Ripple Current

[islander261]

Hello Jim

I have been reading some papers on torch height control. One of the things these lab guys use is related to the ripple current frequency of the plasma power supply when cutting. I know that on our small inverter air plasma units this will be the switching frequency of the inverter or some multiple of it depending on the how many phases are used to drive the output. So can you find out what the ripple current frequency is for the Hypertherm small cutters we commonly use here? At this point even a range will be helpful. I am sure the frequency is quite high because I cannot hear the switching on my TD unit and may be in the same band as most of the arc noise which will make it not practical to use the property. I am also sure this is a topic that has little interest from the general membership here, if anyone wants links to papers PM me and I will send them.

John

[jimcolt]

John,

The PWM frequency in the current inverter units that we build is between 20 KHz and 70 KHz (depending on model). The frequency stays fixed but the on/off duty cycle varies based on current requirements.

I have been working with plasma torch height controls since 1978 when we (Hypertherm) introduced our first one....so if you have questions feel free to ask.

Best regards, Jim Colt

Hello Jim

I have been reading some papers on torch height control. One of the things these lab guys use is related to the ripple current frequency of the plasma power supply when cutting. I know that on our small inverter air plasma units this will be the switching frequency of the inverter or some multiple of it depending on the how many phases are used to drive the output. So can you find out what the ripple current frequency is for the Hypertherm small cutters we commonly use here? At this point even a range will be helpful. I am sure the frequency is quite high because I cannot hear the switching on my TD unit and may be in the same band as most of the arc noise which will make it not practical to use the property. I am also sure this is a topic that has little interest from the general membership here, if anyone wants links to papers PM me and I will send them.

John

[islander261]

Jim

Thank you, I suspected that the high ripple frequency would be the case. Most of the older papers and patents used traditional direct conversion line powered industrial (3phase)plasma power supplies so the ripple was at 360hz. The reason for my interest is the paper mentions detecting the lower peak of the ripple for best indication of the electrode to plate top distance. Does your staff even think about the reported 19khz anode spot oscillation, or is this frequency dependent on cutting conditions and material thickness? I haven't found a clear explanation other than the anode spot moves up and down in kerf. I know that my Torchmate AVHC controller low pass filters the arc voltage at 1khz with a single pole RC filter and then only digitizes to 10 bits with an on-board converter that only has about 8 bits effective resolution when you add up all the errors.

I have read your requirements for good torch height control system in other posts.

John

[jimcolt]

The anode attachment to the material varies continuously.....is has to do with the molten metal flow and the paths of least resistance for the arc, not predictable. The height controls today certainly are well filtered and also do some averaging in order to have fluidic, constant height. The worse thing you can do is to try to make the height control react too quickly as you will end up with rapid oscillation that will clearly show as roughness in the cut face. There are many other things to consider as well such as kerf crossing, the ability to maintain accurate height during deceleration/acceleration, the ability to calibrate continuously to compensate for consumable wear (height is most important, voltage is not) and of course pierce height and delay accuracies. The height controls used on our industrial systems today are pretty much transparent to the machine operator. No settings, all of the data for each material, thickness and power level is stored in the CAM software. Height control auto calibrates on every cut. The height controls on most low cost tables leave a lot to be desired in comparison...and definitely need improvement, however no one wants to pay for them! The component that many aftermarket companies forget is the z axis mechanics......which often is the biggest part of the cost to build a decent height control, but is necessary and is a critical functioning component.

Jim Colt

[islander261]

Jim

Thank you again for the informative reply.

This sounds like the high end industrial machines tightly integrate the materials data base through all aspects of their operation, cam, motion and torch control. From what I think you said the torch height control either uses a set acceleration from the data base or instantaneous speed from the motion control and factors from the material database to adjust the torch height when accelerating and decelerating, is this correct? Is the simpler version of this is having your CAM insert torch height hold commands into your motion control code? Is the continuous height calibration done by sampling the arc voltage when cutting at a known height and stable speed, this voltage is then used to keep the torch at the set point height during that cut? I think most of the lower cost torch height controllers assume the arc voltage is fixed for given torch height.

I found that the methods for kerf crossing to be the subject of more than one patent. I am familiar with many types of filtering used in the feedback path in servo systems along with output rate limiting to protect the mechanics or process under control. I was just trying to get a first order feel for the range that practical plasma controls need to operate in. Since I retired I don't have access to nice electronic measuring equipment anymore to easily run out and make my own measurements.

John

[jimcolt]

During slow downs there is a "% of cut speed" setting in the cnc, when the speed drops below this threshold the height simply freezes, this eliminates diving of the torch. The percentage varies from machine to machine, but usually is in the 80% to 90% range. Sample and hold at the beginning of each cut is used on Hypertherm's industrial height controls....this tends to compensate for consumable wear, changes in cut speed, etc. and maintains physical height better than an operator guessing on the arc voltage needed to achieve a particular physical height. Interstingly the Plasmacam (entry level) cnc machine uses a great sample and hold height control system, flawless, and on a low cost machine. Servos as well. Hypertherm had the first patent for kerf crossing (recognizing the voltage profile of kerf crossing and learning to ignore them), I'm fairly sure it has since expired. CAM could not now the real time speed of a machine, because the mass of the moving parts and the size of the drives, ratios, etc vary from machine to machine...better to let the cnc control the height control freeze for slowdowns.

Jim



John