Mono amplifier. Two must be made for stereo connection. 

I managed to find almost all original parts on our market, which I did not expect at all.
The bigger problem was only with the mains transformer and the filter capacitors. The recommended toroidal transformer for 200W power into 4 ohms is 230V/2x 42V.
I managed to get a 2x36V, 300W toroidal transformer at GM Electronic. So the output power will be a little lower, but more than sufficient for home use. The voltage after rectification and smoothing at open circuit is approximately +51V and -51V.
I couldn't find capacitors with a capacity of 10000uF or 20000uF at 75V at all. So I decided to buy 4700uF/63V electrolytic capacitors and put 4 pieces in parallel on the positive branch and 4 on the negative branch. I have come up with another variant and that is to use 3 pieces of 6800uF/63V each time. 
I did not find mica capacitors on our market at all, so I used ceramic ones. It is said that some ceramic capacitors can "play".
However, I did not observe any problems. 
Carbon resistors are said to be better for HiFi due to noise, but I only learned this after buying the metallized ones.
I used normal electrolytic capacitors, but the LowESR type would probably be better. 

It's worse with zeners. The procedure is such that the 20V zener is connected via a resistor to a voltage source higher than 20V so that a current of 3.3mA flows through the diode.
I found that the Zener voltage changes with temperature and with the time connected to the power supply. I tried to keep the same conditions when measuring all the diodes and waited for the voltage on the diode to stabilize after about a minute.
I bought more Zener diodes, I measured their voltage 3 times for all of them and wrote them down carefully. After the measurement, I put together pairs of diodes to get the most consistent voltage for one channel.

Next, it was necessary to wrap L1 around R49.
For a long time I puzzled over what wire to choose and how best to wind it on the resistor.
Unwound lacquered wire from the output choke of the PC source with a diameter of about 1 mm worked for me.
I wound 11 turns on a screwdriver of a similar diameter to resistor R49. I then slid the finished coil over R49 and bent and soldered the ends to the resistor terminals.
The resistor and the coil are then soldered normally to the board.

It is possible to build an overexcitation detector to the amplifier, which by turning on the LED signals the limitation of the amplifier, when the output transistors are already fully opened and the signal is cut to a size just below the level of the supply voltage. At this moment, the acoustic signal is already degrading in the power amplifier.

(In Project Attachments)

Revival

First you need to try the source part. 
We leave fuses F2-F5 removed and check the connection of the rectifier and the polarity of the capacitors.
Any mistake could be fatal. If everything is in order, there should be a voltage of around 51V on the capacitors of both branches.
The exact magnitude of the voltage depends on the transformer used and the mains voltage.
After disconnection, the voltage on the capacitors will hold for a long time. We can discharge them through a resistor of about 100R 2W.
It may happen that the F1 fuse blows when you turn it on, even if everything is fine.
In my case, the charging current after switching on is so large that I had to install an NTC thermistor in series on the primary side of the transformer, which I took from an old PC power supply, where it performed the same function.
It would be better to use a "soft start" circuit, which consists of a resistor in series on the primary side of the transformer, which after a few seconds is bridged by a delay circuit using a relay to short circuit. However, I did not find a connection that I liked.

If the source is working properly, we can proceed to test the amplifier. 
Trimmer P1 is set to the extreme position with maximum resistance.
We install 100R 1/4W resistors in the fuse boxes for one channel. With the output and input disconnected, turn on the amplifier. 
We should measure a maximum of 2.5V on the 100R resistor, which corresponds to a current of 25mA. If everything is in order, we wait until the source discharges after switching off and install fuses instead of resistors.
We can start smaller. I had switched two wires to the power transistors on one channel, so the resistors started to smoke and smelled pretty bad. 
It really pays to double-check everything, but even so, a mistake can happen.

With the input disconnected we can also measure the DC offset at the output and it will show how carefully we matched the components. 
Now we need to set the quiescent current. 
Instead of fuse F2, we put an ammeter and turn trimmer P1 until we set a current of 100mA. As the amplifier heats up, the current changes.
We adjust the current value until it stabilizes, which may be after about 15 minutes. We repeat the same procedure for the second channel.
In my case, I measured 19mV on the left and 22mV on the right channel.
Now it is possible to proceed to sharp tests and connect speakers and a signal source.
I tested it by connecting the input of the amplifier directly to the output of the sound card.
It is necessary to turn down the volume and increase carefully, because the amplifier is powerful enough to destroy a smaller speaker.
However, I failed to fully wake up the amplifier with the Discman.