Precision in an older manual Sinker EDM

What sort of precision is expected from an older (2000 ish) manual sinker EDM machine?

Looking for a small, yet capable machine to make carbide insert tooling for a cold header machine and
have come across several in the 2-5000 dollar range. They are older and used. The precision we are looking for is within 0.003mm. Not sure it can be achieved without EDM` it to oversized and then lapping / grinding it to spec.


Inputs on this?

The edm pits will be deeper than the tolerance you want to hold, so yes, you must find some other way to finish the surface.

The depth is not critical, diameter tolerance is.

Anyway, cannot depth be compensated somehow? Be running Z x amount higher? Will the depth
change from part to part?

As you might expect. We do not have experience in EDM. But we are looking into it.

Kjeksen said:

What sort of precision is expected from an older (2000 ish) manual sinker EDM machine?

Looking for a small, yet capable machine to make carbide insert tooling for a cold header machine and
have come across several in the 2-5000 dollar range. They are older and used. The precision we are looking for is within 0.003mm. Not sure it can be achieved without EDM` it to oversized and then lapping / grinding it to spec.


Inputs on this?

Click to expand...

Tungsten carbide is difficult to edm.

You would be better off finding a machine that is full CNC IMO.

GENERALDISARRAY said:

Tungsten carbide is difficult to edm.

You would be better off finding a machine that is full CNC IMO.

Click to expand...

Can you explain furhter on why tungsten carbide is difficult to EDM?

Hi Kjeksen:
0.003mm total deviation from nominal is a very ambitious tolerance for any process and is going to be very difficult for you to achieve with sinker EDM in carbide even on a high end modern machine.
There are a number of reasons why; the main one that is relevant to you is that the electrode wears as it burns so you need to have several to achieve your final geometry.
Just making all the electrodes exactly the same is not easy, so a common strategy is to make the trode substantially undersized and orbit it out to achieve control of your cavity dimension.

However you still have to measure your part ideally while it is still in the work tank and can be orbited again if it is still too small after the normal orbiting cycles.
The cut debris from EDM, especially carbide is extremely abrasive so you can't just push in a gauge pin to check size (not that pushing in a gauge pin is a reasonable way to measure at these tolerances anyway).
So not only do you have a process control problem, you have a metrology problem too.

If I were you I would ask for quotes from EDM shops in your area.
Expect to be shocked at the number of NO BID's you receive and the outrageous prices for those who do quote.
This will tell you the difficulty of the task.
It will also tell you that a used manual sinker for 5000 Euros is not going to be even remotely adequate to control the burn to your stated tolerances.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

Kjeksen said:

Can you explain furhter on why tungsten carbide is difficult to EDM?

Click to expand...

I am no expert at it, but I have done it once or twice. But no I cant really explain it. I am sure if you google it someone smarter than me will know.

1 It is picky on the electrode material you can use

2 the electrodes you can use without DC arcing, burn up like snowballs thrown at the sun

Thank you all for the useful input so far!

What precision is expected from a manual Sinker EDM?

From what I have read, tungsten copper is a good electrode for tungsten carbide.

What it all boils down to if we need a CNC grinder aswell to achieve repeatable parts within very tight tolerances.

Hi again Kjeksen:
A manual sinker will be as accurate as the ram motion is.
These machines run a ram (which holds the electrode) up and down on a slide and control it with a hydraulic piston and a Moog servo valve on the really old machines, or a ball screw and a stepper motor on the slightly less ancient ones, and a ball screw, a servo motor and driver on the modern ones.
The condition of the ram, the condition of the screw, and crucially the construction of the ram motion system determine how consistently it will point to exactly the same spot every time it goes up and down.
Since it's a cantilevered beam (supported only at one end), any wear in the mechanism affects its accuracy radically.

A good quality new machine should be able to position its ram repeatably within a micron or so at full extension.
My old (1983) Hansvedt SMB 150 (ball bushings on two round posts) couldn't re-position itself within 50 microns any more, and my newer one (vintage 1996) with a ball slide can do about 4 microns at full extension after 25 years of running.

So best case there's almost half of your tolerance taken up just with uncertainty about where the ram actually is compared to where you're pretending it is every time it strokes down to burn.
With jump flushing, it will cycle several times a second and on a linear motor machine several hundred times a second.

Positioning precision on a manual sinker is as good as the Acme screws that position the table together with the yaw, pitch and roll common to all machine tools, especially those with cast iron slides and manual screws.
Again, a good machine with a good readout on it will be reliably able to position within 10 microns but a beater may be miles off.

Thermal effects dominate in this process, and the presence or absence of a chiller (presuming it's working correctly) may influence the precision by an order of magnitude more than positioning precision.

The variability of the sparking process itself depends on the consistency of the generator in its ability to make identical sparks as the electronics warm up and is typically a minor effect.

So realistically speaking the challenge is multi factorial from the point of view of the machine and its condition, but overall, with a manual machine, I'd be most happy if I could produce my job within 25 microns on a typical used manual sinker of vintage 2000 or newer.

On a brand new, super, top of the line Makino, in a climate controlled facility, cost no object, I would be happy to be able to hold 2 microns for position and size assuming my trodes are 10 times better, my trode holder system is 10 times better and my metrology is good enough to interrogate my developing parts in the micron range while still on the machine.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining