Mark Levinson 333 problem


I have a broken Mark Levinson 333 that I want to repair for myself. It's working so far, but I have a problem on the Volage Gain board, which is why I can't figure it out at the moment.

The following were defective so far:

Q37,Q39 and the CR55 had passage.

After I replaced the parts, the Q36 gets around 100°C, but in the other functioning channel it only gets around 40°C

I had measured all the transistors so far and didn't find anything unusual. What could that be?


  • ML No336_Voltage Gain.pdf
    408.1 KB · Views: 94
This is very wrong.
Collector should be ca. 104 Volt. With 50mA, dissipation should be below 260mW, but now it is way out at 1.45Watt, so be carefull not to fry it.
R79 should have slightly more than 1Volt, but is now only 0.5Volt.
What is the operating output voltage ?
I'll prepare some extra questions.

Good news, that means that a lot of the amp is working and that your output stage is working perfectly.
Now, what are the voltages at:
Both sides of CR23/CR24, should be the same at both sides, ca.103Volt
The voltage over R64 and over R37, should be the same with ca 5 Volt
And from Q27, the base, emitter and collector, should be resp. 103.0, 103.6 and -45Volt.

Hello Hans,

i think I've found the problem. When testing the capacitors on the board, a via was apparently torn off.

But the condition of the board was already very bad. See image.


I'll test again this evening to see if it's any better.

Best regards


Could you do me a favor in return.
When simulating your voltage amp, Current source Q26 goes completely into saturation, which should be never the case with a current source.
On one hand R70 the 130R emitter resistor, with 0.7V on it from Q25, conducts 5.3mA, but only 4mA can go through the collector because the 2*71k5 resistors R109&R119 don't leave any room for more current, and 1.3mA has to be sourced by Q26's base.

So my question to you is, could you measure Q26's B,E and C voltages to see what's really going on.

The 4mA flowing through the collector seems o.k. but to get that in the Sim, I have to reduce R70 to 175R and R109&R119 to 2*68K.

Thx a lot.
Yes I meant q26.
You see, my sim is correct, there is only 0.1Volt between collector and emitter.
The poor transistor is completely saturated and can’t work as a current source.
I regard this as a construction error from ML.
They could just as well have left Q25 and Q26 with the same result.

The ML33 has exactly the same circuit diagram, only with a much lower supply voltage because it’s a bridged amp.
But here Q26 is not at all saturated and can work as a the current source that it was meant to be.