SPP Machine Begins

[Anthoney]

This is as far as I got before the argon ran out.

Heavily inspired by BigSwede.

Close up of blades and feed hole

You know Dad this is all metalwork so far. Right in your skill set. You could do a better job than I have.
The circuit will be simple and we would help if you made one.

I'm not convinced this one will work off the bat. Mystery blades have me worried.

[DAD300]

I appreciate that...really. But I now have gallons of SPP in three sizes. Plus if I need any more I figure to buy it from you guys and help you pay for your machines.

My build / experimenting now is for a 4" column...I have the column and an SPP that worked well in it. But I had power problems running it.

I can wind SPP by the yard now. If I were to build something for SPP, it would be an automated cutter to add to the lathe tailstock.

[Anthoney]

Yeah it was the cutter I meant.
You were an early adopter so it makes sense you have loads by now.
I saw the potential and wanted it from the start but got sidetracked completely for a couple of years.
Hence only just finished my ball I started way back when.

What size SPP worked for you in 4"? Is there a thread on that somewhere?

While off topic do you think the 1/4" or 1/8" convoluted stainless hose would work for a condenser? Is there a reason you went with 3/8"?

[BigSwede]

Anthoney, I think it looks great and should also work great if the linear slide holds up. I've been really happy over the fact that my blades show zero nicks. You couldn't shave with them, the razor edge is dulled, but it still cuts well. The speed of the cut helps, and that's where solenoids shine.

You have a TIG machine, there's another option. They can be hard to find, but there are special fill rods that are made for hard-facing. They use them on big earth mover bucket teeth. What you do, is run a bunch of close parallel small beads, add this fill, which then hardens to near-diamond hardness. You of course then have to grind like crazy, but what it lets you do is add super-hard layer onto mild or unknown steel.

I'm still having some issues with the contact sensor. It'll run for 5 minutes, then hiccup in a weird way; like a temporary short. Bzzzzt, solenoid buzzes like an angry wasp. I'm going to try an experiment with my Auber timer, this guy, reasonable at $35.

http://www.auberins.com/index.php?main_ ... cts_id=115

I'd experimented earlier with success, but the shortest "ON" interval is 0.1 S and even that is too long. Solenoid buzzes, and coil piles up a bit which then deforms the SPP. I did some brainstorming and came up with the answer, a hall effect sensor. A hall sensor is basically a switch that activates when a magnet comes close.

Plan is to set it up so the SSR circuit has this hall switch in there. Solenoid begins its stroke, brings magnet to the sensor. As soon as the hall flips, SSR circuit is open... solenoid relaxes. In the mean time, the 0.1 S on the timer has expired, so hopefully, there'll be one quick SLAP of the cutter rather than a protracted cut. And when your cut is fast and immediate, the cut quality goes way up.

[BigSwede]

Very cool, I discovered that this timer from Auber - and I've used it on various projects for years - has a function that will count loops, and will activate the cut. I know an Arduino or PIC will do this too, but for $35, and no build, can't argue with that.

I know this doesn't help non-USA folk much, because shipping becomes insane, but maybe someone lurking can use the info.

What you'll need is a bit of electronics knowledge. Auber sells hall effect units that interface. It's simple, clamp your SPP bit, add a little disk neodymium magnet to the spindle, have the sensor mounted within 10mm or so, and as the magnet passes the sensor, it delivers a pulse to the unit.

The pulses are counted, of course 1 pulse = 1 turn. Program for a cut at whatever loop count you want.

Ideally, another hall device makes the solenoid cut as fast as possible, but even the 0.1 second cut works fine.

A touch sensor is still the best, KISS extreme, but it's just still giving me grief, and I'm tired of massaging it for length, and tired of the intermittent malfunctions.

[Anthoney]

The Auber pulse counter sounds like just the ticket at a very reasonable cost. May search for a local supplier of something equivalent. I know Hall effect sensors. The solid state descendant of the mechanical reed switch.

Cutter is giving me grief so I know how you feel when you say you are tired of trying to get the touch sensor to work reliably.

First of all it moved as it should; it just didn't cut. Turns out there is too much vertical play in the linear track and the blade would ride up over the wire. Holding it with my fingers and giving the end a tap did cut wire. So I took the shaft from an old brass key and brazed it across the slide just before the feed hole. This made the blade edges collide at first, which did cut wire when tapped but was too much friction for the solenoid. So I filed it a bit and got the blades sliding over each other again, except sometimes they still collide. Either way the solenoid just doesn't cut and even tapping it just mangles the SPP without cutting. Feel like throwing the whole thing in the corner for two years. Bit depressing. Very frustrating. I tried the solenoid all the way up to 35V at 0.82A and still no joy. Now I'm wishing I could work to your precision with physical tolerances and tempering temperatures.

Think I need to abandon the slide and find another way. Never a Spong in the charity shops when you need one.

[BigSwede]

Anthoney, run through this thread if you find time, look for the cutter that DogMaster set up. I like it, gave me some ideas.

http://www.artisan-distiller.net/phpBB3 ... =11&t=4245" onclick="window.open(this.href);return false;" rel="nofollow

He set up a scissors arrangement. I'm picturing two strips of good steel, say 0.187" thick, adequate width and length. Drill a hole for the SPP in one, maybe weld a tube into the hole. Next, spot a pivot point close to the hole, bolt the other strip so you have a big scissors. Grind a chisel edge in the cut blade. Solenoid gets attatched to take advatage of mechanical leverage.

Another thought, find or buy a big solenoid, you want force and velocity both, and a 1" stroke length helps.

Another cutter setup I saw, SPP is fed through a disk near the periphery, and a revolving blade on the axle sweeps over the hole.

Any cutter needs to be rigid and stout and pretty powerful. Even this thin steel is a tough cut.

Final option, look for a style of cutter typically called a cable cutter, has a crescent blade set. Like gardeners pruners. I found a set for $8 at a hardware store, cuts wire like butter, but it would have to be adapted to a solenoid or pneumatic actuator. The latter, just throw a solenoid valve in there, it gets actuated the same as the solenoids we use.

[Anthoney]

In the process of buying an AC mains solenoid.

The disk cutter is based on a "Spong" hence my lament.

May try the scissors with the blades I have. Need a cobalt drill for pivot hole.

May even revisit the slide with the much bigger solenoid.

I know the cable cutters you mean. A cheap set is only about a fiver and would last for awhile.
Might be easiest way forward once I have solenoid.

[DAD300]

Anthoney...I've forgotten where you got your solenoid, but in the U.S. old washing machine transmissions had a great one.

[Anthoney]

It was from a friend and came from a big photocopier. Not up to the job though. The one I am looking at tomorrow is from some kind of postal machine.

[BigSwede]

Zoro has the monster I am using, about $30. I know I am throwing gadgets out there, they all cost and it adds up, but I've always done a mental cost vs labor thing, and less labor usually wins with me.

I think scissoring is going to be more forgiving of tolerances. You can tighten up the two blades by twisting a nut. With a guillotine setup, it's all in the frame, more than the cutters.

[Anthoney]

Yes there is a similar new 110v one over here going for £15 and the second hand 240v one is asking £25 but will prolly take £20 if I ask for £15 to match the other one. I can get a brand new one from hong kong for £20 but that is a long wait for postage. The one I am looking at is just down the road aways so perhaps meant to be. It's over $100 new! and the 110v one is normally £80 new.

[BigSwede]

Yes there is a similar new 110v one over here going for £15 and the second hand 240v one is asking £25 but will prolly take £20 if I ask for £15 to match the other one. I can get a brand new one from hong kong for £20 but that is a long wait for postage. The one I am looking at is just down the road aways so perhaps meant to be. It's over $100 new! and the 110v one is normally £80 new.

wHAAA-A-a-at? That's insane robbery for a simple mechanical solenoid. WTH?! 20 U.K. pound isn't too hideous, but $100 = 50 UK pound. These things should go for $25 or so.

I like Dad300's suggestion, cannibalize a solenoid from a washing machine. IIRC, they are almost identical to the one I'm using.

Way back, I said my solenoid was overkill. I take that statement back. I think power and speed will not only cut better, it'll prolong the life of the blades.

I'm retrofitting my machine for loop counting. On the back of the motor was an old encoder which I removed at the beginning of the project. Good part is, the frame is still there, and the motor has a stub shaft out the rear end. I potted a hall sensor in a brass tube with epoxy, and it fits perfectly in a little aluminum bracket screwed to the encoder frame. The cover even goes back on. All I need to do is stick a neodymium magnet on the stub shaft.

I ordered a couple of cheap Chinese counters. I can use my Auber, but I'd really like one that can program a relay-down (cut) of less than 0.1 seconds. 0.03 would be ideal I'm thinking, inertia alone will finish the cut. The massive challenge will be translating or just figuring out the Chinglish instructions... assuming there even are any.

A nice Red Lion counter that will do what I want is $200!! uh, no.

[DAD300]

Chinglish instructions...I had one a while ago that said, "Make Part A phuck Part B."

[1bigpig]

BigSwede,

I know that you like making things, but I have to ask. You said the servomotor had an encoder. I am willing to bet a bottle of the finest that it also has an "index" pulse. In the Auber specs on page 1, section E, it clearly shows how to wire up said index pulse from the encoder. Now, this assumes it really is an encoder and not a resolver, but still a lot of resolvers will have an "index" pulse. I would be happy to look up the specs and tell you how to wire your encoder if you post the brand/model and possible a picture of the encoder.

Second, the Auber specs states that the timing pulse can be as short as .01--page 1, section B, first line. So, you should not need anything else to decrease the delay. Matter of fact, if you have a zero crossing SSR, making it shorter than .02 seconds, will cause it to possibly not trigger the solenoid.

I hope I am not coming off as a know it all, but as you said, KISS!!!

BigPig

[Anthoney]

Well it was meant to be as it happens.

Really nice bloke, reminded me of another friend of mine, but having a better life because he works for himself instead of for other people. Let me have it for £15 and said I could take it back if it wasn't up to the job. Spent half a day chatting with him and another nice bloke he works with.

Decided to pick my dummy up off the floor, climb back into the pram and have another go with the linear slide. Took an evening plus a half day today but after several strip downs and tweaks it works as it should. Only downside is that during the snagging period I chipped the fixed blade quite badly. Obviously didn't temper it enough. It still cuts so rather than replace it and retweak it all I am just calling it a win. The solenoid was the key. This is a situation where speed and power is a needed factor to make the thing work.
You need a mains solenoid with a laminated core where it is about a two inch cube without the core sticking out and a good long pull. Mine seems to be a 240v equivalent to BigSwede's and was made by the Detroit Coil Company or RossoDecco as they are now known. Worth the £15.

The cutter is bulky, being another mahogany construction and I have yet to set it up with the winder. Once I do I will also need a switch mechanism. The hall effect loop counter sounds bulletproof so I have set my mate to task making a circuit for that one. He really is a good friend. I'll owe him more than a bottle or two once I finally get into production.

Once it is chunking away we will see how consistent and reliable it is over time. Hopefully it will at least be litres of SPP not minutes of cut time before I have to touch it again.

The finished article waiting to be fitted to the winder.

[BigSwede]

I love these inputs, guys, and appreciate them. I know enough electronics to get by, but barely.

BigSwede,

I know that you like making things, but I have to ask. You said the servomotor had an encoder. I am willing to bet a bottle of the finest that it also has an "index" pulse.

You are right, it was an encoder, a nice one. I destroyed it intentionally when I started this project! Why? Dunno, just didn't like the wire hanging off the end of the motor!

Years back I built a little bench CNC mill, so I have some ideas on how they work, and you are also right in that it could have been used in this project. Stupid me, but at the start, I never suspected I'd go the loop counting route.

Second, the Auber specs states that the timing pulse can be as short as .01--page 1, section B, first line. So, you should not need anything else to decrease the delay. Matter of fact, if you have a zero crossing SSR, making it shorter than .02 seconds, will cause it to possibly not trigger the solenoid.

I saw that... I'm impressed that anyone bothered to read the PDF on that thing! I thought too I could program it that way, but if you look at the section F, Function code and range, look at the "counter" modes at the bottom of the matrix, then skip to the "relay modes" section G. Unfortunately, there's a little note... Note, for mode 5 and 6, the duration of the relay pull-in is 0.1 to 999.9s. And the counter mode of the device must use relay mode 5, the one with "resets automatically." Modes 1 and 3 are a one-shot deal. Damn. Maybe I can figure out a way to deliver a "reset" pulse after the cut is completed.

But thank you for posting what you did, I've gotten some great ideas and advice from so many people on this project.

Good part with the loop counter sensor on the back of the motor, with the 10:1 gear ratio, I actually get 1/10th of a loop resolution, as if that makes any difference with the cutter. Given the gear box, I'll have to count loops of the length I want, and multiply by 10.

Anthoney: NICE! Sometimes, you just have to keep plugging, and I agree that with these things, a mega-solenoid is good. I think the key to keeping the blades happy is as much rigidity and accuracy as you can get, with blade A riding right on top of B, no gap allowed to form. Can you apply downward force on the moving blade? Maybe a spring?

Also, consider layers of aluminum foil, SS foil if you can find it, as shim stock to move blades as needed, perhaps closer together.

If you can make a cutter, the rest is easy enough.

[Anthoney]

Thank you.

After I clamped the lower blade tight and added some braze to the sides (it rode up, on just the back braze tacking it, while I was testing) I adjusted the brass finger to get the top blade as just right as I can. Seems good now.

Just waiting on the loop counter to start production. Will give me time to attend to other parts of the column and boiler.

Not quite there, but getting closer. Still quite compact despite the bulky cutter.

[Anthoney]

Rejigged the mounting for the cutter. Now it pivots out to clear the feed for setting up and maintenance. Can't remember what machine I've used, or when and where that had a gate like this on the working face but it is familiar from somewhere in my life. Feel like I used to use something like it somewhere.

Also clamped the feed tube so it can be removed and adjusted in place.

Really finished now bar the electronics. My mate has made it for me, just have to get it.
He is setting up the hall effect sensor tonight. Loop count set by 8 dip switches in binary. Anything from 1 to 256 loops. Cut time smoothly variable between a thousandth of a second and half a second by simply turning a pot. Not a very precise way to set it but easy to implement and I can always get a multiturn pot if it is too hard to dial in. I think it will be good though.

clamp for feed tube.

gate swung wide open. Pivots on a brass rod.

Gate swung shut and bolted closed.

Tried multiple cuts triggered by hand with the spp feeding. Worked fine.

One thing I did notice that on my longer hand driven cuts the still feeding wire piles against the back of the blade briefly and opens out a bit as a result. Nothing consistent but an interesting possibility that timing the feed and the cut might allow for some control of the profile without a ball bearing on a stick to upset it. Might work for copper too. If it works at all. It all snapped back into place for the next cut with no problem. My short feed from mandrel to blade is no doubt a factor in the slightly changing profile by pressure on the blade.

At this stage I can say "It lives!"

[DAD300]

Beautiful...

I agree, if you were to increase the distance between mandrel and blade, or even add a turn, the wire would have more flex to move during closure.

When I'm winding wire, I don't stop machine to cut the long segments. If I try to cut too close to mandrel I pull it off. But if a foot away it just flexes the winding and continues.

[BigSwede]

Anthoney - nice work. Might have missed it, but that gear motor looks like a power seat adjustment for an auto, true?

I know exactly what you refer to when you talk about the SPP strand piling up a bit, I saw iut when I went to a 0.1s cut vs a touch-sense cut. It causes the cut nugget to be slightly asymmetric, in a good way.

If you want to experiment, try this - cut the feed tube in a spot about 2 inches from the cutter, leave a gap in there where there is no confining tube. This allows the SPP coil to bow out, absorb the "traffic jam" at the cutter blade. In fact, I've had my better results when I minimize the amount of feed tube overall. Just enough at the start of the wind to keep her going straight, then a bit more at the end to feed the cutter. As much hanging free with no tube will smooth things out, if that's what you seek.

At the stage you're at, I think it's time to play with bit profile. Try a more square bit. If you can get it to feed correctly, it'll open between loops far more than the flat screwdriver bits.

Isn't it odd how this sort of project can consume man-hours at the expense of still construction? My poor MixMaster build has been on hold for weeks now. It's all good, I'm having a blast.

Cheers!

[Anthoney]

Dad: Thank you. Your work helped inspire me a couple of years ago and I appreciate your opinion.
I think I like my short path and will experiment with it that way.

Swede: Close but it is a bit larger, from a wheelchair push aid. It used to help my mum push my late stepfather around when there were hills. The slide came off the table of an obsolete video magnifier I bought her once and have replaced twice now. The latest one plugs straight into her big flat screen TV. So a bit of a disability based machine.

The entire path from mandrel root to cutter blade is less than three inches. My tube is about 2.5 inches. I see that as a possible advantage if I play with feed speed and cutter stroke. Won't change it unless I have to.

Rock: Thank you. Nice to know you are watching and like what I have done so far.

I know I said it was finished bar the electronics but had another play today and added an exit tube to keep all the little nozzles where they should go. As you say it is fun.

Original 16mm hole through both pieces of wood. A long tube difficult to clear when you stop.

A bit of reshaping to provide easy access and a slope downwards. It still starts with a 16mm tube about 10mm long.

Made a flange to fit over the hole.

smooth 90deg bend in 35mm pipe.

extension pipe added.

The pipes and flange began as all one piece of scrap I had lying around.

[YHB]

Ant, Big Swede,Dad,

This is the best thread I have read in a long time.

Thanks for sharing.

[Anthoney]

On behalf of myself, thank you Brian. As you know I have nothing but respect and admiration for your own powers of innovation and craftsmanship. I couldn't have done what I have without the work of BigSwede and Dad who went first and shared with me. I think the fact this has become a compound thread between us makes it better than if we all just did our own separate threads. It's all to the point. I'm glad BigSwede didn't feel hijacked and encouraged me to post here.

I'm quietly wondering if I should modify my design to use the brilliant three way head design you came up with and Swede smashed out of the workshop and over the horizon. Probably just delay things but tempting, very tempting. It's the best head design I have seen but I am still very taken with Manu's condenser controlled VM as Dad uses.

[BigSwede]

My machine is apart for a rebuild and rewire. As part of the process, and this is just how I do big and fun projects like this, I like to think of alternative hardware.

I know I talked up the Auber timer. I have two of them. It will work. No big deal if someone lurking out there bought one. But it has some issues, especially the OUTPUT cycle time thing. So I'm always on the lookout for industrial counter/timer controls.

99% of the time, I can't find specs beyond the very basics. Anything affordable is a Chinese import. I actually have a box with 3 timers/counters in it that I just can't quite figure out, or they are unsuitable. No big deal, these things cost $15 to $25 each, where a Red Lion or similar is $200+.

Anyway, took a chance on a counter off Amazon that does exactly what we want with almost zero fuss, not a lot of extra crap, easy to program, appears to be very nice quality. The unit will count loops, close the relay for the pre-programmed time, (0.01s) then reset itself. No need for any sort of a reset signal. Best of all, it's $17.

http://www.amazon.com/AGPtek-Digital-Co ... B00LBV8YS2

This is it, look no further... if it works! And I think it will. You get a feel for decent quality and good construction, this has it. I am going to test it, and order a spare if it does what I hope.

The Chinese are killing us... how can they make this thing for $17? For non-Amazon folk, it can be found at brainydeal.com.

Cheers! Got to get this thing back together.

[Anthoney]

Good luck with that.
At my mates now.
This is what he made for me.
He stressed it was just a lash up he had
knocked up in a hurry. I'm happy with it.

Simple loop counter.

Have to go shopping for a hall effect tommorrow and see how it works.
The ones he had to hand were bipolar and we want unipolar.
Going to get a sensitive one and an insensitive one to see which works best.

[BigSwede]

You owe your buddy some hooch! Looks very nice, simple to use. Can't see the uProcessor, is it a PIC?

The little neodymiun magnets today are so powerful, it probably won't matter which hall you use. Remember, most of them react only to the south pole, so if it isn't working, flip the magnet around.

[Anthoney]

Yes the big chip in the middle is a pic. There is a built in a/d that reads the pot and a built in counter that reads the hall effect. Pretty standard inclusions for pics nowadays. It works perfectly.

You were right both halls worked well. Problem is the cutter still jams sometimes. Have to strip it when I get it back to the workshop. Prolly have to replace the chipped lower blade. If it still gives problems after i give it my best shot i may give up with it an replace it with a different sort. Will be a shame, i like the linear design but i don't have your fabrication facilities.

[Edwin Croissant]

Maybe this is an idea for cutting the wire. I think that the second drill can be replaced with two pulleys and a belt:


And here is another one:

[BigSwede]

Thanks Edwin, Yes I like especially that rotational cutter. It looks to be very efficient, but it either needs to be tied in electronically to the rest of the system to get correct length, or it would simply be a matter of tweaking the cutter speed once the wind is up to speed.

The other issue, it looks to be something that has to be custom built. Still, those are some great ideas.

It seems that the cutter is the hard part of putting one of these machines together. Once mine is up and running, and I finally put my MixMaster rig together, I'd like to experiment with cutters that don't require a massive amount of fabrication. Maybe a small cable cutter attached to a stock pneumatic actuator.