@Camarda_Neon_Salerno some things to try:
Disconnect to the IN pin on the LPS.
Connect the IN pin on the lps to 5v and push test again to see if you get full and steady power. Don’t hold for a long time!
I also would remount the flyback.
Not sure what the small red wire on right of flyback is?
For pleasure you would send me an illustrative image
Verify the diagram below is your LPS.
For now ignore the test process on the left of the drawing and pay attention to the P#'s for this test.
Turn off power to the LPS
Remount and rewire the flyback
Disconnect the wires on P2-1(5V) and P2-2(IN)
Connect a wire between P2-1(5V) and P2-2(IN)
Turn on the machine
Push the test button on the supply while taking a video of the meter as before.
Warning: be careful if the supply has the cover off like seen in the pictures above.
With this test the laser will fire will full power so do not hold the “Test” very long.
If this solves the current fluctuation you likely have a bad pot.
If it does not we will try the next test.
Post results here…
Doing the bridge and the power is stable at 21 mA what and then the problem?
please help me
@Camarda_Neon_Salerno apparently there is something wrong with either the connection to pin P2-1,2,3 of the LPS or the other end of that cable assembly where it plugs into that control panel card (Right connector on the picture below).
I would look for bad connections or broken solder joints on that board and the corresponding cables that connect to the LPS P2.
If you do not find anything then you may have a board board (pict. below).
@HP_Persson do you see anything I have missed?
Does the output vary also, like the meter shows? Or if you look at the tube, can you see the beam pulsing ?
To me, it looks like a mA-meter with too big resolution it cannot hold quick enough due to PWM. (one guess). My first meter did similar to this, it was a 100mA meter, it vibrated 
But if it also pulses on the beam there is a board problem as @donkjr showed above. You may need to replace it, or convert to analog (actually pretty easy).
@HP_Persson when he bypasses the control board [IN held at 5VDC] the current is stable. He is also pushing test so the PWM is out of the circuit?
I also connected an analog potentiometer to see if it was the fault of the pwm but the same creates oscillations
@donkjr but if you bridge it, there is no modulation going on? or is it?
Don´t have a oscilloscope so cannot check it myself.
Same pulsing with pot, i still believe the mA-meter is the problem. (too slow meter)
@Camarda_Neon_Salerno if you test engrave or cut, does the beam inside the tube pulse, or can you see on the engraving that the beam is pulsing, with different strength in the engraving? Only look at the result and the tube now, not the mA-meter 
@HP_Persson the test pulls IN up to 5v which is full on when the test switch turns the supply on. This overrides the signal on L.
My understanding is that @Camarda_Neon_Salerno is running this test statically with test button.
Also the meter works fine with the IN pulled high and test button pushed, so logically the meter seems fine?, no?
@donkjr Yeah, sounds logic. But, if the meter pulses like that, no matter if the signal is modulated by the digital board or a pot, he should se variations on the material cut or engraved too.
That would answer the question if it´s a meter problem, or something else.
Isn´t it the other way around? 0v is full power, 5v is no power?
3 am right now, my brain is in slow mode 
@Camarda_Neon_Salerno i must be interpreting something wrong so lets recap:
- When you push and hold the “laser test” button on panel the meter fluctuates and there is no job running on the machine at that time. Is this correct?
- When you disconnect the current wiring and then connect IN to 5VDC, then push and hold the laser test button on the panel the meter does NOT fluctuate. Is this correct?
- When you install a potentiometer (value?) across the 5vdc-IN-gnd then push and hold the laser test button on the panel the meter DOES fluctuate. Is this correct?
- Si
- Si
- No
Both with the potentiometer and with the digital control connected the laser power is unstable while disconnecting power control and I make the bridge that you showed me in the scheme the power and stable at 21 mA
@Camarda_Neon_Salerno how did you connect the pot. At the end of the cable or directly to the power supply?
check your ballast resistor for burn marks
I connected the potentiometer to the pins: P2-3 P2-2 P2-1
@Camarda_Neon_Salerno scratching my head:
The pot connected directly to the LPS the current is unstable.
With everything disconnected from the supply and IN connected to 5VDC the the current is stable.
mmmmm…
With the pot connected directly to the LPS is the current unstable no matter the pot position including all the way to either limit?
+Don Kleinschnitz I did not understand your last message