Collector of Q1316 at -3.2V is much lower than the marked value of +1.6 which suggests the oscillator is being driven harder.
However, there is a diode to ground connected there (CR1316) which should prevent any voltage lower than -0.7V or so.
I can't see that it would cause your present problem but do check that diode.

Can you get the 20MΩ resistor?

 

Oh i should have been drunk last night cause i did not take the good measurement.
On the base of Q1306, i measure 0.76v. it's a little bigger than on the shematic but it sounds good.
This problem drives me crazy.
The 24.5M resistor is not changeable. it's a kind of plastic plate with 3 resistors (R1303A, B, C).
I tryed to add 180K but it does not change at all.
Do you think it can be a material problem ? Everything else seem to work, exept the vertical adjustements which is totally
untuned.
Thanks

 

It's higher than it should be - still points to R1303A. You need to check R1303A. Get that 20MΩ resistor.

 

Forget the last one (-3.2v is on the base), there is a -0.7v on collector of q1316, so, there is a problem.

 

Yes, -3.2V on the base is wrong too.

 

I have a 20M resistor waiting for your measurement. I'll do this tomorow night..

 

I may lose sleep........

 

Finaly,
i made measurement of the resistor and i get 14.5M... I don't understand cause i wasn't able to measure it.
So, i add a 10M just before the base of the transistor and then, the bulbs are flashing and with minimal brightness, i can see clearly the lines.
After few measurement, something near Q1316 has burned (maybe the Q1318 again) so, i'm back to the starting point...
Maybe i've touched two componment with my multimeter stick. I'm waiting for a new Q1318 (2n3055).
Too bad for me

 

Just to be clear, R1303A in parallel with 20MΩ reads 14.5MΩ? If that's right then R1303A is 52.7MΩ (should be 24.5MΩ) and that is certainly your fault.
Where did you add the 10MΩ

 

With 22M in parallel, i read 9.2M.
So, my R13.3A = 14.5M.
After this, i unsoldered the parallel R and read back R1303A = 14.5M (maybe i did it wrong before).
So, i added a 10 M in serial in order to match the 24.5M.
I think i'll wait for my new transistor before doing wrong again

 

With 22M in parallel, i read 9.2M.
So, my R13.3A = 14.5M.
After this, i unsoldered the parallel R and read back R1303A = 14.5M (maybe i did it wrong before).
So, i added a 10 M in serial in order to match the 24.5M.
I think i'll wait for my new transistor before doing wrong again

When you are reading 14.5MΩ, is at least one end of the resistor disconnected from the rest of the circuit, because it must be to get a correct reading.

 

I disconnect one end and get about the same result but the value is always changing. I'm not sure my chiness meter is good for big values.

 

I disconnect one end and get about the same result but the value is always changing. I'm not sure my chiness meter is good for big values.

What readings does your meter give for the 20MΩ or 10MΩ resistors and are the readings stable?

 

The resistance change and become stable after a while. What is sure is that you point the problem pretty well.
I don't know how i'll fix it cause this resistor will not be easy to replace.. i have only big values 1/4W..

 

the value is always changing

A thought: Your fingers should not touch the connections while you are measuring these resistors as you are lower resistance than the resistors.

 

this resistor will not be easy to replace.. i have only big values 1/4W

You could replace it with a string of lower values in series to make up the total required resistance. You will need to be sure of the voltage rating of the resistors and the actual voltage they will get in that string. You also need the space to be able to do that and make sure it is all well insulated against 2.5kV. The power rating of the resistors is not important. The total power in the original resistor is only about 1/4W.

 

I'm aware of that. I'm a noob for scope but have good skils in electronic (metal detectors and avr projects).
I'll loof for a replacement for this resistor

 

I can add 4.5M to my 22M and put some shrinkable tube, it will be very much smaller than the original one...
And then, replace the transistor.

I'll post the result soon.

 

2.5M, not 4.5. I have too big fingers for my phone

 

Is the 22MΩ rated for at least 3kV to include a safety margin? Most common resistors are 300V-500V and you would need at least 6 of these to be able to withstand the voltage.