It is currently Wed Dec 13, 2017 5:29 pm

 

Everlast CNC Port Errata - How it really works

Discussion forum for Everlast & Longevity plasma cutters
Rodw
Posts: 87
Joined: Sun Aug 21, 2016 1:49 am

Everlast CNC Port Errata - How it really works

Postby Rodw » Wed Aug 09, 2017 5:34 am

Guys, I spent last weekend working out what was wrong with the Everlast CNC port in my PowerPlasma 50 purchased July 2016 from Everlast Australia. The documentation is riddled with errors so if you have one of these machines and are trying to get it going on a plasma table, ignore everything except this information. The documentation is identical on a number of other machines, so there is a good chance this applies to other Everlast machines.

This is the board I am talking about (shown with connectors removed):

Image

The 12 port CNC port is directly behind this board and you can see my connector (provided by Everlast Australia) is plugged in.

So here is another shot and notice the position of the link in the photo
Image

Error #1
The drawing in the manual is from the bottom side of the board (that is OK and easy to work out)
Error #2
The manual says that the 2 pins on the left select the divider ratio depending if they are shorted or not. This is incorrect. These 2 pins are connected on the board so shorting them has no effect. The correct position for the link is to connect it to the leftmost 2 pins or the rightmost 2 pins. More on this later.
Error #3
The manual states it is connected to the workpiece clamp via a 100k resistor. It is not connected to anything.
Error #4
The manual states that the pin numbers on the board are identical to the pin numbers on the connector. This is incorrect but the connector pinout is correct. The board signals has the divided voltage and the raw voltage swapped around. This is not a problem because you should not need to worry about the signals on the board.
Error # 5 - and this is a big one
The 50:1 and 16:1 voltage dividers are incorrect! They are actually out by a factor of 1.5
The correct divider values are:
50:1 divider is actually 75:1 (I measured 76.145:1)
16:1 divider is actually 24:1 (I measured 24.529:1)

What does this mean?
If you want 75:1 division, place the link on the 2 pins closest to the CNC port connector
if you want 24:1 division, place the link on the 2 pins furthest from the CNC port connector.
Use the pinouts for the CNC port with confidence.

EDIT: Errata on the errata

Further testing has revealed that the board actually is 16:1 when tested with a multimeter. The hardware I'm using to measure the divided voltage has a a 10 v range internally that can be scaled with an external resistor. The resistors on the Everlast divider are being counted as being part of the external scaling resistance so on my hardware Mesa (THCAD-10), the division varies with the scale set by an external resistor. eg
20 volt full scale reading becomes 24:1
10 volt full scale reading becomes 32:1

My next post will outline the tests I did to support these conclusions.
Last edited by Rodw on Sat Sep 02, 2017 5:32 pm, edited 2 times in total.

Rodw
Posts: 87
Joined: Sun Aug 21, 2016 1:49 am

Re: Everlast CNC Port Errata - How it really works

Postby Rodw » Wed Aug 09, 2017 5:35 am

For the last 12 months I've been building a plasma machine and table using LinuxCNC and a Mesa 7i76e plasma controller. When I fired up the machine for the very first cut, the torch voltage came in around 19-20 volts. This was clearly incorrect. I first thought that there was a bug in a component I had written in C for LinuxCNC. But that checked out 100% perfect after I wasted a few days troubleshooting. So I decided to check it out more closely once I had the machine dialled in on one material and a bit of time.

So the first thing I did was pull the machine apart and change the divider link position. This had no effect! I pulled the board out of the machine and inspected it in bright sunlight and I found the link problem.

But changing the link did not deliver meaningful results.
So I found a set of multimeter leads so I clipped off the probes and soldered them into the CNC port connector
Image
And ran them up to my multimeter and tried some cuts
Image
But the divided voltage still made no sense
Image
So I called in a mate who knows more about electronics and me and asked him to help me draw out the divider circuit. We worked out the link was shorting out a 67k resistor thus taking it out of the dividing circuit. I still did not have a clue about how things were working so I decided to waste the following morning drawing up the circuit myself. This let me uncover some of the other errors but I still could not work out what the divider ratio was.

So after talking to a couple of people and taking on board Tom Caudle's feedback on this forum, I remembered I had a couple of 48 volt power supplies laying around. So I connected them in series to get 96 volts.
Image
And I plugged them into the CNC board.
Image

Image
And read the divided voltage
Image

I'd also been lent another multimeter so I plugged it into the soldered in leads and noted about 2 volts loss across the resistors on the raw arc voltage (1.95 volts actually)
Image
While this was going on I could also see the voltage in LinuxCNC and I got it to log torch voltage and a few other stats at 1000 times a second.
Image

Initially, I was getting 150:1 and 75:1 dividers from this measurement process but then I remembered that I had already scaled the voltage as pte the Mesa THCAD manual which I was using for voltage measurement to double the full scale range. Once I measured the voltages before this scaling resistor and adjusting the calibration in LinuxCNC accordingly, I was getting perfect results.

So doing another cut after all of this using the same settings I used in the earlier photo with the sparks, I observed a constant voltage of 84-85 volts which is pretty well spot on with the voltage shown in the photo of the grey multimeter.
Last edited by Rodw on Wed Aug 09, 2017 11:11 pm, edited 1 time in total.

User avatar
acourtjester
Elite Contributing Member
Elite Contributing Member
Posts: 2818
Joined: Sat Jun 02, 2012 6:04 pm
Location: Pensacola, Fla

Re: Everlast CNC Port Errata - How it really works

Postby acourtjester » Wed Aug 09, 2017 9:35 am

Thanks for the report, I'm sure it will help others going on the same path :)
DIY 4X4 Plasma/Router Table
CandCNC DTHC-IV & Feather Touch
Hypertherm PM65 Machine Torch
Drag Knife and Scribe
Miller Mig welder
13" metal lathe
Small Mill
Everlast PM 256 tig

rrossetti
Posts: 9
Joined: Thu Apr 17, 2014 4:34 pm

Re: Everlast CNC Port Errata - How it really works

Postby rrossetti » Thu Aug 10, 2017 3:17 am

Hi
here is the fix for the 50-1 divider board to work correctly

as used by us at R-tech

You do not have access to download this dxf file.
Become a Contributing Member to gain access to the thousands of files shared by our members. Click Here


Rodw
Posts: 87
Joined: Sun Aug 21, 2016 1:49 am

Re: Everlast CNC Port Errata - How it really works

Postby Rodw » Thu Aug 10, 2017 3:39 am

rrossetti wrote:Hi
here is the fix for the 50-1 divider board to work correctly

as used by us at R-tech


Awesome. Does that also give 16:1 with the link in the other position?

rrossetti
Posts: 9
Joined: Thu Apr 17, 2014 4:34 pm

Re: Everlast CNC Port Errata - How it really works

Postby rrossetti » Thu Aug 10, 2017 5:58 am

hi never tested 16-1
not needed on our setup
http://www.r-techwelding.co.uk/r-tech-c ... 4ft-x-4ft/
cheers rob

Rodw
Posts: 87
Joined: Sun Aug 21, 2016 1:49 am

Re: Everlast CNC Port Errata - How it really works

Postby Rodw » Thu Aug 10, 2017 6:32 am

rrossetti wrote:hi never tested 16-1
not needed on our setup
http://www.r-techwelding.co.uk/r-tech-c ... 4ft-x-4ft/
cheers rob

Rob, Thanks, I just have to change one resistor and I can scale to anything. Can you give me an idea of what the maximum torch voltage you've seen on the Everlast 50 amp? Nice table by the way.

Njcnc
Posts: 5
Joined: Tue Aug 29, 2017 11:04 am

Re: Everlast CNC Port Errata - How it really works

Postby Njcnc » Tue Aug 29, 2017 11:53 am

Oh man, thank you both so much for this info. I have been running Everlast plasma on two separate builds now with raw arc voltage and Proma THCs. I hated having to run raw arc voltage out of the until to the Proma. I'm heading home today after work to jumper that board and see if the Proma will read the 50:1 after the mod.

Njcnc
Posts: 5
Joined: Tue Aug 29, 2017 11:04 am

Re: Everlast CNC Port Errata - How it really works

Postby Njcnc » Sat Sep 02, 2017 1:44 pm

Wanted to post back that I went ahead and did the mod pictured above. I have a powerplasma 60s and the board looked a tad different but found the same thing. Pin 4 at the plug is a black wire going to pin 7 on the board. Pin 6 at the plug is a red wire going to pin 5 on the board. Jumped the resistor at pin 7 on the board and fed it into the Proma's 50:1 divider. I got a bit over 90 volts firing in the air and about 62 cutting. The Proma will read as low as 45v so this is going to work out, no more running raw arc voltage out of the case. Here was a cut done in 1/8" plate at 75ipm, @45amps and the THC set at 62v. Had almost no dross, just a quick tap with a hammer took it off.

IMG_20170902_133816.jpg

You do not have access to download this dxf file.
Become a Contributing Member to gain access to the thousands of files shared by our members. Click Here


Rodw
Posts: 87
Joined: Sun Aug 21, 2016 1:49 am

Re: Everlast CNC Port Errata - How it really works

Postby Rodw » Sat Sep 02, 2017 5:59 pm

Hmm,

Further testing has revealed that the Everlast manual is correct about the division ratios. They are indead 16:1 and 50:1
BUT:
Many THC's must have some sort of scaling resistance internally to scale the input voltage to suit their internal ADC. They may see the Everlast division resistors as an extension of their divider network and display erroneous results.

So shorting out a resistor as shown may work on one THC but be incorrect on another.
In my case, my THC hasa 0-10 volt sensing range internally and can be configured using an external resistor to whatever you like;

So with the divided voltage disconnected from my voltage sensing hardware, a multimeter shows the dividers are indeed 16:1 and 50:1 but as soon as plug the wires into my sensing hardware, the voltage division at the same point becomes:

Code: Select all

Native 0-10V range   Dividers of 32:1 and 100:1
Extended 0-20V range Dividers of 24:1 and 75:1
 


So it seems you need to test against a known voltage to confirm the final dividers on your hardware. Just also be aware that an unloaded DC switch mode power supply may still have a significant AC ripple in it.


Return to “Everlast & Longevity Plasma Cutters”

Who is online

Users browsing this forum: No registered users and 1 guest