shop edm

[MrFluffy]

More conventional progress, I traced the leak last night to mostly the fact that the gutter designed to take the leaking residue that gets past the seals had a slot aprox 2" x 1/4" punched through it. So the seals are designed to catch most, and the rest is taken away by the gutter, which was instead ducting it onto the floor and machine table.
It was near the tracer table attachment point that was also torn off, so I think something big and heavy was dropped there causing this.
I have filled it with belzona 1111 (industrial higher strength version of jbweld) and today when I get out there I will seal the infil with a additional layer of silicon.
Tonight I hope it will run without leaks so I can leave it to do some longer cutting unattended.
The z gaiter still needs replacement and its beyond taping up really, it fell to bits when I tried to repair it, but a friend is checking their heavy plant stores for a suitable inventory piece. Hopefully it will match up to a bellows seal from a industrial machine which will be significantly cheaper.

[awemawson]

Well I hope that the bellows you find are cheaper than the ones I had to replace on my Eurospark H425 'sinker' edm - similar shape to your Z axis - just a convoluted cylindrical shape with ten convolutions. It was 280 UKP and this is the second time in about 18 years of ownership that I've had to change them On my Fanuc Wirecut Model M EDM the splash is contained by what look like 'way cover' flat convoluted bellows arranged with the lines of convolution 'standing up' vertically, and they slide in a top and bottom chanel. There are a pair, each are 115 mm tall and stretch to about 600mm - it cost me approx 110 UKP to have them made.

Expense these hobbies .......... !!!!!

[MrFluffy]

Yes, the quote for the same Y axis seal in a corrugated curtain style was $460 + shipping.
However, I've got very deep pockets and very short arms, so if I can't find one for the Z for under £50 I'll be sick as a parrot all week. There are chinese sellers of concertina style rubber seals and theyre in the under a tenner bracket.
All I need to do is find somoene who does a generic in 150mm diameter and 200mm long.

[MrFluffy]

Have a old section of tractor innertube as z for now, no photo's it looks nasty but it'll do for now.
What I do have is a photo of something I cut today. This little smiley wearing shades is my old company logo (and how I ended up being called mr fluffy). Material is 5mm thick 316 stainless. After some tips from Jmullet on cnc zone to use a higher MA value to stop wirebreaks and not to reduce V, I saw 9mm/min cutting rates and produced this.


Now feeling very very pleased with progress!

[awemawson]

Very nice. I dream of 9 mm / min - I can approach 5 mm at times if I'm lucky in 6 mm mild steel .

When you say increase MA - is that milliamps of current or some other variable? What wire speed were you using?

[MrFluffy]

The A of MA is the length of the OFF pulse when it encounters a unstable cutting condition. The M part of it is a variable which it uses to decide how unstable the pulse has to be before slowing down or correcting in some way. So the cutting conditions say a MA of 16 for this material, but JMullet suggested to start with 29 and work back until I got good speed without breakage along with a different servo, and that turned out to be 18, and the book servo setting. I managed to do a stock edged start instead of a start hole without fracturing the wire, if thats luck or the conservative settings only more testing will tell.
I was pushing it harder and harder at the end once the test piece looked ok, I wanted to get to 10mm/min because well, its a nice round barrier isn't it. But then I broke the wire on the final skim cut pushing it too hard. I still have the wrong diameter (too thin) wire in for the guides and am waiting for a reel of "Megacut" plated zinc wire to turn up, only reason for the fancy zinc plated wire was tha it was cheaper than hard brass, but will give it a whizz speeding it up to see.
Wire speed was 10, which the book says is 1340mm/min and right out of the book of cutting conditions. Reading that back I just noticed I had the wire tension wrong, its not a straight figure in grams but a translation for the machine settings, I had the wire too taut, and did indeed manage to break it at one point, but weirdly after the lower die and round the exit roller. I think the wire has issues because if you pull a length of it between your fingers, you can feel the odd mark in it, which obviously translates to weak spots.
Tweaking, but have a toolchain that works now too. From sketching it in cad to part. I need to get the cutting conditions entered into some kind of spreadsheet to put in the header as a C condition entry for the stock I want to use now, but as I'm not awash with raw material, that can happen when I buy a bit for each job.
I'm going to finish boxing the cabinet up and restore the last cab filters and redo one of the discharge lead ends, then I can start getting it networked properly & build/aquire some basic tooling. I'm using the strap clamps from the mill at the moment and hosing them down with oil afterwards to stop them rusting and of course there's the whole interference with the machine issues with sticky up tooling, but a 3R rail and vice are significantly out of my budget, so will have to make something up for myself.

[Old Hat]

I've seen impressive self-builds for work holding, like pairs of vertical vise-like overhangs.
And filtration builds that were so good that they built and sold same to other shops.

Google Images can be your friend, but hit'n the jackpot searchterm can take a while.

[awemawson]

I think that you said in an earlier post that you are using 0.1mm wire. I'm using 0.25mm hard brass.

I suppose that progress through the material is inversely proportional to the area of the 'cutting face' so for a given thickness of material a 0.25 mm wire is removing somewhat more than two and a half times as much material as a 0.1 mm wire. So perhaps our cutting rates are similar in volume terms.

My machine is hand threaded, and unlike modern machines, the feed side of the wire comes into the lower guide, so you have to thread 'upwards' which is bally difficult to see (especially with varifocal glasses !!!!) For this reason I've always run the machine conservatively to reduce the wire breakage frequency. Some time I must knuckle down and do some proper scientific testing of the various parameters and make a table up of optimum settings.

BTW the way to start at an edge with good flushing is to have a bit of Perspex with a groove in it 'cupping' the wire - can be super glued or held on with magnets

[MrFluffy]

The "megacut" 0.25 wire turned up today, and oh my, is it easier to thread. I literally could just push it through the two guides, vs trying to thread spaghetti through the eye of a needle for the 0.15mm hard brass. It is hand threaded but from above so easier to see. I broke the wire a few times and it was almost a joyful experience to rethread as compared to before.
I'll give that edge start tip a go, I just pumped the MA value right up and went for it, you can hear how spluttery it is, there's another vid later with it humming away when the flushing has got going properly. It was managing 1mm/min after a tweak of the parameters, material was 45mm thick 6083.
Quick youtube, dodgy camera work and it interferes with the camera mic.

https://www.youtube.com/watch?v=Q5cPGbMDcBg

It was cutting this, which was a L series 12tooth belt pulley for another project. The reason there's only 1/4 of it is because the wire escaped the exit bin and was snaking round the floor sparking out on things, so in the interests of not getting electrocuted by it, I hit halt. I'll have to build a inverted funnel for the plastic replacement bin I have, and looking at photo's that's what they were delivered new with.


Still more improvements to be had, but getting there.

[awemawson]

Really nice looking cut there. My used wire is wrapped up onto a take off spool which is quite handy. Some machines have a 'chopper' to cut the used wire into short lengths (about 1") to make handling easier.

I'd be interested to see your rate of cutting with the 0.25 compared to your thinner wire with the same materials and settings to see if my 'volume removed' theory holds water

Looking at the videos the current meter seems not to register much - is that actually the case ?

[MrFluffy]

It starts off at about 45v and seems to sit at 30v during the cut with slight variance as it corrects itself and the milliamp meter barely registers too, as the meter reads up to 200v either its just ticking over or the volt & amp meters aren't working correctly. I suspect a bit of both as it does rise with changes in parameters, I'll hook the oscilliscope back up and take a voltage measurement with that during the cut and see if it backs the meter up.
I will have another read of the cutting conditions book and see what takes the most current and see if I can run it a little harder as a test but the upper ranges of performance might involve thicker wire again to handle the power.
Also, when the wire escaped the bin, it was arcing out to gnd, which I think means one of my discharge cables has too high a resistance, as otherwise the preferred path to gnd would be via them and their low impedance connection, not via a thin wire, so I've ordered some new terminals to redo the temporary repairs I did prevously that I'm now not 100% happy with. It may affect the current flow having a higher table resistance, and might let me bring that MA value down to the book values for better speed too as I'm still running 2-4 points higher than the book on average.
Ill load the stainless test piece back on and do a comparison with 0.25 wire, be interesting to find out like you say.

I thought about making a chopper, but JMullet says it just coils into the bin and behaves itself, but my bin was purchased from the supermarket the other day so suspect its not quite as perfectly sized as the original. Plus I think the original machine has a invered funnel on the wire exit, which again is broke and missing on my machine. Thats what I think it needs to encourage it to go the right way, nothing fancy, just some perspex cut into a funnel shape upside down the size of my bin.

[awemawson]

I'll await your stainless test with interest!

On my machine the entire wire run is 'live' and that includes the take up spool of waste wire. The 'Discharge Electrodes' are how the cutting voltage gets on to the wire, and current flows from them, through the wire, through the work piece and to ground. As I understand it, it is correct that your waste wire is live and would spark to ground. It will only be grounded (and then partially) while a discharge is happening.

[MrFluffy]

Theory tested... I did manage to get it up to 11.8mm/minute but then I broke the wire. 9 seems do-able reliably. When its struggling it drops down to 5mm/min. Mostly it seems to slow to 5 when its having flushing issues near a edge or similar as George suggested. The voltage is still fluctuating a bit too much for my liking though but the new crimp ends haven't arrived yet to fix the dodgy discharge lead.

I cut this with it tonight, first start hole in the centre & easy to thread, going great, until the very last curve of the last spline I got another interpolation error, then drove it the wrong way in manual mode. I'll try to load and finish it off by mdi tomorrow. This is just a test piece to prove fit of my measurements.
Final piece is for the middle of the clutch for my drag race bike, and what started me on the rocky road to wanting my own wire edm.


I'll run through the code next time in dry run mode on the machine itself, lesson learned.

[Old Hat]

and what started me on the rocky road to wanting my own wire edm.

??
Rocky Road?
I'd like ta see what you can acheive when yer "On Fire" !

I know guys that are years on a big project.