Nah, it became an obsession. It wasn’t an entirely useless diversion though, as an IR remote control passive pre-amp is on the workbench using the same technology. Anybody using this method would probably use a 5v relay and remove the need for the 12v regulator. The 12v relay was the only one in the box at the time. For practical reasons the relay switches the mains to the HT transformer, rather than the HT itself.
LCD and micro controller
Acknowledging the fact that the LCD is an indulgence and a bit of an overkill, it still fulfills the original requirement of providing a 30 second delay for the HT and provides 2 voltmeters to monitor the bias current on the output valves. The AVR microchip contains an ADC which is capable of measuring up to 5 volts directly, without resorting to divider circuits. As a consequence, by placing the correct value of sense resistor on the output valve cathode then a measurable voltage can be produced. Considering the LCDs are available for less than £2 and the AVR micro chips cost £1.50, then the only real indulgence is the time spent writing the code.
If anybody is interested in trying this method then the AVR code is available here. If unfamiliar with microcontrollers generally then be aware that some hardware will be needed to get the program into the chip after it has been compiled. A free compiler is available from the Atmel site here, and cheap programmers(under£10) can be found by searching for ‘attiny programmer’ on any search engine.
Summary
Probably not the most elegant layout, but it is symmetrical. The output transformers are James 6123 and the power transformer are bespoke windings from BLS and the chokes are Hammond, supplied by Bluebell.
