Titivating a Wire EDM Machine

[awemawson]

All this of course is a distinct case of ''scope creep  ''

So today was a blasting spraying sort of day. Wind a bit too breezy but I think I've got away with it.

[awemawson]

So now it's all back together and looking slick in it's new coat of paint  A bit of touching up needed in places, mainly as I got a reaction between the paint and the blue workshop paper towels that I used to separate the panels as they were fully hardening.

[awemawson]

I do have one issue to sort - if you look closely in the last picture you can see that the sliding covers that I had all the trouble joining together have warped. The original one was slightly curved but the one I made was nice and flat. Foolishly I decided to lightly glass blast them to get the finish texture on both to be the same. They warped all over the place and despite my best efforts I cannot get them to lay sufficiently flat not to rub on various bits of the machine. The easiest solution is to make another pair and that is what I've decided to do. But this time they'll get sprayed 'as is' 

Now it's back together and offers the prospect of being used in earnest I decided to attack the wire threading issue head on. Two issues - visibility and tapering the wire end.

The first issue is that you cannot see the 'funnel' that you are aiming the end of the wire at from underneath the arm assembly. OK a 6" mirror tile with three 'feet' super glued on the back to keep it out of the water makes a significant improvement.

The second issue is that the wire is only microns smaller than the guide hole and needs tapering as discussed before. I opted for the Zippo lighter approach as it can be free standing when lit leaving both hands free to manipulate the wire. Light up, anneal the wire, and stretch to break - works a treat. I used to have a Zippo years back but needless to say couldn't find it anywhere. Ebay provided me one very cheaply and it came with FREE ENGRAVING. Had to put some poser text on it naturally 

[awemawson]

At the week end I managed to bend up the replacement sliding covers, but this time they DIDN'T get glass blasted before painting, which was what distorted them last time 

I took the opportunity to slightly redesign them. The stationary  'kerbs' that restrict their movement are 8 mm deep. The covers slide on top of these kerbs and have a down turned flange that stops them passing the kerb. The originals had 8 mm down turns, which meant that not only were they resting on the kerb, but also the top of the machine and scratching the paint when they moved. I have reduced the down turn to 5.5 mm giving a theoretical 2.5 mm clearance over the top of the machine if everything is flat and true. So now (again theoretically) the covers only touch the kerbs and not the machine top. In practice so far it seems to work ok. Pity I didn't remember to re-set my corner notcher for the reduced flange - if you look at the first picture you can see the corner is more 'open' than is ideal, but never mind eh ! My 'joggler' that I made earlier came into play joggling the join between covers and worked well. I had made it originally for a 4 mm 'double joggle', but inserting strips of 2 mm aluminium on each side reduced the joggle to 2 mm very satisfactorily.

Second picture shows the slight clearance I have created between the cover and the machine top.

Perhaps now I can get down to it and start using the machine in earnest 

[awemawson]

This evening I managed to tick off another 'round-tuit' : The machine came with a 'remote pendant' - basically a box of switches and lamps on an umbilical cord replicating the axis motion buttons. But as my machine was originally equipped for automatic wire threading this particular pendant abounded with a load of extra buttons and lamps relating to features no longer on the machine. The pendant had had quite a bit of its controls removed already when I got it. Presumably this happened when when the machine had automatic function removed. I also robbed some apparently redundant buttons to get the main console functional, in some cases having to make frankenmonster switches from several others  I did at the time blank the few holes I created with grommets.

All this looked a mess, and as I now knew which functions were needed and which were redundant, having a picture of the simple control in the Operators Manual, I decided to do a 'proper job' and finish the task.

Only a case of removing surplus controls and making the wires safe with heat shrink sleeving. Amazingly as obtained there were loads of wires with bare ends just floating about. Probably their other ends are disconnected - but are they  Not good practice to leave them bare even if they are isolated elsewhere.

The pendant is quite handy when moving the table to place the wire close to the work using the Mk1 Eyeball method.

[awemawson]

More playing about today. Really this is for PeteW who asked a few posts ago how I would accurately put centre holes in those 20 tooth gears that I cut earlier. I'd been playing with gear parameters in the FeatureCAM gear generating program, as I'd not been happy with the tip radius that I'd cut last time. So I generated a 25 tooth Mod 1.5 gear with what I hope to be the correct geometry, and to demonstrate the method this time put a 10 mm shaft bore and a 3 mm keyway - all cut on the EDM machine.

A/ Having downloaded the G code to the machine I mounted the 1/4" mild steel plate in which I had drilled a 3.3 mm centre hole.

B/ Then I cut the wire and threaded it though this hole.

C/ Then using an inbuilt machine function accurately located the centre of this hole and set it as X=0 Y=0

D/ Next I inserted an M01 (Optional Stop) command between the G code cutting the bore and the rest that does the actual gear teeth.

E/ Then I set the machine running with the 'Optional Stop' button pressed. (*)

F/ This cut the bore and keyway then automatically stopped waiting for my attention.

G/ Next I cut the wire, moved the table so that the re-threaded wire was outside the blank, and set it going

H/ Several hours later I removed the gear and cleaned it up.


(*) it was at this point I realised the the button was incorrectly labelled 'Option Skip' not 'Option Stop' as it should have been, so while it was making the gear I printed out another button label on overhead projection film and fitted it.

Mild steel always comes out with light surface rust, which I usually take off with phosphoric acid - this time I used the wire wheel.

[AdeV]

Tidy!

How long is "several hours", and - when used in its original production environment - would the operator have been pushing the machine along faster?

What's the maximum thickness steel it'll cut?

Cheers!
Ade.

[awemawson]

Ade that took about 3 hours, but I know it wasn't optimised. Approximately 1 mm of cut per minute as it was set up. Theoretically it should do up to about 10 mm per minute on that 1/4" plate for a roughing cut.

It will however do up to 200 mm thick tool steel 

[AdeV]

Thats quite impressive - especially if it can maintain its 20mm/min speed in that 200mm steel....

What about ali, brass, etc.? All cuttable? I may have to start watching eBay again....

[awemawson]

If it conducts electricity it'll cut it. Speed will reduce with increased thickness - think in terms of 'area cut per minute'. They don't often turn up on ebay.

[AdeV]

If it conducts electricity it'll cut it. Speed will reduce with increased thickness - think in terms of 'area cut per minute'. They don't often turn up on ebay.

I'd be interested to see what it makes of graphite.... or diamond. If you could cut perfectly sized dead-nuts parallel graphite cylinders, you'd be mates for life for anyone building a Stirling engine...!

eBay is a code for eBay, Apex, Riley Surface World (yes, that's a real company), Peaker Pattinson, and a few others...

[awemawson]

One of its main applications is to cut copper and graphite as electrodes for 'sinker' EDM machines, I have reams of tables of cutting 'conditions' for graphite 

[awemawson]

An advantage of a machine that plods on doing it's own thing is that you can get on with other things while it does

Before I set the machine cutting another test this morning I removed its 'running hours' meter, which has been rather flaky in operation. So while it ploughed it's own furrow I dismantled it to try and find the cause.

It's a mechanical Omron 'Time Counter' model KTH-R running off 110v AC - I'd already identified an LCD possible replacement at a very modest price if it wasn't fixable but better to fix if possible. Construction is a simple plastic jacket surrounding the metal interior works, that is sprung over a couple of dimples to retain it. Pulling it apart it was pretty obvious what the issue was - the original lubrication on the gear train was very stiff and solid. Rigging a 110v supply to it, it only took a gentle spray of WD40 to loosen up the solid grease to a more suitable consistency for it to start working freely. A bit of a clean up and re-assembly, and as I type this it's been running for over an hour with no problems. It can be popped back in the machine when its finished it's current job.

[awemawson]

At long last the bespoke bellows that I commissioned have arrived - in fairness slightly earlier than the six weeks I was quoted.

So they got fitted this afternoon - hopefully I won't now have to leave a folded towel down the side of the machine where they were missing 

[spuddevans]

Looking very smart there Andrew 


Tim

[awemawson]

Thanks Tim, it certainly is a bit prettier than when I got it, AND it now works 

Still loads to learn about the nuances of the programming - quite a bit of subtlety involved to get a decent rate of cutting without breaking the wire.

I scored an ebay triumph for it the other day - nine 8 kg rolls of the brass wire - 72 kgs of wire - should out last me even at my current enthusiastic rate of use

[awemawson]

Been cogitating on the power this machine consumes - the book of words says that it needs a 7kW three phase supply, and if it genuinely is pulling 7kW all the time it's running, at 14 pence per unit that soon mounts up. So I thought that I'd measure it.

Although I have a 'spare' three phase certified meter sitting on the shelf, it's not the easiest thing to lash up temporarily, as it has to be mounted so that the back is vertical. I fixed up my 'Owl' 'Intelligent Wireless Energy Monitor', using three current transformers feeding it's transmitter - one on each phase. For current monitoring and test purposes I have a 16 amp 'commando' plug wired to a cable mounted socket using 4 mm 'singles' cable that allow the current transformers to clip on to them.

Firing up the machine it was interesting to monitor it. On the back of the cabinet there is an isolator, which if turned 'on' feeds power to a three phase transformer in the de-ioniser unit, as well as to the main unit, but the main unit only powers up when the 'On' button is pressed. It's quite easy to forget the isolator and just press the 'Off' button on the machine - turns out that that transformer is drawing 150 watts just idling and not powering anything  Must remember to isolate!

Anyway the power with everything running is less than 3 kW so not as bad as I had feared - but still about £0.37 per hour just for electricity !

[millwright]

So cutting that gear cost you £1.11p in elecricity, not taking your time and material cost into account or other workshop costs (lighting, heating and air). I dont think thats bad at all. Good job theres no charge for all the time you have spent getting it back on line.
i have enjoyed the log with all its trials and tribulations. what you have achieved is a bit beyond me but has been good to follow, as was the Traub lathe.  hope they can pay for their keep now.

John

[awemawson]

Thanks for the kind words John.

No way will they ever earn their keep (other than in enjoyment) as I refuse to take paying jobs. I retired to leave that sort of stress behind. No problem helping friends out, but as soon as it becomes commercial obligations come with it. The previous owner of the Fanuc Tapecut wanted me to do some work for him and his customers, but I've refused for these reasons.

I do these machines for the enjoyment that they give me with their challenges. Pig headedness really, as if someone says something can't be fixed I rise to it!

[Will_D]

And "TG" for people like you!

[awemawson]

And "TG" for people like you!

  

[DavidA]

TG = Thank God.

I think.

Dave.

[dsquire]

And "TG" for people like you!

  

TG stands for Thank God. I fully agree with Will. The journey is sometimes much more enjoyable than the destination. Keep on fixing them Andrew and I'll be watching from this side of the pond. 

Cheers 

Don

PS: Looks like DavidA beat me to the explanation. I guess that makes it unanamus..

[awemawson]

Today I finally made a start on making a replacement take up spool. The original spool is in a sorry state. The two sides screw together to allow taking the used wire off easily. The threads have stripped, and some previous owner has wrapped the male side with PTFE tape to allow it to grip. Also the side cheek on one side has cracked and been rather crudely glued - time for a replacement.

[awemawson]

I decided to use polycarbonate, as it machines well and is pretty tough and resilient. The original spool cheeks are moulded, with a tapered lead in to guide the wire provided by a perspex disk insert. To replicate this I chose 6 mm polycarbonate sheet which I can taper on the lathe having cut the basic shape on the CNC mill.

The hub diameter is a nominal 60 mm and I found a suitable rod of 70 mm on eBay for an acceptable price