[Oracle_9]

Hello,

I am in the early stages of my wire-edm project. It will be a small DIY type desktop cnc, which I will be making. The cutting area will be about 12" x 12" and the height of the material-to-be-cut is undetermined yet. Typically from thin metal sheet to 1/2" thick. I have obtained plans from Camtronics which comes with a very simple software that would run it, but it is a very basic software...and its about 54 kB file.

Maybe to test the system, I would use the provided software, but I would like to use something a little better when I will be using the machine in the long run. I am looking for a software a PC can run the wire-edm via printer port. And that has functions and features for the wire-edm process.

Does anyone know of such a software? And if so, offer any suggestions?

Thanks.

[jcc3inc]

Sir,
My understanding of wire edm is that the current is measured, and if a short is noted, the controller backs up along the cut path until the short is cleared, then progresses forward again. This is a specialized function requiring storage of a discrete number of steps on a continual basis, so that the exact path may be retraced. I believe that this function would not be too hard to implement, but there are few who want this function.
Do you know in detail the performance requirements you want? I'd be interested in your comments. I have built a 3 axis controller similar to the TurboCNC by DAK Engineering.
Regards,
Jack C.

[Oracle_9]

Hello,

First off,

Having re-read the software description in the manual, it should be ok.

The provided software has features to move manual move the wire to a start position before the EDM process begins, a feature to set how much steps for the wire to advance during the edm process, and how much to retract, a pause time for the wire to stay stationary to allow time for the workpiece to clear out. There are other features as well, but too much to include..mostly on settings to configure your PC and the CNC.

Basically, it functions like you described, except it does not retract all the time, only for some reason, the wire is in direct contact with the work piece. But, typical it remains stationary for a given time for the workpiece to clear away, and then advances again.

All the parameters of how much it advances/retracts/pause time, have to be inputed by the user. This is mostly trial and error until you get the desire values.

The provided software should work well for this particular small wire edm cnc from what I have read and been told. I'll try it out when the time comes and see.

The main things that I would like changed are more of a personal preference. Right now the program operates under DOS. And it has no visual features, only text commands you type in. Preferably, I would prefer it to be under Windows and to have a visual characteristic, like buttons for menu and othe commands. A better way to import files. Something to syntax check the g-code for errors, a joystick function would be nice too.

Just to avoid an arguement, I do realise that wire-edm cnc is not a simple thing. And the program associated with this is not simple either. A lot of factors involved, such as the circuitry involved. Such as what the industrial wire edm's can adjust current depending on the type of metal and thickness used, whereas my simple wire edm has a fixed current (I think??), and the type of circuit used to monitor the short/discharge...etc.

Anyway, I am looking forward to start building this and eventually test it out, as time permits. I was hoping if there was a software that is very flexible and that I could adjust the settings in order for my simply machine to work. I'll keep an eye out just in case.

[mdynac]

to my knowledge there is no reasonably priced software to do what you would like. 6 grand would get you there.......


I use the Linux based EMC Software to run cnc stuff, it has several GUI options and the best part of it is that it is free!!!!!
www.linuxcnc.org

I am currently working on an EDM function to make this software work with Wire EDM Machines.

I sure would like to get a copy of that camtronics edm software, just to see how it works. Unfortunately thier EDM setup is quite "amateur hour" compared to what I do in the EDM field, so I will never purchase it.

a few hints to help you out when you get your EDM up and running:

The three most important aspects of the edm process are:

1. Flushing

2. Flushing

3. Flushing!!!!

Use deionized water with some type of particulate filtering system.
and some good flushing nozzles that give good flush pressure.
The gap VOLTAGE is monitored during the edm process and not the Current in order to determine a short condition. Buy yourself a DI meter from McMaster Carr or somewhere (like 60 bucks) to monitor your water quality. Keep the water cool!!!! Chill it somehow or have a big reservior like 30 to 50 gallons.
Dirty water will cause the gap voltage to become very low, and could cause excessive wire breakage. The water is the dielectric and must be maintained at a constant, or your erosion will be erratic.


Use .012 thou wire to avoid excessive wire breaks (at least to start with, and once you are eroding okay, you can go with smaller wire.)

Ensure very smooth wire feed and maintain consistent wire tension.

Well hope this helps, good luck!!

michael

[pgruendema]

Hi Mdynac:
I have spoken to Ray Henry about the matter of using EMC(1) to drive a wire EDM. He assures me that this has been done but i didn't press him for the details as i wasn't pursuing Wire at the time. My present interest is in using EMC(any version) to drive a 3 axis ram type edm. Have you checked in with Ray or any of the other developers of EMC to get their suggestions for edm'ing with emc?
If i recall it correctly (don't count on that!) there was some kind of external interface which monitored gap voltage and that gave the signal for emc to advance to retract.
pgruendema

[mdynac]

well we are just starting work on an emc2/edm function, however it is intended to be a wire edm control. and far from finished.
Ray Henry's solution is explained on the wiki pages at www.linuxcnc.org



there are many different sinking edm functions out there, tho none are cnc.

[medved]

I have been thinking of writing software for Wire-EDM too (in dos qbasic) to be able to experiment with.
Well it might be too complicated project for me, but i started by adapting free (G-code to)plotter programs form Luberth web site.

I would like to ask Mdynac some more about flushing.

1-if wire is going to be retracted sometimes - there is danger of backlash, when you change the direction, you can loose the precision - unless you have LINEAR position encoders to track position

2-i use stepper motors, one step being 0.01mm - is this gap too big?
This way i am LIMITED to 0.01mm increments, so i suppose there must be many sparks done during one step(i could imagine another system, if i had 0.001mm precision, there could be many steps for ONE spark)

FLUSHING DURATION:if one spark cycle can take 0.01second or less to finish, the FLUSH cycle to finish and settle can take 100 times more, 0.5second or more - if i make flush every spark or even every 10th spark, than it could take days to make small cut into metal.
How do big machines solve this problem?


Why it there such secrecy about power supplys for EDM? There are books advertising they help you build one, but to me it seems their only goal is to sell their book and to prevent spreading free information.

[mdynac]

I would like to ask Mdynac some more about flushing.

1-if wire is going to be retracted sometimes - there is danger of backlash, when you change the direction, you can loose the precision - unless you have LINEAR position encoders to track position

well if you have backlash, you have a problem. On the edm's that i service, backlash is so minimal that (we use a laser interferometer to measure it) it is comped in the software....unless you are making parts for the space shuttle, i see no backlash problem....just make sure that your mechanical setup is tight.... most edm's use closed-loop servos with rotary encoder feedback for positioning, some use steppers....i prefer servos.


2-i use stepper motors, one step being 0.01mm - is this gap too big?
This way i am LIMITED to 0.01mm increments, so i suppose there must be many sparks done during one step(i could imagine another system, if i had 0.001mm precision, there could be many steps for ONE spark)

.01mm = .000394 inches....should be just fine, most real edm's have 50 millionths resolution. an edm machine does not spark, step, spark, step.... it moves forward in relation to the gap voltage. low voltage = slow down, high voltage = speed up. backup occurs when the voltage drops too low (a preset value) and short is imminent.




FLUSHING DURATION:if one spark cycle can take 0.01second or less to finish, the FLUSH cycle to finish and settle can take 100 times more, 0.5second or more - if i make flush every spark or even every 10th spark, than it could take days to make small cut into metal.
How do big machines solve this problem?

first of all in wire edm, flushing is constant, the modern prefered flushing method is to completely submerge the part in DI (fill the tank) then the upper and lower flush heads inject high pressure DI into the contour at rates up to 140 psi.......the flushing heads are very important, they damn near sit right on the part and inject the high pressure DI into the cut, from the top and bottom....the heads also incorporate the power feed contacts, which tranfer the spark to the wire, and the wire guides, which keep the wire aligned.....if your are going to homebrew a wire machine, much thought should go into the flushing heads.....



Why it there such secrecy about power supplys for EDM? There are books advertising they help you build one, but to me it seems their only goal is to sell their book and to prevent spreading free information.[/QUOTE]

i have no idea why there is secrecy to edm power supplys.....i've seen them all Sodicks, Agies, Japax, Charmilles, Andrew, Elox, Fanuc, you name it, from old to new....no big deal...just another cicuit, and nothing special....any benchtop pulse generator or function generator can deveop the spark pulse, you just have to amplify it up to usable current levels....
hell, some brand new $80,000 machines use a 555 timer as a source.....

good luck

[medved]

first of all in wire edm, flushing is constant,

Hi

I dont understand how can continious flushing work.
If i understand pictures and explanations of wire-edm process, in the gap a delicate process takes place, first iones are bulid up and then plasma channel opens. I imagine if there is flushing(medium is moving) during this buildup, then plasma channel (and its prerequisites - ions) is actually flushed away , and there will never be any spark, or - one will need much much higer voltages to make a spark(compared to situation where dielectric medium stands still and channel building isnt disturbed by flushing).
This is why i supposed one must wait for flush cycle to finish, for dielectric fluid to calm down before continunig with next spark.
Could you explain, how edm can work in continious flush mode?


And one thougt about filtering dielectric(water in this case).
There are magnets(?) to prevent limestone/carbonates to settle on household plumbing.
Could we use such (electro)magnets to filter-capture wery small particles, to use it beside mechanical filtering of flushing fluid?

thanks for help

[mdynac]

*

Hi

*I dont understand how can continious flushing work.
If i understand pictures and explanations of wire-edm process, in the gap a delicate process takes place, first iones are bulid up and then plasma channel opens. I imagine if there is flushing(medium is moving) during this buildup, then plasma channel (and its prerequisites - ions) is actually flushed away , and there will never be any spark, or - one will need much much higer voltages to make a spark(compared to situation where dielectric medium stands still and channel building isnt disturbed by flushing).
This is why i supposed one must wait for flush cycle to finish, for dielectric fluid to calm down before continunig with next spark.
Could you explain, how edm can work in continious flush mode?

http://www.manufacturingcenter.com/t...99/1199edm.asp

----the above link, explains some more...

----the flushing accomplishes several things....

1. it is the dielectric between the wire and the work, we maintain this at a very low conductivity...

2. when the pulse finally arcs thru the DI, the flush then removes the edm'ed particles out of the cut

3. it cools the wire and the work

if the flushing did not do these three things, you will be breaking a lot of wire.....

the pulse rate is variable up to as much as 100,000 times a second (100khz) so all that talk of ions and plasma (which does happen) it has plenty of time to do it's thing.... dude, in my 100khz example, ionization occurs 100,00 times a second (it don't take very long!!!!) you are talking about the molecular level of edm, i am talking about usage....if the molecular stuff didn't work, edm wouldn't work. the edm process is still working with constant flush.....i suppose if you forced di into the cut at 10,000 psi you might have trouble......

again, the 3 most important parts of edm is flushing, flushing, flushing.





And one thougt about filtering dielectric(water in this case).
There are magnets(?) to prevent limestone/carbonates to settle on household plumbing.
Could we use such (electro)magnets to filter-capture wery small particles, to use it beside mechanical filtering of flushing fluid?

---standard 3 micron paper filters do the job just fine, some systems use diatonamous earth filters....paper filters are easier....

---you can use magnets, couldn't hurt, i know of no one using that method besides there is a lot of non conductive junk in the water also.....particulate.....






thanks for help

a