Dimension problem - Circles out of round

[TLJC]

I am cutting a bunch of circles for a customer. 12 gauge mild steel. They are supposed to be 52" diameter. They are 1/8" out of round. What could be causing this issue? 40 amp consumables, 1/16" cutting height, 150 imp, .25 pierce delay. The pieces look good, they just aren't round. Any thoughts?
Thanks

[Kentuckyfarmer]

I am new at this but from what i have read and limited personal experiance check all things mechanical. I had a set screw work lose in a pulley, circle turned into eggs in a heart beat. Any play or losenes will mess with the circle shape.
Hope this helps, others with more experiance will have more and better advice
Mike

[SCHIDTHEAD]

Ingredients: 1 carbide tipped metal scribe
1 can red primer
1 mic or flat steel measuring apparatus
1 pc. of flat steel

Spray a light coat of paint on your steel and put it on the table. secure your scribe to either your torch or remove the torch and use a wooden dowel in your holder so it doesn't move. Set measurements on x and y axis and run so the scribe clearly marks the lines on the steel and mic or measure the lengths. Then do the same with a equal sided square and mic/measure the sides and from corner to corner to check square and adjust your controller .ini x and y settings to set. If all is true to measure then nothings loose on your table. Is it dual or single axis drive? If it is dual you may have a sync problem with the motors and I cant remember right off how to check but it may come back to me here in a few...

[jmsrbrt]

I am cutting a bunch of circles for a customer. 12 gauge mild steel. They are supposed to be 52" diameter. They are 1/8" out of round. What could be causing this issue? 40 amp consumables, 1/16" cutting height, 150 imp, .25 pierce delay. The pieces look good, they just aren't round. Any thoughts?
Thanks

Agree with Kentuckfaremr...most likely a mechanical issue such as a loose setscrew or improper belt tension.

[TLJC]

Thanks guys,
I will go over everything mechanical first. I will then do the scribe test.
I will keep you posted.
Thanks again!!

[SignTorch Vector Art]

if it is consistently off by 1/8" across 4 feet, yet the end-of-cut consistently meets back up with the start-of-cut (not off by up to 1/8" sometimes) it could be the axis steps-per-inch setting in the CAM software.

[Bigrhamr]

Under Settings-Machine-Scaling check that the numbers for x and y are the same. That is the number of encoder counts per inch of travel. Assuming you have the same gear reduction and pulley size on both axis the numbers should be the same, if there is a difference in gearing this is where you would adjust for it.
Is the long measurement always on the X axis by any chance?
My G-5 has 28-1 gearboxes (tag on the box) and 15 tooth pulleys. Is yours the same? And what are the scaling numbers set at now?