oscilloscope tube traces flickers(help sought)

[Unica_Research]

I made a oscilloscope tube driver circuit last week. it's flicker when you connect blanking signal to everywhere(GND, +140V,blanking signal). if you make blanking line open circut,so traces brightness as normal but a retrace line gonna arise.
could you tell me how to solve it?
Thanks

[Unica_Research]

more picture

[Unica_Research]

picture 3

[Unica_Research]

picture 4

[perdrix]

Of course it will flicker at line frequency triggering - that's 50 or 60Hz, readily detectable to the human eye.  Try trigerring on the square wave itself (assuming it's about 1kHz or so).

[StillTrying]

Supply ripple on the -2kV, cathode and grid which cancel until you add a steady voltage to the blanking/grid, just a wild guess.

[Unica_Research]

Thanks for reply
Even 1 KHz is used for triggering, it will produce brightness flicker like water ripples.

[Unica_Research]

wonderful suggest!
I will try. 

[shakalnokturn]

How did you go about filament supply?
AC and with a winding isolation >2kV?
Why the choice of R1 and R2?

[Unica_Research]

Thanks for reply
My filament power supply is independent. When the filament power is turned off, the cathode is still hot and emit electrons to light up the screen, but the screen still flickers until cathode cools down.
In other words, it has nothing to do with R1 and R2.
I can't solve this problem until now......

[StillTrying]

Is the 1kHz exactly 1kHz?

On picture 4 the brightness ripple period is about 8.5 1kHz cycles making the brightness ripple frequency 117.6 Hz, which is very close to 2 X 60Hz ripple, is the mains there 50Hz or 60Hz?
The flag and the web says it's 50Hz.

I don't think it's anything to do with AC heater leakage because I think that would produce 1 X mains frequency brightness ripple.

We don't now anything about where the -2kV or blanking signal comes from, is the brightness ripple still there with just the blanking signal's GND connected but not the blanking signal itself, to check for a GNDs problem.

[shakalnokturn]

How about moving D2 to the top of your resistive divider and tying the bottom of the divider to ground?

[Unica_Research]

your suggest is working, but a little flickers still now.

[Unica_Research]

I have measured a waveform of grid by high voltage probe.
No -2KV, a rectangular wave arise in grid.
when I turn -2KV power on,it reduce peak-peak voltage of Grid rectangular wave.

P.S.   blanking signal is a rectangular wave.

[shakalnokturn]

The brightness is quite sensitive to grid voltage variations, if HV supply is not stable (ripple) this ripple will also affect the brightness because of ripple leakage through C10.
You may want to consider better filtering of -2kV (C12) and regulating this voltage.
If you're using a ceramic capacitor for C10, the capacitance decreases with voltage which may explain why you're loosing amplitude when grid bias is active.

[Unica_Research]

 HV supply is not stable, final.
there is a Line ripple on HV, I found it by other oscilloscope with high voltage probe.
Ripple disappeared after replace power supply.
I solve it...

Thanks for your help.........

[StillTrying]

Supply ripple on the -2kV, cathode and grid which cancel until you add a steady voltage to the blanking/grid, just a wild guess.

In that case I think my wild guess was quite good.