My Everlast 50S plasma build observations

[msimpson99]

The following are some of my observations while doing some extensive experiments with the Everlast 50s plasma cutter on the CNCCS.

Most of these will present them selves when not using the ARC OK signal. When setting up a new machine I oftain will wait setting up this signal lead so I can get the timing setup properly.

Observation #1
While this is more general and not directly related to the Everlast 50s. The UC300ETH only recognizes dwell times in milliseconds. This is one reason I have Mach3 set to milliseconds in the book. (Note this also true of the UC100 and the All the Smooth stepper controllers)

In comes the problem with SheetCam creating pierce delays. First, you must set the time unit in applications settings to ms. But, even after doing that SheetCam still outputs the dwell setting in seconds. This will not work with MAch3 and the UC300Eth.

You must make the following changes to the Mach3 Post processor you are using.
In Pen down change
post.Number (pierceDelay ,"0.###") to post.Number (pierceDelay *1000,"0.###")

In Pen Up change
post.Number (endDelay ,"0.###") to post.Number (endDelay *1000,"0.###")

These will change the gcode output to ms.


Observation #2
When doing cuts in thin sheet at high speed, oftain the first cut will be delayed due to the time needed for the pressure to buildup in the torch. Depending on what I am doing I will do a single pierce on the scrap area of the stock. While that one will fail, all others will fire properly.

Observation #3
There is a latency between the Start signal and point at which the Everalast actually recognizes the signal and fires the torch. This is caused by many things. One of which is the large relay we are using to send the signal to the plasma cutter. I found that adding about 400ms to the pierce delay you desire seems to work pretty well.

Note that changing to the use of the ARC OK signal changes the timing and normally you go back to the normal pierce delay signal. One might wonder why not just always use the ARC OK signal. One reason is that if you want to add dimples or very light pierces to the stock, this works best not using the ARC OK signal. Note that sheet CAM has the ability to turn the use of the ARC OK on and off between operations.

Observation #4
The CNCCS was designed to be able to travel at high speeds with a high speed acceleration. From the research I have done .3Gs is the recommended acceleration. I have done some experimentation and found that setting the acceleration so that you achieve at least .15Gs will greatly improve cut quality on curves. As the acceleration increases the cut quality does as well, but with diminishing returns. The CNCCS can easily achieve .3G acceleration but certain things needs to be done to the machine with acceleration this fast.

1. If you are using the leveling casters I recommend, you need to retract them.
2. The resting pads for the casters is made of rubber. This can cause the machine to wiggle at high acceleration, you have two options here. Replace the pads with something harder. Or you can use some sort of brace that ties the machine into a nearby wall.
3. by doing very high acceleration, you need to pay particular attention to the machine tuning. You want all the preload screws on the bearings to be adjusted properly to eliminate any play. Also if you are using a floating head, it needs to be fairly heavy duty and keep the torch from any kind of play during starts and stops.

[msimpson99]

Real world Example.
This is a small base for piece of jewelry I make. 22Guage mild steel. It measures 1-1/2" x 3/4".
I'm cutting these at 20Amps using 40Amp consumables on the Everlast.

The following pictures are without cleanup. I only scraped the dross off.

Top

_MG_5775.jpg

Bottom

_MG_5776.jpg

At 250IPM and an acceleration of 1.5G, most of the dross chips off and the piece requires very little sanding. Sorry but I don't have any pictures of the same piece at lower Gs, but the dross is very hard to remove, lots and lots of sanding. Low speed dross really stinks.

Here is the part after sanding and powder coated with copper vein.

_MG_5779.jpg

I will be doing a video soon of the whole process of making the jewelry piece.

[msimpson99]

I did do a quick little video of the CNCCS with the floating head cutting a small part out of 22 gauge steel.

Notice the two sacrificial pierces at the start of the job. I found that since I am not using the ARC OK signal this gives me the most consistant cutting, especially for small parts on thin sheet.

https://youtu.be/FahojZp3X6o

[Rodw]

Try as I might, I struggle to see any observations specific to the Everlast other than perhaps it takes time for the air to reach the torch. Perhaps thats why Thermal Dynamics mount the air solenoid on the torch. In my experience I would think that would not be an issue if you waited for an ArcOK. But I understand with thin material why you don't wait for it when piercing holes.

Observation #5
One specific issue of the Everlast 50s and ArcOK that could be an issue for you is that if your ArcOK circuit on your controller does not draw enough current to exceed the minimum switching current of the Everlast's internal AcOK relay, the relay will oxidise and result in intermittent faults. This will happen if there is high impedance (resistance) between your CNC controller input and ground (assuming sourcing inputs). The solution is to add a pullup or pulldown resistor (depending on sourcing or sinking inputs) that is sized to ensure the current passing through the Everlast's relay contacts AND is rated to cope with the current. I wasn't the mathematician, but we found that if using 24 volt logic power, we needed a 1 k, 1 watt resistor to ensure reliable arcOK triggering.

[msimpson99]

When working out things on a new build, especially a new PC (Windows 10) and interface (UC300ETH) I like to get things working without the ARC OK. This helps me determine the timings and latency. Once I get things dialed in, only then do I hookup the ARC OK signal. So the ARC OK signal is not even connected to this machines controller.

I use a G540 and it does have a pullup resister built in. Once connected the ARC OK signal does work very well. I will add a couple Gcode snippets to sheetcam that will allow me to turn the use of ARC OK on and off as need before certain cuts.

Observation #6
I agree, I believe the very first Pierce does have some latency which has something to do with the air build up. Even my Hypertherm machine torch has this to a lesser degree than the Everlast. I would lay odds that the Hypertherm solenoid has a larger orifice to allow it to build air pressure much faster.