You need to consider hysterysis in terms of height....
1 volt is approx 0.5mm of torch height according to published material.... Hence 2v can be 1mm of torch height (act maybe a little more....) If your torch is flying around at 1.6mm above tge plate, that is only around 3 volts of height....

One of the failings of torch height controllers can be sampling speed in terms of input (sampled voltage) to output (up and down motion).... That is where the cheaper thcs can fail to match up when you are running higher ampage on thinner plate (higher feedrate required)

Hello Robert, thanks for your input.

The figures you were quoting was something I have failed to find, so I can base my dead band initially from that.

I use a drag tip at the moment and these are really only cheep things so not made to any great spec,  I will be using my lathe to modify these to remove the castelations and give more of a point to the nozzle to allow me to get a variation in height.

But first will run as standard with a few normal tips and measure arc volts if I can to see where it runs now, then try to set up with THC to get back around that same level.

Knowing the 1 volt per 0.5mm will help a lot.

I guess it will also depend on nozzle diameter, speed and width of cut/thickness of metal.  Could you point me to the doc if possible.

At the end it may be a complete failure with my set-up and approach, but I am still having fun on the journey.

Adrian

To avoid retyping a load of similar information, below is a link which provides a lot of the references you require, and provides guidance on published feedrates etc.
http://www.forum.cncdrive.com/viewtopic.php?f=4&t=168

HOWEVER, please note that this information may not be applicable to your specific application as I was discussing something specific here, to do with integrating THC with a motion controller, but the links and information should be off use (voltages and maximum expected feedrates referenced).  Also your plasma cutter is not the same as a hypertherm as far as feedrates pierce and cut heights are concerned.

Plasma is all about power (Watts), and Watts = Volts x Amps..... hence although your plasma may say 40A.... that is 40A at what voltage?

I found this manual for the Cut40, although there are likley to be a tonne of similar manuals with machines that look different all called Cut40 from different factories in China... so bear with me a little (not being negative or sarcastic toward them at all)
http://www.bjhowes.com.au/CUT40.pdf
https://www.google.co.uk/search?q=Cut40&ie=utf-8&oe=utf-8&client=firefox-b&gfe_rd=cr&ei=FIc5WOGdGu7W8gfL_rHICg#q=Cut40+pdf


The manual quotes 20 - 40A at 96V  (hence at best you have 3840W).... there is also a power factor consideration at 0.93% .... possibly.... because although its suppose to be an IGBT plasma cutter if you have a look at the VERY basic schmatic you'll see there is a transformer dispite the quoting of PWM at 100kHz etc .... hence I'd say you probably have around 3571W of metal cutting power say 3.6kW.... note there is also an efficency of 85% quoted... hence there may be some more losses somewhere too

I had for many years (still have it, just not used) a hypertherm powermax 350, which could be probably used to provide comparable cutting tables for your CUT40 (at least something to start with), given the cut rating is 10mm.... similar to the cut40 manual posted above (but the power rating is only 27A at 91V so 2457W... hence you may be able to get better than the PMX350 (which is the same energy rating as the PMX380, both of which are "older" transformer technology although they are not scratch or HF start (they use blockback start).
https://www.hypertherm.com/Xnet/library/library.jsp?file=HYP104427&format=pdf
https://www.hypertherm.com/Xnet/library/library.jsp?file=HYP104478&format=pdf
You can browse the hypertherm manuals here for any model (including FULL service manuals, sign of a good manufacturer):
https://www.hypertherm.com/en-US/customer-support/documents-library/



Also reading the manual the plasma cutter uses HF (high frequency) plus requires "scratch start"..... never good with CNC because of noise and interfearence potential plus not sure how your'd going to do the "scratch start" operation as it requires your machine contacts with the plate then backs away for around 1mm (according to the ref'd manual), then begins cutting.

my travel speed is slow circa 900mm per min at the moment possibly up to 1200 or 1500 if I get this to work, still very slow in comparison to commercial equipment.

Your feedrates are very (TOO) low.... plasma is all about acceleration too.... becuase the cutting current is constant and you need to change direction, you need to be able to have you machine accelerate and decelerate its axis as fast as possible. or you will end up with rounded corners and lots of dross.

How are you driving this plasma table?

What ampage stepper motors are you using?

What stepper drives are you using?

I started off with a load of ..... well not very good stuff and you end up buying twice to replace out the poor stuff to get the cut qulity where you want it.

Kind of like the old addage.... "measure twice, cut once"....... "design once, buy quality, never look back".

with plasma, you will come across the name Jim Colt a lot across many forums, its worth reading his posts

http://www.cnczone.com/forums/archive/index.php/t-94386.html
here is some guidance on acceleration (although really it's a function of feedrate..... higher the feedrate, you'll need more than the 40miliG listed here in my opinion as 40miliG (0.04xG[9.81m/Sec/Sec *1000 = 393.2mm/sec/sec)

because if you were running a modest 100"/min (2,540mm/min) feedrate, the distance travelled is 2.3mm

but now at a 200"/min feedrate, the acceleration and deceleration distance is 4.6mm  (which is probably the top end of your cut40 at a guess).

or at a top end (small scale) cutting feedrate of about 400"/min 10,160mm/min is 9.2mm.

you should get to know this formula well for your design:
Distance Travelled = Feedrate ^2 / (7200 x acceleration)
Distance travelled (in inches or mm)
Feedrate (cutting feedrate in inches /min or mm/min)
Acceleration (in in/s/s or mm/s/s)

http://www.engineeringtoolbox.com/acceleration-velocity-d_1769.html

note this is cutting feedrate (G1, G2 and G3 motion), but if you set the acceleration too high it will stall if you go for high rapid motion

There is a point when you wonder how fast do you really want your machine to go for rapid motion (g00).... depends on your pub... I'd go with knowing that you have good stable acceleration and your rapid feedrate is around your maximum cutting feedrate for a non-production enviroment (do you really need >600"/min??) as it won't save much time on a hobby if your maximum cutting feedrate is 400"/min.

a 4' traverse time will be less than 3 seconds saving ..... [4' (~1220mm), 600"/min (15240mm/min) = 4.8sec, excl accel + decel; 400"/min (10,160mm/min) = 7.2 sec; 200"/min (5,080mm/min) = 14.2 sec ...... bit slow for me though, but bread and butter for me is 2 to 6mm plate, occassional 1.5mm]

The generally accepeted rule is reduce feedrate to 60% for holes or radiused corners of less than 1" diameter (serves me well).
https://www.hypertherm.com/en-US/customer-support/faqs/how-can-i-cut-better-holes/

Your feedrates are very (TOO) low.... plasma is all about acceleration too.... becuase the cutting current is constant and you need to change direction, you need to be able to have you machine accelerate and decelerate its axis as fast as possible. or you will end up with rounded corners and lots of dross.

How are you driving this plasma table?

What ampage stepper motors are you using?

What stepper drives are you using?

I started off with a load of ..... well not very good stuff and you end up buying twice to replace out the poor stuff to get the cut qulity where you want it.

Kind of like the old addage.... "measure twice, cut once"....... "design once, buy quality, never look back".

The 900mm per min is typically my G1 cut speed.

I just can not get a lot faster so it is not possible with my set-up size 23 steppers on X and y and type 17 on Z.  Currents around 3 Amps X and Y and just under for Z, all from 24Volts supply and control max rating is 24Volts.  Again Chinese set up cira £60 purchase.

I am aware it is cheap but it will have to do.

Table size is 760 by 460 ish mm.  All home made and very, very budget.

I will have a read through the information you have provided, yes I have the manual for mine rated voltage is 100 Volts in the user guide.

The input is 4.7KVA at rated output of 40 Amps, so really wants a 22Amp supply but duty cycle is 60% at that those figures.

I need to run the machine under 25 Amps if I can reasoning is it is on a 13 amp socket.

220 Volts x 13Amps = 2860 KVA * 0.85% efficiency I have 2431 KVA available, 100 Volts rated so around 24 amps cutting current.

I can push a bit higher but the garage supply which is really only rated at 20 Amps has to power PC, lights and a compressor.

Thanks for the rest of the information I will read through the posts you have listed, thanks for that.

Adrian

p.s. even your modest feed rates are to fast for me, it is a case of do what I can for fun.

the 24 volts power supply will be your problem to some degree....

What stepper drivers are you using (link, or picture?)

Be careful with TB6600's or their smaller brothers .... I killed a few of them at 42V (before moving onwards).... but they should be good for about 35/36V.... it can be the back EMF that can be the problem.

What / how are you driving your table?  belts, rack & pinion...... not threaded rod please...

How are you driving your Z axis?    with THC accelleration is disregarded hence the axis shoots up and down with technically near infinate acceleration, because the stepper motors do not ramp u to the feedrate set, hence you really want a very course threaded Z axis (2-3 turns per inch) .... yes you did read that right 9 to 12mm pitch ...
https://www.aliexpress.com/item/1pcs-T10-10mm-screw-length-400mm-picth-2mm-lead-10mm-trapezoidal-screw-with-1pcs-brass-copper/32659186417.html?spm=2114.30010308.3.140.s33IlW&ws_ab_test=searchweb0_0,searchweb201602_1_116_10065_117_10068_114_115_113_10000007_10084_10083_10080_10082_10081_10060_10061_10062_10056_10055_10054_10059_10099_10078_10079_10073_10097_10100_10096_10070_10052_10050_424_10051,searchweb201603_6&btsid=4169b973-4f52-4d3c-9886-bc20e58025db

If you wish to see the project of building it then:

http://www.mig-welding.co.uk/forum/threads/cnc-plasma-cutter-start-of-bits-buying.52149/

The controllers are TB6560AHQ based the steppers on X and Y are 23HS9430B
X axis is 20 steps per  mm, and Y, which is my problem is 100 steps per mm as I am driving a 4mm lead ball screw directly at 1/2 stepping, it was more reliable and faster than 50 steps per mm with full steps.

Yes my Z axis is threaded rod rather then try and get a ball screw which was 5 inch long.

Adrian

Just found the information the Z stepper is NEMA 17 Stepper Motor JK42HS 1.8° 42x42mm [JK42HS40-1704]

Adrian

Those drives don't have a good reputation at all on the mycncuk forum.
http://www.mycncuk.com/threads/1373-eBay-TB6560-Stepper-Motor-Driver-Boards

Ballscrews are unnecessary (and not good in my opinion) for plasma because of the amount of very hard abrasive swarf floating about even with good extraction over time).

Don't cut aluminium, stainless or galvanised (zinc) coated steel without extraction and good makeup ventilation as it is carcinogenic.