I see a lot of edm plans on ebay that say you can build one in a weekend. Every once in a while I notice they get a bid. My question is this, has even one person, ever completed one, and actually edm'ed anything? The whole thing kinda reminds me of "galaxy quest" where the ship communicates with Tim Allen on the barren planet, and asks him if he can put together a makeshift lathe to save himself from the stone giant that is about to crush him.
I would be cautious of these. I don't believe you could make a good machine in a weekend. Even if it works, it's going to look like it was built in a weekend. It sounds like those "turn your CRT monitor into a video projector" plans.
But if anyone has made an EDM machine from these plans and found it to be good, I'd be very interested to hear.
There were plans in an old Popular Mechanics/Science/Electronics that you could easily throw together in an hour. You used an existing drill press to manually feed the tool into the work. It involved working with AC line voltage with no isolation transformer, and used a light bulb as a ballast. You adjuted the feed by how bright the bulb lit up. Very kludgy, very dangerous, but I'm sure it worked.
If I absolutely had to burn a tap out of something and had no other choice, I might give it a try. I used to keep my eyes open for a good sized isolation transformer to try it with, but all of the local electronic surplus places have pretty much dried up. I could probably get one at the MIT flea market, but I never get over there any more.
Whether this is the sort of thing the guy on eBay is selling, I have no idea.
Years ago I had a friend built one from plans in Popular Mechanics, or some such source. It worked well enough that he discovered why one DOES NOT submerge an arc in water (the plans specified Kerosine but my mate "knew better").
Bruce-in-Bangkok (Note:remove underscores from address for reply)
There are other plans for cheaper systems that use manual ram positioning, possibly a vertical milling machine and electrical [not electronic] power circuits using one or two lightbulbs and some large caps to control the spark [120 hz with 60 cycle in and a bridge rectifier. It may be possible to get 360 hz with three phase.]
Home Shop Machinist has had several articles over the years on a home made EDM. For reprints click on
HSM may have some back issues:
Inexpensive EDM, An Gregrich, Arnold Subject: EDM Issue: Volume 15 Number 5, Sep 2002
Sequel to the Inexpensive EDM, A Gregrich, Arnold Subject: EDM Issue: Volume 15 Number 6, Nov 2002
An easily-built system is described in Popular Science, March
1968, pages 149-153. However, it is electrically dangerous from a shock standpoint, and suggests using kerosene as the dielectric which is a strong fire hazard. This machine was reworked a bit in the January 1991 issue of Home Shop Machinist (HSM), with a few notes in March 1991 (page 11), July 1991 (pages 4-6), May 1992 (pages 10-11), May 1994 (page 44), and May 1995 (page 19). EDM was described in February 1993 in Electronics Now (pages 79-81) along with a list of EDM resources, with a followup in the May
1993 issue (pages 78-80) that references the Popular Science and HSM articles. This machine requires "manual" feed of the electrode. Please note that non-flammable dielectrics should be used, and are available from supply companies such as MSC (I have also heard of using de-ionized water as the dielectric).
Those are for a really simple, probably dangerous to the unwary user, tap burner, that would be useful for an emergency job where you needed a hole or a tap burned out (removing broken taps!) or such. They work , but are not EDM in more than the basics.
Nowhere near the clean cutting and precision that a machine that would regularly be called an EDM machine, would be able to put out.
A fellow named Langlois ran a multi-part series in Home Shop Machinist, about building a capable EDM machine. NOT a weekend project!
I'm pretty sure the series was reprinted by Village Press, the publishers.
I kludged something like this back in the early 70's. I used an isolation transformer and a voltage doubler circuit with a light bulb to limit the current. The light bulb works well as the resistance changes with current. This was done with a friend and used a wooden dowel in the chuck. Feeding with the drill press is pretty poor. We used the stop and then leaned on the feed to get a small enough motion. We used kerosene, which is not really a big hazard as the sparks all are under the surface.
So I would say very kludgy, but not very dangerous. And it did work but slowly. We could burn square holes in HSS.
Actually it's what is known as a "relaxation oscillator" circuit and it works off of DC, and spark frequency doesn't really depend on line frequency in any way (although the ripple will modulate it. Essentially you slowly (that being a relative term) charge a capacitor through a resistance, and then when the voltage exceeds the dielectric strength of your oil, you rapidly discharge ten capacitor through the gap via the spark. As I recall from when I built one, I think the pulse rate when cutting right was in the order of a few khz.
And, the author, Ben Fleming, now offers a PC Board for the major circuit components - by perusing the above yahoo group, you'll prob'ly find that the Fleming design is somewhat better than the Langlois one, and several cuts above the wallplug / lightbulb variants.... The book is, IMHO, worth the money /mark
As one who has built both machines.....the one in Popular Mechanics decades ago, and the Robert Langlois' excellent stepper driven ram, I can attest to the functionality of both.
Each set-up uses the "Russian" circuit, which is better known as a relaxation oscillator working as described above.
I was one of the group that worked with Robert to build his machine for each of our home shops. It is a nice machine...I purchased Roberts populated circuit board and did the rest myself. I use distilled or ionized water with brass electrodes as this combination seems to work quite well. And I keep a jug of this water on hand just in case I break a tap late at night. Unfortunately I haven't broken a tap since I built this machine. :-))
It also will burn holes in tungsten carbide inserts or holes of any shape in hard conductive material.
Using readily available brass tubing in square, rectangular, or hexagon cross section, from hobby shops, these shapes can be burnt into work pieces easily.
The PM set-up can indeed be dangerous without an isolation transformer! And I sure didn't have one! I had an autotransformer and used a 600 watt projection bulb as the series resistor. This worked very well when using baby oil as the dielectric. The feed, of course, was manual on the quill feed which I had insulated with a slip- over plastic tube (which is still on the feed lever today!). This setup allowed me to burn out a broken tap in an hour or so.
Naturally the automatic set-up with pump circulation and stepper motor feed is much nicer.
For accurate work it is necessary to either filter the dielectric really well, or use new water. Any conductive suspension will increase the overburn or, when badly contaminated, lead to parasitic discharges on the surface of the work.
The relaxation oscillator, while simple and reliable, is rather slow in operation. I would sure like to upgrade my machine with a more modern switching supply if any of you can point out a suitable design.
In conclusion, my EDM machine uses a ram slide salvaged from an old office type writer carriage slideway, with slight modifications. The tank is a stainless steel rectangular pot from a restaurant steam table...I bought 2 and use one for oilchanging our fleet. The tank is mounted on a small X-Y table (Atlas??) so that features can be accurately located.
One thing that hasn't been clearly articulated in this thread is what kind of EDM is to be built, as there are at least two drastically different types - die sinker EDM and wire EDM with very different capabilities and applications.
I've seen various references to and pictures of home built die sinker EDM machines, but almost no references to home built wire EDM machines.
Old EDM machines can be had for a song. No longer useful for industry and very few hobbyist after them. I have both a wire and a sinker EDM. Fun toys. Look on eBay or, better yet, at machine shop auctions
A wirecut EDM would be a tricky project to do at home. First, unless you're only going to cut straight lines in one axis at a time, it needs a pair of coordinated stepper- or servo drives for the table. Second, and again, unless you're going to only cut straight lines, you need CNC or a tracer to follow a pattern.
Not the controls, though. Not only do you need positioning, you also need extremely precise servo action. In other words, when the wire has to retract due to low voltage (that's the servo action used in motor-driven EDM axis drives), the wire has to retrace its path. If it was moving in a direction, say, 37 degrees from the X-axis, it has to know to retract by one, two, or three pulses in each axis simultaneously or you will get an arc, which will break the wire...or you will get a mechanical hangup, which will break the wire. That's not off-the-shelf software.
EDMs get along especially well with steppers but it's a b*tch to get them to work right with DC servos. That's because a cheap, hobby-level DC servo drive doesn't know it's supposed to accelerate like crazy through the first two pulses. Steppers have max torque at zero rpm, so they're a natural for the job. They sure make traverse rates slow and noisy, but that shouldn't be a problem with a hobby machine.
As for water, wire EDMs work well with oil dielectric if you aren't in a hurry. Some of the most precise wire EDMs use it today. The first wirecut machines, pre-CNC (Andrew Engineering and Japax, 1959 and 1960, respectively, sold as attachments for ram EDMs) used oil. So did the first Sodick machine in 1980. Since then several others have used oil.
If you want water, take a look at the older deionizers. They're just little water softeners. As for filtration, one of the better aftermarket filters made in the '80s used four rolls of toilet paper to filter the water. d8-)
By wire "feeds" I assume you're talking about guides? They use diamond today, but some early ones used carbide. That should be OK for a hobby machine. Guides were mostly round until around 1978 or 1980, except that Agie had that three-sided deal with one moving side. Plain round holes work fine until you try to split tenths. And you aren't going to split tenths with anything you could build in your basement, anyway.
All in all, a challenging project, but the mechanical parts are manageable.