View Full Version : EDM [Electric Descharge Machining on old machine]??

01-08-2005, 09:39 PM
Is there a EDM Forum here??

Have posted this on practical machinist forum, but sometimes no response for
many days or weeks, so I thought this way I have more luck!

I use just regular copper...I do hardly any sinking on my old agie 1968 model,
but now I have to edm 8 cavities, something like 3/4 wide, 2" long and .2
deep, with a radius on the bottom.

I was thinking, mill first, rough out a bit, then edm.
To get down to the bottom with sharp edges, I have to redress the shape about 3 to 4 times, that is what I hate the most!

Is there a electrode material, what stand up going down in one shot?? [on my oldie]

Someone mentioned tungsten copper??
Does someone has experience with this, using old Agie like mine??

Another possibility, farm it out and get rid of the sinker??

Any help appreciated!

01-09-2005, 01:06 PM
I would use graphite. Poco has some really good selections. Now what material are you sinking? I would definitally mill first. I usually leave .005-.008" of material thenjust clean that off. I have an older Eltee Pulsitron. When possible I will mill everything I can on the mill, then only pick the corners. Or detail areas I can't mill.

01-09-2005, 01:54 PM
Thank you Kmed,
By how much would graphite be better, in terms of wear?....20%, 50% or better?

I have experimented ones, many years ago with harder beryllium type copper, or whatever it's called, also with graphite, but it did not appear to make any difference.
Actually, I just tried yesterday again with graphite again....had some laying around when we had to
machine some rings, same thing, wear was still about 30%

But then, I don't know, if I had the right grade???

Does it make that much of a difference in what grade you are using??

Material to be edm'd is QT 100 Plate, similar to 4140H.

I'm thinking, to mount all 8 electrodes on a plate, that way, the pockets will be done at ones.
With today's machine, one could get away with only one electrode,.... so what they say?

Do I miss something, in the setting of the machine???

I always make sure, that I use just enough power to get it to spark nicely, keep the amps as low
as possible, and use a coarse setting when roughing. Is this the Idea??


01-09-2005, 03:52 PM
I use the EDM-1 for roughing and EDM-3 for finishing. I also stay away from magnetic work holding as I found that my flushing cannot flush out the burned off scrap and get more wear on the electrode, and burrs. This is due to the burned off scrap being magnetized to the work piece. What radius is needed at the bottom of the part? Maybe a roughing electrode and a finish electrode is what you need. I would still rough machine as close as you possibly can before moving to the edm. Berrilium copper is a PITA to work with and would probably be easier to just machine multiple graphite electrodes rather than the berillium copper. Poco also has a graphite they use to edm copper with. But I believe that is highly dependent on flushing of the work area. Maybe use a higher pressure on your flushing and see if that makes a difference. for more info you can go to www.poco.com.

Tom Rogers
01-09-2005, 05:17 PM
It is possiable to edm in a "no wear" setting. This is more like a minimal wear setting. It uses reverse polarity, electrode positive, work negative. With out knowing how accurate you need to be I can't say if this will work for you. You are on the right track by roughing as much as possiable before doing any burning. If conditions are right only a total of 10 electrodes would be needed for 8 cavities. 3 hits each, i.e. 3hits for the first cavity, throw the first one away, (or re dress), add one un worn hit for the next, etc. If you would like to discuss this e mail me at jkupkey@mail.epowerc.net. sincerly; Tom Rogers,

01-10-2005, 01:54 AM
Thank you Kmed & Tom,

looks like no magic here!

Can you tell me the wear rate on graphite, would it be much better then copper? I have now app.. 30% on copper.

I have also tried to reverse polarity, that was when I did some carbide burning but felt that it didn't made a difference.

When flushing with too much pressure, it was actually behaving negative.

Finish is not a big issue, it's a rubber mold plus minus .002" tolerance, radius on bottom, across the 3/4" with, is about 2".

Tom Rogers
01-14-2005, 07:07 PM
Yes graphite is better than copper, at least on steel. You are right about flush pressure being a factor. Generally a pressure of less than 6 or 8 psi works best, with smaller electrodes deflection can be a problem with even these pressures. It is usually best to start low, 1 to 2 psi when the electrode is in the cut, pressure should drop to almost 0 when your electrode retracts. Try increasing in small increments till you achieve the best burn. Some times a pressure of 1/8 to 1/4 psi is best, you need an accurate pressure guage and a good manifold set up with a by pass to achieve good control at these low pressures.
A brief description of the theory on low wear burning. A setting with more than 50 % on time, and as high as 90% on time, combined with low flush pressures contributes to a plating of the electrode by the steel being burnt. Your are burning on the verge of a dc arc. A fine tuning of all burn parameters can lead to wear rates as low as 2 to 5%. It has its draw backs, very fine finishes are tough to accomplish. But over all accuracy is impressive even with only 3 hits. Shallow burns can even be done in only 2 hits.
As a general guide square inches equal on time, .25 sq in 25us on time, these are only ball park , your surface finish requirements may dictate a lower value, If I understand your electrode size of .75 x 2.0 square inches = 1.5 so a value of 150 us on time would be the ideal, but surface finish would be about 150 rms, kinda rough so a setting of maybe 45 or 50 would be more appropriat with an equal ammount of off time, 50% duty cycle to start. Over burn would be about .002 and drawing about 4 amps. You will need to determine how many transistors to turn on to achieve this with your burner. I have some charts on overburn, surface finish, power required, ect, though not machine specific they seem to provide a good start on other machines, time and patience are required to determine what your machine likes, time well spent when you can do more in fewer hits, especially if the carbons are intericate.
Tom Rogers

01-14-2005, 10:17 PM
Thank you Tom for your detailed information!

I really have to get more serious with this machine and study the manual!

Just haven't done much burning, to really get "into" things!

I have made 9 electrodes, possibly I have to redress some....will see.

Customer has changed radius on bottom, from 2" to 1.096".....a good thing, I haven't started them early.

I will first rough out the center, 3/4" wide area, and not do the ends 30 deg.at all.

If somebody would like to model the shape and give a tool path in Bob CAD V 19, I will be more then happy to look at it, that would give me some hope in bobcad? and would upgrade perhaps? I think, it could be done by just machining it, after all, according to the sample part,
we can allow a 1/32R. in corner... drawback, tiny endmill!

The outside shape is the same as the dxf file shows, I will explain, how the opposite will look [Electrode] because that's how you would model it.

Think of a piece of key stock, 3/4 x 1.99 long, the ends taper at 30deg.[per side] to .219 radius center. Over the 3/4" wide, the entire length, put on a radius, convex, of 1.096, then on the vertical mill, swing your head 30 deg.and mill the ends with the face of the endmill across the ends over the 1.096R. .196" down [deep]
The total depth of the cavity is going to be .196" deep.

Naturally my electrode I have made .006" smaller per side for spark gap!

Can Rhino model this and have it machined in bobcad??
I played around ones in Rhino [demo] I think it's a good thing, ones you know how!...but all takes time, especially for the older folks!

01-15-2005, 10:03 AM
Do you have a dxf of this? I could model it for you in solidworks and give you a iges file or what ever else you want. I could also give you tool paths from solidcam, but would need to know what controll you use so I can modify or write a post to work for you. I have never had much luck using bobcad to program surfaces, but it can be done. I also havent used it since v18. Rhino would definitally model it for you.

01-15-2005, 08:28 PM
Thank you Kmed for your offer!
At the Start of the thread is a dxf file, this gives you the basic Idea, how it looks like, the rest you should be able to follow at my follow up. I can't draw it properly in bobcad 17 & 18 Version, it needs to be modeled.

But everything is just fine wow, have test milled center cavity, using a 3/16 bullnose endmill with 1/32R and burned a test piece in Aluminum....it looks good!