back to

To get more information contact me at:

Albarry reengering
by Andrea Ciuffoli - starting in the 2008


I don't like to repair amplifiers designed and build by other person so this is an exception.
Some months ago I have ear this amplifiers connected to Spendor SP1 loudspeakers, driver by a simple TDA1543 DAC with passive I/V and passive pre-ampl.
The emotion transmit by this "simple system" was very high so now I have decided to repair it with the purpose to see what there is inside.


Follows the complete schematic of this amplifier with the exclusion of the bad protection section created with opto-insulator and many other transistors.
The original protection circuit is not enough to protect loudspeakers and this final stage.
In any case the fault on this amplifier has been generated driving a pair of Quad ESL 63. This loudspeakers include a protection circuit that shunt the input when it consider the signal can create damage to the panels (see schematic)

The final stage is composed by the cheaper solution 2 pair of darlington TIP142 and TIP147 able to give 10A each.

There is a common single differential stage with BC556 followed by a second amplifier using a not common transistor ZTX653/ZTX753 1A 140MHz and hfe=225, better than normal BD139.

So the amplifier since to be the more simple possible and only the use of the ZTX can be considered a plus.

There is no capacitor in parallel to the BC546 used like bias regulator and temperature sensor.

I suggest to integrate a simple but good protection circuit like the one used in the Velleman K8060 kit or VM100 board.

The performance declared for the M1008 are 100w on 8ohm and these data will be confirmed by measurement and simulation.

The power supply is composed by a toroidal transformers 33+33Vac and a 8A diode bridge and 2 big capacitors should be 22000uF 63v.

On the pcb board there are some power supply bypass one 100uF 63v and a little mkp.

Many resistors on the signal path should be Holco and on output transistor are used  Enamelled Wirewound Power Resistors like the RWM Vishay

In the original project the bias current is set to 5mV on 0.47ohm so 10mA, if you increase it the amplifier will be destroyed.

To repair it in any case it is necessary change 2 x BC556 and 1 x ZTX753 and 2 x ZTX653 + all final.

So the final schematic should be the following.


This amplifier has been simulated with Winspice compatible with Spice3f4.

The original schematic without the 22pF on feedback circuit present a strange peak at 1MHz that it can create danger oscillation.


I have used in all the measurements the Clio system by Audiomatica with the Transit by M-Audio to use the USB port of my portable pc (instead of the PCI audio board).

Here output signal after driver stage load by 2 x 10K without final stage.

Here output signal on out driving an 8ohm resistive load