Nobody have any input?

It extrudes both pla and abs 3mm (2.8) ok, good quality and so. Printable but:

I get diameter variances that goes in waves. and i *think* waves superimposed on the big waves.

The upper wave show what it looks if i measures over aproximately 50cm to 100cm depending on speed.
But if i measure over a longer period ( couple of meters) i start to see a wave superimposed over the 50-100cm.

the 50-100cm variation is about 0.05mm and the second is about 0.1mm, but at certain moments
the waves adds onto each other so the variation sometimes is 0.15-0.2mm.

So for like 5m i get ok diameter, then the next 5m is way to big variation, the next 5m is then again ok.

At first i thought it was because i had a winder that start, stop at intervals. So i made a winder that runs continusly
at fixed speed, and it was then i got a visual of it.  The shadow on the sensor slowly went up, then down and after a couple of oscillations it got bigger and bigger then suddenly smaller and smaller until it was back at the small up/downs.

So. Is my problem with the auger? The nozzle? Too small meltzone? Too small/big gap between augerend and nozzle?
It must have something with pressure buildups to do i think.
The nozzle has meltfilter, then a breaker (just a plug with about 10holes, the plug is coned inwards closest to nozzle, the nozzle is flat. maybe i should have a nozzle that is coned on the inside?

The motor is a BIG 230v motor (2hp?) with 34rpm gearbox into a 16mm auger. Motor is run from a Variable Frequency Drive.
The VFD makes sure the motor runs the same speed even at load.
I have two melt-zones aprox 10cm apart. 2 PID controllers both pidtuned. First melt-zone a little cooler then the second.
Vertical mounted with loop to the floor then up to winder.

I have tried speeds (augerspeed) from 6rpm (1kg in 8hours) up to 25rpm (1 kg in 3 hours), and i have tried different temperatures.

I really hope you all understand what i mean (english is NOT my first language).