Leak Logo Leak TL/25A
Power Amplifer

Information about this amplifier is very limited as it does not appear to have been advertised to the extent that the TL/12 was. The units photographed here date from about 1953.

Physically the TL/25A is on the same chassis as the TL/12 (the chassis is even punched for TL/12 components), however the output transformer is vastly larger in size - looking more like a 50 Watt transformer. Another physical difference is the addition of a sixth valve on the chassis between the output valves and mains transformer. It is likely that these amplifiers were built to order, rather than being available ex-stock, making them very rare.

The circuit is similar to the TL/12, but differs in the following ways:

The TL/25A was a revision of the original TL/25 from the late 1940's which used cold-cathode tubes to derive the screen supply instead of the KT61 used here (information supplied by Patrick Cheffins). Like the TL/12, all the components used were of first rate quality, with large oversized transformers, and oil-filled paper capacitors in the power supply (no electrolytic capacitors.)

The photos included here, are of amplifiers still in theatre service in Australia some 40 years after installation. The sound system in the theatre concerned was installed for the stage show "Kismet" in 1953, and featured six Leak TL/25A amplifiers, one Leak TL/12, and seven Leak Varislope III pre-amplifiers.


Measured Performance (Test Date: July 25 1988)

1. Distortion at Rated Power Output (25 Watts)

 Frequency  Output Power  Harmonic Distortion
 20 Hz  12.8 W (max)  N/A
 30 Hz  25 W  1.8 %
 100 Hz  25 W  0.14 %
 1 kHz  25 W  0.1 %
 10 kHz  25 W  0.5 %

2. Distortion at Maximum Power

 Frequency  Maximum Power  Harmonic Distortion
 20 Hz  12.8 W 2 %
 30 Hz  25 W  1.8 %
 100 Hz  32 W  0.23 %
 1 kHz  34 W  0.22 %
 10 kHz  32 W  0.7 %

3. Frequency Response at 1 Watt, relative to 1kHz

 Frequency  Output (re 1 kHz)
 3.8 Hz  -3 dB
 5 Hz  +2 dB
 10 Hz  0 dB
 1 kHz  0 dB
 10 kHz  -0.5 dB
 20 kHz  -1.3 dB
 34 kHz  -3dB
 62 kHz  -6 dB

4. Square Wave Response:

Rise Time: 10 micro-seconds
Overshoot: Aprox 5 % ringing, superimposed on an overall underdamped output (ie: no actual overshoot).

5. Feedback:

26 dB at 1kHz.
Decreasing at approx 6dB per octave above 4kHz.
Decreasing at approx 6dB per octave below 30 Hz.

6. Special Notes:

a) With 440 V on the anodes instead of 480V as above, the maximum power was 25 W.
b) With 440 V on the anodes, and EL34 valves instead of KT66's, maximum power is reduced to 22W.
c) With KT88's instead of KT66's, and with a 5AR4 rectifier intead of the 5U4, the anode volts were 480 V, and the maximum power was increased to 37 Watts, with 0.2% THD at this power.

7. Test Conditions:

AC Mains supply: 233V, with voltage selector set to 230 V.
Measured Anode Volts: 480 V
8 Ohms load, and strapped for 8 Ohms.
2 x new KT66 valves, and 5AR4 rectifier.


Designers and Manufacturers of Specialised Electronic Instruments for the Communications Industries