Titivating a Wire EDM Machine

[awemawson]

This mornings job was to refresh my memory as to how the two wire guides are set to be vertical. Fortunately the book of words 'sort of' covers it - however their procedure only covers one axis (which they don't point out) so it needs repeating with the jig (see later) in two places.

However firstly I had to re-thread the wire - it's always been a pain to thread but today I developed a couple of techniques that do help somewhat.

Firstly, to get the wire through the sliding seals without concertinering it into a zig zag I used a length of thin nylon tube. Slide wire down tube, poke tube through seal into working area of machine, pull off tube. Simple and easy.

Secondly getting a 0.25 mm wire threaded into a hole in a sapphire that is only microns bigger - it's fun! However if instead of cutting the wire with (very sharp) cutters, you break it. It must neck down before snapping and produce a tapered end that threads in far more easily .   It does however dig into your fingers somewhat ! I presume that cutters must actually distort the wire into a chisel shape that flares a bit at the edges.

[awemawson]

Now fortunately my machine came with the 'verticality jig' so it was just a case of (once that bally wire was threaded) plonking the jig onto the holding frame, adjusting the X & Y position so that the wire glances the jig, then tweaking U & V to get the same degree of light flickering on the top lamp as is displayed on the lower one. Once this is done, the U & V co-ordinates are used as offsets internally so that this position is regarded as zero for them and the wire is vertical.

[awemawson]

So of course in my excitement at getting close to finishing I broke my good resolution not to water it up before painting - had to try it out !

So as I type this it is cutting out a 20 mm square in a bit of 1/4" mild steel hot rolled plate just as a test. It's so far cut a length of 23 mm in 15 minutes - nothing optimised so far - just playing.

It's not the most photogenic process - the covers and water gets in the way but here are a couple of pictures - the first through the covers, and the second with the splash covers off (briefly). Not the best but hard to get in there to take them !

[awemawson]

So 82 minutes after starting the cut we have cut 85 mm of cut length, and amazingly the part has come out pretty well bang on target size 

No doubt cutting speed can be increased by tweaking, but by contrast with how it was before I retro-fitted the lower wire contact it's quite fast. I previously had to slow it down to avoid wire breaks - this time none  For the same thickness material (1/4" it would have cut pretty much the same rate independent of how hard it was.

Note the use of a rare earth magnet to stop the cut off part sagging and fouling the wire as the cut finished. Amazing amount of debris produced from a small cut. It's mostly very fine steel particles, but there will be brass from the wire mixed in. I may experiment with a magnetic particle separator to save the main filter some work.

I need now to get some time in playing with it to learn to use it 'comfortably' rather than always being on the edge of my seat !

[awemawson]

What expansion port(s) do you have in those SFF machines? If they'll take PCI cards, you can fit one (or two) of these: http://www.ebay.co.uk/itm/261083309079

And then you've got 4-6 serial ports (including the 2 on the machine)....

I'll certainly explore that suggestion Ade - thanks

[awemawson]

Yesterday, just to get some machining experience on the EDM, I set it cutting out a pair of 18 tooth 1.5 MOD gears from 1/4" mild steel plate. The gear profile was defined in FeatureCAM and down loaded to the Fanuc as G code.

Came out bang on as per the defined gear  (which I should have given a bit more tip radius to !) but it makes what you tell it to make 

Today's exercise is to get my head around optimising the cutting conditions to speed the process up. Those gears took best part of four hours EACH so not an ecconomic way unless it's in something very hard or is an exotic tooth profile. (Those gears have a cut length of 184.6 mm and took 3 hours 45 minutes and 26 seconds each)

[awemawson]

So first results of tweaking: That 20 mm square a few pictures above now cuts in 47 minutes as opposed to 85 previously, so almost halved cutting time. Don't yet understand all the parameters . . .but I aim to 

[Kjelle]

Weellll... This has been a rather captivating story so far... 

Kjelle

[Pete W.]

Hi there, Andrew,

I think those gears are impressive.                    

3 hours, 45 minutes and 26 seconds is a far shorter time than you'd wait for the Postman if you bought them from 'outside', especially if your time of need was 17:00 on a Friday evening!

Here's a question, though, which machine/process would you use to make the central bores and how would you register with the gear-tooth PCD? 

[awemawson]

Thanks chaps - I must admit to a certain self satisfaction at the results! It's also nice to see some people are actually reading this drivel. 


Pete, If I was doing it 'properly' I would have drilled a hole, then threaded the wire through it and got the machine to 'find the centre'. Then I'd have cut the bore accurately on the machine, then cut the wire, move to the outside, re threaded the wire and cut the gear. It would then have been accurately centered. But you know the threading issues.

[Spurry]

Thanks chaps - I must admit to a certain self satisfaction at the results! It's also nice to see some people are actually reading this drivel. 

Andrew
You are too modest. A brilliant thread, very well written and photographed. Keep up the good work..........pleeease.
Pete

[spuddevans]

I've been watching too, and very impressed with your efforts and results. 

Tim

[awemawson]

Thanks for the comments Pete & Tim.

I decided that I have to get to grips with the wire threading issue. It breaks down into two areas:

a/ The lower guide is in a confoundedly hard place to look at

b/ I'm trying to thread a 0.25 (and possibly burred in cutting) wire through a hole that is allegedly 0.255 mm diameter

Now there's not much that I can do about a/ (I think most modern machines run their wire downwards which would help) but I did a bit of research and asked a few questions on a professional forum about the actual wire and it's cutting. Apparently in early Elox machine manuals they advocated heating the wire with a cigarette lighter, and stretching it to break. Another manufacturer embodied a pair of  pillars with a voltage between them - stretch the wire across, anneal it and stretch to breaking point.

I tried the cigarette lighter method - it works and produces a thinned wire, but I'm not a smoker and the lighter is hard to manipulate as the wire is held taut enough to break. Strangely I'd thought of the electrical method independently so I thought that I'd do an experiment.

Digging out a 12v transformer, I rigged a pair of screwdrivers in my vice, holding them with non conducting soft jaws. Drivers about 3 1/2" apart, 12v on the wire - touch it on and POP  instantly the wire melts - volts too high.

So out came the ancient variac that I keep for scaring Health and Safety people  A bit of experimenting came up with 3 volts being enough to anneal the wire. Takes a certain technique to take it off before it breaks, as if it fuses rather than breaks it forms a ball on the end.

Anyway from this crude test I've shown that I can taper the wire from it's original 0.25 mm to 0.13 mm - half it's thickness. It is possibly worth making up a little fully insulated box with a couple of pillars poking out, that I can use to stretch wire ends. (or I could stick with a lighter!)

[dawesy]

Excellent project Andrew.
Really enjoying reading it so far even if I don't fully understand some bits (I'm still a newbie )

[dsquire]

Andrew

It is allways a pleasure to read your posts and watch as you prepare retired machines for a come back. You do have the ability to get the best out of these tired old machines. I am looking forward to seeing what you will be cutting with this when you have it completed..

Thanks for sharing with us Andrew.   

Cheers 

Don

[awemawson]

I spent today running hard wiring up in the gods from my three cnc machines back to my desk in the workshop. The replacement USB to 4 x COM port box arrived by post this morning and seems much more stable than the one that I returned (for which the paypal refund came in this morning  ) The thermo-plastic that they are making Sub D connectors from these days seems pretty poor quality. If you solder a male 25 way up the stuff melts so much that the pins point in all directions - had to plug the male into a female to keep them straight. I've been soldering these things for the past 45 years and not had this issue before 

Then fighting a loosing battle trying to find a suitable DNC program to link them up. CADEM make what seems the most appropriate, but they are still failing to return contact - probably don't want any more customers 

Tomorrow brings filling pot holes in the drive with concrete so probably not much will get done then 

[dawesy]

Think everything is made by accountants and not engineers now Andrew. Good idea for keeping the pins straight though. I'll remember that one ;
I never understand companies that don't respond to customer questions. Had similar from pace radiators recently. The replied asking for sizes ( which were in the first mail ) I replied with a diagram but had nothing since.
Hope you get it sorted.

[awemawson]

Time to start slapping some paint around 

I've had to make several panels which were never painted, and the previous owner has made a hash of a brush finish on most of the original paintwork. Now my RAL 7044 silk grey paint and etch primer has arrived it's time to get on with it.

Weather forecast is predicting a dry week end, even if it is raining as I write this. So hopefully I can glass blast the old paint off tomorrow and start spraying. I doubt it will be fine enough for outside painting. So I started this morning clearing out the welding shop to hopefully give me enough space to blow some primer on as soon as the panels are de-painted. Glad I put all the machines in there on wheels 

Next job was to remove the panels - the machine looked decidedly naked without them !

[awemawson]

There are a total of ten panels needing cleaning and painting, ranging from large and heavy down to quite small.

I would really like to be able to strip the paint on the remaining horizontal surface surrounding the table as that too has been badly brush painted. If I could find a satisfactory way of stripping it I'd roller paint it to avoid over spray all over the work shop.

If only the original Nitromors stripper was still available it would do it easily, but the EU have banned all the good stuff in it that actually works 

Any suggestions for stripping selectively without throwing stuff all round the workshop would be appreciated.

[nrml]

Can I suggest trying screwfix's ''No nonsense paint stripper'' along with a tungsten carbide bladed paint scraper. The combination has worked very well for me. You need two thick coats of paint stripper left on for an hour each (as per the directions) before scraping. It seems to be the only paint stripper that actually does what it says on the tin and it costs only £3.20, so you don't really lose much if it doesn't work for you.

BTW Thank you for all the excellent threads you have posted. They have all been most interesting.

[awemawson]

Thanks for the comments NRML

I've managed to locate some stripper with Dimethyl-chloride as it's base, and that is expected on Tuesday. Meanwhile this morning I succeeded in glass blasting all the panels that needed stripping, and got a coat of etch primer on them outside as the weather stayed fine.

No pictures as my iPhone decided it's battery was flat despite telling me it was 100% only two hours before - nothing to do with 'upgrading' to ios 8.1 yesterday I'm sure 

One of the panels that I made a few years back, the one with a rectangular tunnel in it to give access to the water flow gauges, I had given a coat of primer to back then from a rattle can. Forgot that and sprayed it with  my etch primer and of course it reacted and gave crows feet every where. Ended up using paint stripper (B&Q ultra safe ultra useless) which as it was only fresh came off ok'sh . Re-sprayed and it along with the other nine panels will get a coat of  semi gloss tomorrow.

[awemawson]

So today I got a coat of top coat on both sides of all the panels and two on the outsides. If you look at the first picture and see the nettle bed behind the panels you'll see we had an early frost - no actually it's fine glass grit from the blasting yesterday. First rain'll wash it off.

Oddly the 'grey' etch primer is distinctly green

[awemawson]

With rain threatening, and the paint having hardened significantly I decided to bring all the panels indoors to fully harden off for a few days whilst I attacked the main carcass of the machine.

I'm waiting for some decent paint stripper to arrive, but I thought I'd see what a hot air gun could do on those bits where it's safe to do so. Painfully slow  The metal takes the heat away too soon. So I brought out Big Bertha the Blow Lamp. Now it's not appropriate to use this everywhere due to the proximity of wires etc, but a lot of the surface area can be cleared. It worked a treat. Took about 20 minutes to do the bit that you see in the picture. The horizontal surface was far easier than the vertical, and I'm sure it's that over the years water got under the original paint as it just stood there - it looks a bit rusty - probably cleaned up somewhat when the previous owner brush painted it. It'll need a bit of rubbing down and primering but it's quite pleasing progress.

[awemawson]

This morning I managed to finish stripping off the vast majority of the paint on the machine carcass horizontal surface using the blow lamp. Still to remove the 'kerbs' to allow me to strip up to the edges. Still to use chemical stripper on some fiddly bits and also the edge of the work area 'tray', which is stainless steel, so if it cleans up ok I won't paint it.

The surface has obviously previously been stripped back where the rust shows but needs quite a bit of preparation prior to painting. I may risk a 'strip and clean' disk now 99% of the material has been removed as hopefully it won't make too much mess with the little that remains.

I remembered that yonks ago I had bought a 'touch up' spray gun for a project that never happened, and I've never used it. I'm going to experiment and see if it is feasible to use it without stuff going every where as a sprayed finish is very much preferable to a rollered one.

Then it was time to go to town and sort my iPhone 4S out. Upgrading to ios 8.1 had resulted in it's wifi refusing to work even after a 'factory reset'  After much faffing about I now have an iPhone 6 - an exasperating experience

[awemawson]

Today I made a great discovery 

I had thought that the top that I've been stripping was part of a welded structure forming the machine carcass. As I was working out how to fix masking for spraying, I got to wondering how the heck they put the machine together in the first place Surely it MUST have been built from the bottom up ? The top MUST come off. Sure enough I found 8 concealed screws.

Huge benefit in that being able to remove the top meant no masking, and I could glass blast it to get it decently clean. Down side was that quite a bit else had to come off. The 'curtain tracks', the arm, and the work tray. Actually only took 20 minutes as I've had them all off before.

Net result is I've been able to properly clean up the metal work (including a 'water inlet panel' that I'd passed over last time.)