heartless wrote:

going back to the OP's original question - this is definitely, 100% a mechanical issue... Something is loose or is flexing somewhere.

ANY flexing of the frame will be amplified at the printhead... and these cheap i3 knock-offs are famous for frames that flex.
Same with loose, sloppy belts... belts need to be quite snug. If they vibrate when the printer is in motion, they are not tight enough.
Check the bed, too... is there any rotational movement of the bed? (can you twist it) if there is, that is a problem.

Personally, I do not like this type of printer, but to each their own..

Thanks for the clarification!

I have the opposite issue. I have a MonoPrice Maker Select V2, which is a cheap i3 knock-offs are famous for frames that flex. And the frames is indeed flex. I have not add the z braces to stabilize it yet. The print quality is great,  no noticeable z-band when observe from one foot away with naked eyes (print at 0.2 mm layer height and 60 mm/s). The layer can be observed when look very closely or use camera.

It is very strange, the frame is even damaged in the left, slight push will cause frame shake along y-axis direction and I think the x-belt can be tightened (the frame will not shake during printing, only shake when pushed). The printer is noisy, lower motor current to quiet level will cause layer shift, so it is hard to understand the good print quality.

going back to the OP's original question - this is definitely, 100% a mechanical issue... Something is loose or is flexing somewhere.

ANY flexing of the frame will be amplified at the printhead... and these cheap i3 knock-offs are famous for frames that flex.
Same with loose, sloppy belts... belts need to be quite snug. If they vibrate when the printer is in motion, they are not tight enough.
Check the bed, too... is there any rotational movement of the bed? (can you twist it) if there is, that is a problem.

Personally, I do not like this type of printer, but to each their own..

Thanks for the clarification!

ANY flexing of the frame will be amplified at the printhead... and these cheap i3 knock-offs are famous for frames that flex.
Same with loose, sloppy belts... belts need to be quite snug. If they vibrate when the printer is in motion, they are not tight enough.
Check the bed, too... is there any rotational movement of the bed? (can you twist it) if there is, that is a problem.

Personally, I do not like this type of printer, but to each their own..

This printer is a basic metal frame I3 clone so I expect with proper tuning  it is capable of good prints. The trick is finding and correcting minor deficiencies.

IMHO there are a few printers out there that are cheap junk .And there are certainly well built ones, but usually expensive printers . The old adage you get what you pay for.  And in the middle ground is real life printers that take time, patience, and sometimes a few quality parts to cross that  line . We often seek the holy grail of a sub $ 300 great printer.  You save money but they take time to tweak.
In realty it usually takes a bit of effort to get the printer we have to do what it s supposed to and what we want it to do.

One one has masted his or her printer there are only a couple reasons to purchase a second third etc printer. 

1) one needs or wants a larger printer.  While one can splice parts this is of ten an unwanted task and since no one has invented a printer stretcher a larger printer can be a handy thing to have.
2)  success.  Printing so many good parts that one needs an additional printer to keep up with demand. 
3) Experience and wanting to get a real  printer. But beware there will still be a learning curve. 
I am talking real upgrade here like a ultimaker or genuine prusa .
4) wanting to build a kit to educate yourself in how a printer is built set up and tuned.

However continuously buying printers trying to find the best one is IMHO a futile effort and you end up wasting money.

I hope this helps

You don't always get what you paid for. I was amazed by the apple like design of Cube3D and read about it and it's high end model CubeX Duo, the review on amazon for CubeX Duo is flooded by angry user. It is $3,000 printer but get 77% one star review. And many comments start with give one star because 0 star is not an option.

Because the people doing the reviews are probably new owners like the OP and don't know what a good print is supposed to look like. I have seen many prints posted here and on other sites that the poster called amazing but to me looked like trash. Layers separating, ringing around hole, bad infill, and so on. Again if you have never seen a GOOD print then you won't know what one looks like.

Agreed. When I have my first 3D printer, and I know nothing about how to print correctly, I print at 100mm/s and still amazed by the quality even the corner is round and z-band are obvious. Then one day I slow it down, I was amazed again, then I see the prints for SLA printer, it was amazing quality! Then I see people post process the print with acetone steam bath, the quality is amazing...

IMHO there are a few printers out there that are cheap junk .And there are certainly well built ones, but usually expensive printers . The old adage you get what you pay for.  And in the middle ground is real life printers that take time, patience, and sometimes a few quality parts to cross that  line . We often seek the holy grail of a sub $ 300 great printer.  You save money but they take time to tweak.
In realty it usually takes a bit of effort to get the printer we have to do what it s supposed to and what we want it to do.

One one has masted his or her printer there are only a couple reasons to purchase a second third etc printer. 

1) one needs or wants a larger printer.  While one can splice parts this is of ten an unwanted task and since no one has invented a printer stretcher a larger printer can be a handy thing to have.
2)  success.  Printing so many good parts that one needs an additional printer to keep up with demand. 
3) Experience and wanting to get a real  printer. But beware there will still be a learning curve. 
I am talking real upgrade here like a ultimaker or genuine prusa .
4) wanting to build a kit to educate yourself in how a printer is built set up and tuned.

However continuously buying printers trying to find the best one is IMHO a futile effort and you end up wasting money.

I hope this helps

I just go read some reviews about anycubic i3 mega. Every review said it is great printer, however, the sample print they post are terrible quality. How did they conclude this is a great printer?

Because the people doing the reviews are probably new owners like the OP and don't know what a good print is supposed to look like. I have seen many prints posted here and on other sites that the poster called amazing but to me looked like trash. Layers separating, ringing around hole, bad infill, and so on. Again if you have never seen a GOOD print then you won't know what one looks like.

Look where the X and Y drive rods are housed in the case. Should be some sort of bearing or sleeve with a housing of sorts. Make sure those are not cracked and there is no play in them. This is normally caused by a degree of give or play in the axis making the long part of the egg shape.

This printer moves the Y axis by moving the bed. Make sure there is no slop or give in the belt and make sure the pulley is secure on the motor shaft. Sam for the X which is moved on the gantry the print head rides on. Make sure there is no slop in the belt and the gear is secure. Also make sure the guide rods are square on both axis. Literally measure the space between them at both ends and the middle. Of it is not exact on all three spots then your rods are bent or the mount is wrong. A very small deviation will cause a big difference at the print head.

Above bolded words are words confuses me. Since I'm newbie in both 3D printer and English, I have no idea what they refers to or what they means. Maybe some picture will help.

This is not a VREF issue.  There is clearly something mechanically incorrect.  Ovals are typically caused when X isn’t operating perfectly perpendicular to Y.  In your case because the ovals are inconsistent, you have slop in the system.

I mentioned that, but it looks like you where the only one who noticed it Ward. I have never really figured out why people ask for advice but then ignore it or overlook it when it is given.

To OP the only thing Vref will effect is moire if too high and binding if not high enough.

As I stated earlier the issue is with your axis and some slop or give somewhere. Only you can find it though. Like I said though measure the distance between each guide at each end and the middle. IF they are not exactly the same, that could be your issue. A very small amount of variance translates to a large amount at the print head.

Look where the X and Y drive rods are housed in the case. Should be some sort of bearing or sleeve with a housing of sorts. Make sure those are not cracked and there is no play in them. This is normally caused by a degree of give or play in the axis making the long part of the egg shape.

This printer moves the Y axis by moving the bed. Make sure there is no slop or give in the belt and make sure the pulley is secure on the motor shaft. Sam for the X which is moved on the gantry the print head rides on. Make sure there is no slop in the belt and the gear is secure. Also make sure the guide rods are square on both axis. Literally measure the space between them at both ends and the middle. Of it is not exact on all three spots then your rods are bent or the mount is wrong. A very small deviation will cause a big difference at the print head.

+1