Great content yet again! :) To get the flux off, easiest known method is to glass bead blast the part, gets rid of everything and makes it look nice as all the discoloration aslso disappears :)
Looking good! For expend wire storage I will use a 20 l plastic container sliced on the top to make 2 half’s and put the wire thru the spout , that way it’s safe as the spent wire is a electrical potential when the machine is on..
My Mitsubishi cuts the used wire into little lengths. About 3-4 mm long. The chopper in a simple mechanism. Almost like a 4 flute carbide endmill chopping the wire as it exits thru a carbide block with a hole the wire passes thru. Than the pieces just fall into a bin.
@@baxedm9806 They are but are expensive from Mitsubshi. There are also aftermarket ones available. The one I use has changeable cutter blades and wire tube. But it also is expensive. Just do a search for Wire EDM chopper. You will get the idea. It does make a little noise but less than the actual burning process. As to speed. I think it actually varies the speed based on the wire speed. You may be better to design your own to get the cost down. Maybe use some small carbide inserts from say a milling or lathe tool.
@@EdgePrecision Wire EDM guy here as well (though on FANUCs). Yes they do vary the speed as there's usually a parameter that is set to achieve a consistent cutting length. The FANUC ones use a square carbide insert, looks like one from a lathe, but has no codes etched on it so can't confirm. YMMV, but on our FANUCs the wire chopper gives a lot of trouble, even with regular maintenance. our Agie works great but costs 10x more and new ones do not come with wire chopper anymore, a sign that we are not alone RE the reliability of the system.
IMO the strain-gages will also have a limited bandwidth. An optical interruption-sensor and a long-travel spring would be my first go-to for a tensioning-mechanism.
Average strain gauges have an incredible bandwidth of at least 100 KHz, some even 300 KHz. So it is not the sensor that determines the BW it's the amplifier used. With low cost amplifiers, achieving a electrical BW of 1 KHz under gain 1000x conditions is easy. The mass of the sensing flexture onto which the gauge is mounted will limit the BW. This can be kept low and stiff so the BW will not be the problem, the noise immunity might...
Just use a larger Bin for the used wire, or.. design a "chopper" that chops the used wire into small bits ( the Mitsubishi FA20P has such a chopper, it cuts 30 mm strands.. works like a charm.. The Mitsubishi DWC200H used a large bin.. but that was a pain to clean out.. the Brother HS100 has a collecting spool that you could take apart, so when it was full you just screwed the outer flange off, dumped the used wire in a bin, and put it back into the machine.. A small bolt was used to attach the wire to the spool. works like a charm
My current MK1 DIY machine uses the brother HS100 wire re-reeling concept. The disadvantage is that you need to tie knots for that concept. Im working on a wire chopper, which will be featured on this channel.
@@baxedm9806 I never found tieing Wire to the spool a problem.. but if you have many wire breaks in a part.. it can be quite annoying.. i agree on that.. If you use a bin, either with or without a chopper.. that allows for future modifications like auto wire insert. The Brother HS100 i worked with, didnt have that.. the DWC 200 H and FA20P did.. that was great.. To get that to work on your machine, i'm afraid you need a total redesign of your wire guiding in the machine.. the small diameter rolls will make the wire "Curl up" when not tensioned after breaking.. You also need a "puller" in the lower arm, but that can be any system.. even like a little Turbine driven by water.. as long as it keeps the wire tight during insertion, until it reaches the Tensionroller, that keeps the tension on the wire during machining..
10:04 You state that that is a realistic wire speed. I did some rough calculations estimating the right wheel to be 30 mm OD, with a time to complete one revolution at about 5 seconds. That's 12 rpm, and given that my guess of 30 mm OD is in the ballpark, that is ~1.1 meter per minute of wire. Our FANUCs typically run at 6-7 meters per minute on roughing cuts, and up to 12 on skim cuts. Least I can use to cut anything and the wire not breaking (0.30 mm wire on very thin stock) is about 4. I haven't watched all your videos so i don't know what your final goals are, but 1 m/min seem a bit too low for me. RE your tension issues, I usually run into that problem when there is too much tension on the wire and it is yielding somewhere on the wire path, usually on my machines between the lower guide roller and the pinch roller out back. Sometimes it's grooves on the pinch roller themselves, due to wear. They are made out of some sort of ceramic material and we need to move the wire path a bit to the side when the groove reaches 0.03-0.05mm depth. That's about every 800h of operation. I have watched your newer vid where you made a closed loop tension circuit, that should solve most issues but the low wire speed may still bite you in in the back, as wire spent has less area and that can quickly make the constant tension circuit overload the wire breaking it. Also, since spent wire is smaller, it can ride in the pinch roller groove and make it slip. Anyway, these are my 2 cents, I am enjoying watching your project, keep it up!
Hi, thanks your elaborate feedback. I typically also run at much higher speeds than 1m/min. Usually in the same ballpark as what the Fanuc machines do. There are some applications in which a very low wire speed is also used. For instance, a part which has a relaxed tolerance on dimensional accuracy, but needs a low surface roughness. In such a case you can set the cutting frequency of the arc really high, and the current really low. With a really low current you can lower the wire speed drastically as well in that particular case. These are edge cases though, but anyways, Im desiging my machine such that it can be used for the edge cases as well.
Rather than having a motor control both the tension and feed, how about adding a pulley block with the wire looping through several pulleys? Could the tension be set by having different masses on the other end of the pulley system?
Could you add a second axial load on the opposite side of the motor? You could change the clamping mechanism so that the same spring and toggle apply force to two, opposed clamping wheels. This would ensure the clamping forces are the same. The two axial loads should mostly cancel out and allow higher clamping forces without changing the axle.
BTW, have you looked/considered the Stepper Killers from Teknic, also known as Clearpath servos? They can handle speeds from zero to 2-4k RPM with tension control, will also break if the tension is suddenly lost and of course can be driven with contant speed also (well, duh, servos) :) The only downside is the cost of 259 USD minimum, but they do not require anything else than a power supply and control signals, so no separate drivers etc.
I have seen those, and might go down that path. However I first want to try to get it to work with cheap motors. If it doesn't work out the way I want it, I'll switch to clear path motors or something similar.
Hi, great content (as always)! Just a quick question: Why didn't you use stepper motors for the wire feed mechanism? As 3D printing is so popular these days they can be bought really cheap and you wouldn't have to use a close loop speed control like you do with the dc motors. Greetings from Germany
Hi, really cool project. With the way you are collecting spend wire. Why are you not chopping the spend wire in small bit like they do on commercial machines? You are just recycling the wire anywar, no? If you explained it before I appologies. That makes the waste wire much more manageable wouldn't it. A good quality side cutter and and a Smal motor excenter mechanism would do the trick. Again really nice efford in any case.
@@baxedm9806:D, Well and I just watched your next video. I might be a bit overkill but you might wanna watch what the guy from Duet3d did for there COREXYUVAB 7Axis 5 Material 3D printer did (ruclips.net/video/pQ_FTlTxPm8/видео.html). They had a similar diecopling issue. But it is the insane approach. And If my experience of runing wired through tubes taught me anything, you wanna mount the spools suspendet over but not attached to your XY mechanism and keep those tubes short. Well I am looking forward to your future endavours.
Really interesting ! That's a good video to understand the challenges behind a simple looking wire feeder !
Great content yet again! :)
To get the flux off, easiest known method is to glass bead blast the part, gets rid of everything and makes it look nice as all the discoloration aslso disappears :)
Looking good! For expend wire storage I will use a 20 l plastic container sliced on the top to make 2 half’s and put the wire thru the spout , that way it’s safe as the spent wire is a electrical potential when the machine is on..
You make a great Job!👍🇩🇪
My Mitsubishi cuts the used wire into little lengths. About 3-4 mm long. The chopper in a simple mechanism. Almost like a 4 flute carbide endmill chopping the wire as it exits thru a carbide block with a hole the wire passes thru. Than the pieces just fall into a bin.
I like that idea. What was the approximate RPM of the cutter? Was it noisy?
Are the carbide block and cutter consumables? Would love to have the part numbers so I can see if I can order them somewhere.
@@baxedm9806 They are but are expensive from Mitsubshi. There are also aftermarket ones available. The one I use has changeable cutter blades and wire tube. But it also is expensive. Just do a search for Wire EDM chopper. You will get the idea. It does make a little noise but less than the actual burning process. As to speed. I think it actually varies the speed based on the wire speed. You may be better to design your own to get the cost down. Maybe use some small carbide inserts from say a milling or lathe tool.
@@EdgePrecision Wire EDM guy here as well (though on FANUCs). Yes they do vary the speed as there's usually a parameter that is set to achieve a consistent cutting length. The FANUC ones use a square carbide insert, looks like one from a lathe, but has no codes etched on it so can't confirm.
YMMV, but on our FANUCs the wire chopper gives a lot of trouble, even with regular maintenance. our Agie works great but costs 10x more and new ones do not come with wire chopper anymore, a sign that we are not alone RE the reliability of the system.
IMO the strain-gages will also have a limited bandwidth. An optical interruption-sensor and a long-travel spring would be my first go-to for a tensioning-mechanism.
Average strain gauges have an incredible bandwidth of at least 100 KHz, some even 300 KHz. So it is not the sensor that determines the BW it's the amplifier used. With low cost amplifiers, achieving a electrical BW of 1 KHz under gain 1000x conditions is easy. The mass of the sensing flexture onto which the gauge is mounted will limit the BW. This can be kept low and stiff so the BW will not be the problem, the noise immunity might...
Just use a larger Bin for the used wire, or.. design a "chopper" that chops the used wire into small bits ( the Mitsubishi FA20P has such a chopper, it cuts 30 mm strands.. works like a charm..
The Mitsubishi DWC200H used a large bin.. but that was a pain to clean out..
the Brother HS100 has a collecting spool that you could take apart, so when it was full you just screwed the outer flange off, dumped the used wire in a bin, and put it back into the machine..
A small bolt was used to attach the wire to the spool. works like a charm
My current MK1 DIY machine uses the brother HS100 wire re-reeling concept. The disadvantage is that you need to tie knots for that concept. Im working on a wire chopper, which will be featured on this channel.
@@baxedm9806 I never found tieing Wire to the spool a problem.. but if you have many wire breaks in a part.. it can be quite annoying.. i agree on that..
If you use a bin, either with or without a chopper.. that allows for future modifications like auto wire insert.
The Brother HS100 i worked with, didnt have that.. the DWC 200 H and FA20P did.. that was great..
To get that to work on your machine, i'm afraid you need a total redesign of your wire guiding in the machine.. the small diameter rolls will make the wire "Curl up" when not tensioned after breaking..
You also need a "puller" in the lower arm, but that can be any system.. even like a little Turbine driven by water.. as long as it keeps the wire tight during insertion, until it reaches the Tensionroller, that keeps the tension on the wire during machining..
10:04 You state that that is a realistic wire speed. I did some rough calculations estimating the right wheel to be 30 mm OD, with a time to complete one revolution at about 5 seconds. That's 12 rpm, and given that my guess of 30 mm OD is in the ballpark, that is ~1.1 meter per minute of wire.
Our FANUCs typically run at 6-7 meters per minute on roughing cuts, and up to 12 on skim cuts. Least I can use to cut anything and the wire not breaking (0.30 mm wire on very thin stock) is about 4. I haven't watched all your videos so i don't know what your final goals are, but 1 m/min seem a bit too low for me.
RE your tension issues, I usually run into that problem when there is too much tension on the wire and it is yielding somewhere on the wire path, usually on my machines between the lower guide roller and the pinch roller out back. Sometimes it's grooves on the pinch roller themselves, due to wear. They are made out of some sort of ceramic material and we need to move the wire path a bit to the side when the groove reaches 0.03-0.05mm depth. That's about every 800h of operation.
I have watched your newer vid where you made a closed loop tension circuit, that should solve most issues but the low wire speed may still bite you in in the back, as wire spent has less area and that can quickly make the constant tension circuit overload the wire breaking it. Also, since spent wire is smaller, it can ride in the pinch roller groove and make it slip.
Anyway, these are my 2 cents, I am enjoying watching your project, keep it up!
Hi, thanks your elaborate feedback. I typically also run at much higher speeds than 1m/min. Usually in the same ballpark as what the Fanuc machines do. There are some applications in which a very low wire speed is also used. For instance, a part which has a relaxed tolerance on dimensional accuracy, but needs a low surface roughness. In such a case you can set the cutting frequency of the arc really high, and the current really low. With a really low current you can lower the wire speed drastically as well in that particular case. These are edge cases though, but anyways, Im desiging my machine such that it can be used for the edge cases as well.
Thanks again Mike, iT would-be be great!!!
What if you could use the same wire in a continuous loop
Rather than having a motor control both the tension and feed, how about adding a pulley block with the wire looping through several pulleys? Could the tension be set by having different masses on the other end of the pulley system?
Could you add a second axial load on the opposite side of the motor? You could change the clamping mechanism so that the same spring and toggle apply force to two, opposed clamping wheels. This would ensure the clamping forces are the same. The two axial loads should mostly cancel out and allow higher clamping forces without changing the axle.
Could you run both the feed and retract pulleys off one motor by using a belt drive on the back side to Turn both at the same rpm?
BTW, have you looked/considered the Stepper Killers from Teknic, also known as Clearpath servos? They can handle speeds from zero to 2-4k RPM with tension control, will also break if the tension is suddenly lost and of course can be driven with contant speed also (well, duh, servos) :)
The only downside is the cost of 259 USD minimum, but they do not require anything else than a power supply and control signals, so no separate drivers etc.
I have seen those, and might go down that path. However I first want to try to get it to work with cheap motors. If it doesn't work out the way I want it, I'll switch to clear path motors or something similar.
muy buen trabajo
Hi, great content (as always)! Just a quick question: Why didn't you use stepper motors for the wire feed mechanism? As 3D printing is so popular these days they can be bought really cheap and you wouldn't have to use a close loop speed control like you do with the dc motors.
Greetings from Germany
Steppers do not have enough torque and introduce unwanted vibrations in the wire, which negatively affects cutting precision.
Hi, really cool project. With the way you are collecting spend wire. Why are you not chopping the spend wire in small bit like they do on commercial machines? You are just recycling the wire anywar, no? If you explained it before I appologies. That makes the waste wire much more manageable wouldn't it. A good quality side cutter and and a Smal motor excenter mechanism would do the trick.
Again really nice efford in any case.
A custom wire cutter will definitely be covered in a later video. Im already sourcing parts for it.
@@baxedm9806:D, Well and I just watched your next video. I might be a bit overkill but you might wanna watch what the guy from Duet3d did for there COREXYUVAB 7Axis 5 Material 3D printer did (ruclips.net/video/pQ_FTlTxPm8/видео.html). They had a similar diecopling issue. But it is the insane approach.
And If my experience of runing wired through tubes taught me anything, you wanna mount the spools suspendet over but not attached to your XY mechanism and keep those tubes short. Well I am looking forward to your future endavours.
Have you looked in to the work done by the ODrive guys?
Nope, never heared of them. Any references?
@@baxedm9806 ruclips.net/user/madcowswe
@@baxedm9806 ruclips.net/video/p4ltHDpxrbI/видео.html