GME's New 80/20 CNC Build - My Design

[Ntl]

Why would aluminum be more true than steel?

A aluminum extrusion is made different than steel tube. completely different animal with how it's made as well as the tolerance it has. He will be fine with flatness as long as he got his mounting surface level and on plane. It is interesting that it's so hard to find the flatness tolerance for 80/20 extrusion though. But regardless I guarantee it's way flatter than the 1018 tube the Saturn is made of.

If I personally built a extrusion machine I would buy pre machined lengths for the surfaces that need to be flat (gantry and rail mounts). I don't remember off hand what the company name is.

But like I said I'm sure he'll be fine with his table he's being very thorough with everything. And if our tables are turning out good parts his definitely will too.

Dan

[wmgeorge]

They also mount on the side because it helps keep the rails clear of dust, in fact they install a dust shield above it.

[GME]

Grabbed a bag of popcorn and i will be following closely this post!
I was to build a machine a bit like cncrp pro 4x8 but i might start from scratch like you did !

Thank you for watching. I hope I don't disappoint.

Gary it appears your machine is going to be pretty much like the Saturn2, even to the placement of the linear rail on the top. That leaves you no up/down adjustment (other than shims) and subject to the straightness of the aluminum extrusion. The CNCRP places their rail on the side to allow adjustments, they seem to have a proven design with no issues with accuracy.

I don't think they mount them to the sides to provide adjustment. That's an adjustment that few of their customers would be able to do. They mount them with a jig, parallel to the rails. You certainly don't want to be shimming the rails.

They use a Jig to aline with the top of the rail, side mounting gives them a way to adjust the up/down if needed. Same method CAMaster uses. On top your counting on the top of the rail being perfect otherwise you need to shim. There is a reason why CNCRP does what it does. Steel rails can also be ground flat, if needed. Why would aluminum be more true than steel?

Bill, I agree with Gerry. Few CNCRP owners are going to be able adjust the rails up and down. If you had to do that, it would most likely be because proper care wasn't taken when putting the frame together, or an extrusion was bent. Perhaps more to the point, few would even think about adjusting the rails.

Gerry and I also agree on the issue of shimming under the rails. You would never want to do it.

As we know, building a CNC involves compromises. The CNCRP machines are no different.

For openers, CNCRP instructions call for placing the jigs at each end of the extrusion, clamping the linear rail at each end, and tightening the linear rails screws in-between. This necessarily assumes the linear rail is perfectly straight. There is no straightness tolerance on linear rails, even on Hiwin, so, IMO, that is a mistake. There is no provision given to adjust out any lack of straightness. On the other hand, using the jigs closer together and working from one end to the other necessarily assumes two side rails are perfectly straight and the tops of the side reails are coplanar. Not necessarily the case. CNCRP had to figure out the best overall compromise. Reasonable minds can debate the methodology they chose. I'm not going to express an opinion on it. I went in another direction.

Perhaps more to the point, the CNCRP and CAMaster side mount assumes that the two sides of the machine are precisely the same distance apart. I don't know for sure, but assume CAMaster mills their machines so that the 2 sides are precisely parallel. However, the CNCRP present a different situation. I've already spoken to the 80/20 cut tolerances (.015" tolerance on cut length). If the cross pieces vary as much as .015", then the linear rails will vary that much as well. So, while you could theoretically make some minor adjustments up and down, you would still be left with some significant misalignment issues across the width with no practical way to adjust for them. Short of milling to length, like I did, it is what it is.

I am not suggesting that that CNCRP machines are bad. They aren't. They are designed for a specific market segment, and have a many satisfied customers. I just wanted to build in more accuracy and rigidity than their machines provide. I'm just in a different market segment.

On the issue of "why would aluminum be more true than steel," I suppose it's all about how it's made. I cannot represent that one is more true than the other. I don't know one way or the other. However, it is possible to mill aluminum extrusions flat and true, just like with steel. I recall reading in the Zone that some supplier provides the service, but I don't recall who. Maybe Misumi? Anyway, I would have done it, but given the lengths involved, I had no confidence in my ability to make the extrusion better than they come "out of the box." Actually, I believe I could do it, but not with the equipment I have. Milling the ends is one thing, but milling the surfaces of 60" lengths is something else. Sending them out for milling was not in the budget. It's one of those compromises I had to make.

Gary

[GME]

A aluminum extrusion is made different than steel tube. completely different animal with how it's made as well as the tolerance it has. He will be fine with flatness as long as he got his mounting surface level and on plane. It is interesting that it's so hard to find the flatness tolerance for 80/20 extrusion though. But regardless I guarantee it's way flatter than the 1018 tube the Saturn is made of.

Dan

Dan, the 80/20 specs are on page 60 of their catalog. I'm not in love with their permissible variations, but it appears that the way I constructed my machine, with milled uprights, that I got a pretty good surface. For those who may be interested, I've reproduced the 80/20 specs below:

They also mount on the side because it helps keep the rails clear of dust, in fact they install a dust shield above it.

Bill, it is debatable whether side mounts do a better job of keeping dust off the linear rails. Side mounted rails usually means they are mounted closer to the cutting action. No question that dust shields can, and almost certainly do, make all the difference. On the other hand, top mounted rails like mine, and on the Saturn line, are above the cutting action. Hopefully, that will help with dust. When I was using my first machine, which also had an elevated gantry, I didn't have a problem with dust collecting on the steel plates the skate bearing rode on, even when cutting MDF without my dust collector on, and without the dust brush in place. Not something I did often, but I did look at it. I understand that I will have to be far more careful with linear rails vs a skate bearing system. Linear rails are much less forgiving. I'm hoping that adding side skirts to vertical segments will help contain errant dust.

Gary

[ger21]

I understand that I will have to be far more careful with linear rails vs a skate bearing system. Linear rails are much less forgiving.

From my experience, they are more forgiving, if the bearings are the double sealed variety. We ran a machine for 10 years where the bearings were always covered in chips, and never had a problem with it. Just lube it regularly to flush any contaminants out.

[Peach52]

For GME:
On your post number 4 link for your picture are not good !
As i'm looking to buy a use benchtop mill i was curious about your setup, i'm not a machinist but a simple wise hobyist!

[GME]

Some may wonder why I chose to elevate my gantry the way I did versus that more often used side plates. In a word, rigidity. It seemed to me that two horizontal extrusions separated by a series of verticals would be more rigid than using side plates, or as CNCRP uses, a single piece of vertical extrusion in conjunction with smaller side plates. I'm not a mechanical engineer, and don't have empirical evidence that my choice was the best one. I can think of reasons why makers would choose side mounts over elevated gantries, the main one being cost. Whether using extrusions or welded steel, and elevated gantry takes more material and labor, which translates into a higher price tag. With welded steel machines, side mounts probably make sense, because you can elevate the gantry with welded steel uprights that are extremely rigid by both material used and by design. However, with extrusion machines, the materials used to elevate side mount gantries are far less rigid.

I know of three possible choices: a moving table, side plates in some configuration, or the one I chose.

I immediately tossed out the movable table from consideration. To get four feet of Y-axis movement would have required an eight foot table. Not a option for me, and more design issues than I wanted to undertake.

So, I was faced with choosing between vertical side plates and elevated frame models. Although I was milling the extrusions either way, there was always the chance of introducing some minor error. With top mounted rails, any width error were easily accommodated, although any height errors would remain. On the other hand, with side mounts, any height errors could be accommodated, but any width errors would remain. Although, at first blush, it might seem like a coin toss, it wasn't. With side mounts, any error would put significant strain on the linear rail system. With top mounts any error would be vertical, meaning that the rail would just ride up and down along the length. Now, if the two sides had any height disrelationships at any point along the length of the axis, there would be some unwanted strain. However, I believed, maybe erroneously, that potential height disrelationship would have less impact with top than side mounts. Also, I had 100% confidence in my ability to mill the short vertical pieces the exact same length. I constructed a simple milling jig to ensure the length. On the other hand, while I was confident in my ability to get my cross pieces to the same length, I was less than 100% confident. Again, I could adjust for some slight disrelationship in width, and my design technique should have more than adequately take care of height - at least to the extent that height could be accommodated for short of milling the horizontal faces of the extrusions.

Gary

[wmgeorge]

Gary & Gerry I did not mean to imply the end user would adjust the would adjust the side mounted linear rails other than on initial setup. That was the big complaint some users voiced on the Saturn2 design, the linear rails on some were not level and the only adjustment was shimming. Mine however and many more were just fine.

BUT if the rails did need leveling at some point in the future the adjustment is available on the side mounted rails. The side to side alinement or parallisism is taken care of in the design by using extrusions cut to the right length and you said that measurement was within .002 inch. So in theory that would be the side to side tolerance, but that can also be adjusted or corrected as I understand reading the setup instructions.

There is a reason why at least two commercial manufacturers are making them the way they are, and I never hear a complaint about either one cutting out of square or uneven.... do you?

[GME]

[QUOTE=Peach52;2290094]For GME:
On your post number 4 link for your picture are not good !
[/QUOTEorry, but I don't understand. Are you saying you cannot see the pictures? Your link takes me to Ebay for some suspension parts, so I must be missing something.

For GME:
As i'm looking to buy a use benchtop mill i was curious about your setup, i'm not a machinist but a simple wise hobyist!

I'm just a hobbyist too. I've never been a machinist; no formal training. I learn a lot from books and what I can find on the web, and then practice, practice practice. I'm lucky, because I'm pretty handy and have always been able to do about anything I've tried. It helps that I'm one of those "perfectionist" kind of guys.

Nothing special about my setup. My mill is a Precision Matthews PM25MV, which is similar to the Grizzly 0704. I chose the PM mill, because it has ground ways and is a belt drive, unlike the G0704. IMO, the PM is a better machine, but also costs more.

If you get a benchtop, don't make the mistake I made on your choice of vise. My first one was a 5", and also billed as precision, it really wasn't. 5" is really too large for a benchtop. I replaced it with a Shars 440V, a 4" vise. The jaws open to hold 6" material, which is great. Most don't open that far. It is precision made. Don't get me wrong. It's not a Kurt, which is the industry standard and carries with it a big price tag. Also, get some good angle plates. They will allow you to hold things a vise won't do so well with. I read of folks who use angle plates and skip the vise, but I wouldn't go that route.

Although they call it a benchtop, I bought a stand for mine. I also made a sub base for my stand out of 80/20 and added levelers to the sub base. That way, I was able to get my machine perfectly level with a machinist precision level. I used hockey pucks for leveling feet. They give a large bearing surface, and help with vibration dampening, not that my machine has much vibration.

I also added glass scales with DRO. I could have bought the machine with the DRO setup, but by installing my own, I saved more than enough to make it worth the effort.

I wouldn't even think about doing the work I did for my CNC without the scales/DRO setup. Way too much backlash in the screws that move the table to operate the machine easily. The DRO takes all of the guesswork out of setups and cutting. Remember, these are not huge knee mills. You can adjust the nuts on the table screws to take out most of the backlash, but setting up cuts and holes counting wheel revolutions is still not something that I would want to do. I would have preferred precision magnetic scales, but they were too expensive for my taste. The good ones cost a lot more than the glass scale variety.

I cut mostly aluminum on my mill. Best to use a mist coolant/lubricant. I use a Noga Mini Cool, which, I believe is about the best on the market. There are much cheaper ones on the market, but it's one of those cases where you get what you pay for.

For mist coolant, I use a soluble oil and water mix. Works great and doesn't rust the machine. It can be pretty messy though. Depending upon how long you run the mist, and how finely you have it set, you can get an oil slick all over everything in the room. I have mine set to a pretty coarse mist, which helps contain it. I've toyed with making a plastic sheet enclosure to contain it even better, but have been too busy with my CNC build to do it.

Get yourself a good roll around cart to have handy near the machine. Can't beat it for keeping the cutters, collets, edge finder, angle plates, and everything else you need close at hand. It's also a great place to put your vise at the end of the day, when you remove it and clean things up. You don't want to let the vise attached to the mill after using mist. The water and oil mix gets trapped under it and won't quickly evaporate out.

Oh, and I cannot understate the importance of ensuring your mill is "tuned up." Adjust the gibs. Tram the spindle side-to-side and front-to-back. Side-to-side is pretty easy. Front to back less so. Never trust that the machine is adjusted properly, unless you've checked it all yourself and revisit tuning on some sort of regular basis.

Finally, I suggest buying a mill that can be readily converted to a CNC sometime down the road. I don't know the range of mills the current kits cover, but I believe the PM25MV and G0704 are probably the most common, with the G0704 more so.

More questions?

Gary

[GME]

Gary & Gerry I did not mean to imply the end user would adjust the would adjust the side mounted linear rails other than on initial setup. That was the big complaint some users voiced on the Saturn2 design, the linear rails on some were not level and the only adjustment was shimming. Mine however and many more were just fine.

BUT if the rails did need leveling at some point in the future the adjustment is available on the side mounted rails. The side to side alinement or parallisism is taken care of in the design by using extrusions cut to the right length and you said that measurement was within .002 inch. So in theory that would be the side to side tolerance, but that can also be adjusted or corrected as I understand reading the setup instructions.

There is a reason why at least two commercial manufacturers are making them the way they are, and I never hear a complaint about either one cutting out of square or uneven.... do you?

Bill,

Your recollection about the 80/20 tolerances is off a bit. The .002" tolerance is .002"/inch for square across the end. So, a 15/30 extrusion could be out-of-square by .003" in the narrow dimension, and .006" across the other dimension. The length tolerance is .015". So, for the machine generally, where the width is governed by the cross pieces (the segments the spoilboard mounts to), the side-to-side tolerance is .015". Seems to me that is a lot for linear rails, and 3X the published height tolerance of about .005" for Hiwin HG20s. I believe height tolerance would govern in the case of side mounts.

Cutting out-of-square is a somewhat different issue, and is most evident in larger pieces, where any error is magnified. Unless you are cutting close fitting pieces, you may not notice there's an issue. Thickness issue are pretty much the same. Irregularity may not be so dramatic as to notice it. That doesn't mean the errors aren't there. I suspect that few folks are interested in dialing their machines into a gnat's ear, or even give it a thought. If they were, I suspect we'd be hearing about it. If they were looking, they were bound to find something, given the 80/20 tolerances. Remember, you and I have disagreed in the past about the level of precision I want in my machine. I'm guessing many more people hold your perspective than mine. I can live with that. To each his own.

As I've said before, none of this is intended as criticism of the CNCRP line of machines. They are designed to a market segment and are leaders in it. I'm not about to argue with their success. They do what they do as well as can be done in their price point. They don't scrimp, and are regularly making improvements. On top of it all, they offer world class customer service. If I was in the market for an "off the shelf" solution, CNCRP would be one of the first places I'd be looking.

Gary

[wmgeorge]

Gary, there are adjustments built in to take care of that in the machine setup, so the machine cuts and engraves true. If not there would be thousands of upset customers.! Added, you are quoting cut spec’s from 80/20 whatever, my guess is CNCRP cuts their own in house to much tighter standards. It does not make much sense to do otherwise. I am also guessing there are different grades or sources (US or Import) for the extrusions.

You are building the perfect machine and there is Nothing wrong with that at all. Best of luck. BTW I sold my small milling machine and it was wonderful, regret selling
.

[GME]

Gary, there are adjustments built in to take care of that in the machine setup, so the machine cuts and engraves true. If not there would be thousands of upset customers.! Added, you are quoting cut spec’s from 80/20 whatever, my guess is CNCRP cuts their own in house to much tighter standards. It does not make much sense to do otherwise.

You are building the perfect machine and there is Nothing wrong with that at all. Best of luck. BTW I sold my small milling machine and it was wonderful, regret selling
.

Bill,

I not aware of the adjustments you refer to that take care of a potential +/- .015" variation along the width of the machine. They aren't documented in the CNCRP literature. Seems to me that if you get things in proper alignment at one point on the machine, it will be off at another. How do you adjust for that sort of variation?

I believe your guess about CNCRP cutting their own extrusions misses the mark. I've seen many build posts referring to the extrusions coming from 80/20, either before or after all of the other parts arrive from CNCRP. The cuts aren't all that expensive, so taking it in-house may not make sense. And, it isn't just about cut lengths. The CNCRP uses a type of fastener that requires blind holes 13/16" in diameter to be milled in the ends of the extrusions. There are 4 such fasteners at each end. If you have a mill and tooling, and know how to use them, no problem. Not possible without a mill. So, capturing the work in-house would require a good saw and an decent mill, and someone who knew how to use it. Having your mill operator out of commission means orders stack up and service suffers. IMO, much more reliable and cost effective overall to just have 80/20 do it.

However, it isn't cut and milling costs that makes the biggest difference. SHIPPING does. The 80/20 costs a king's ransom to ship. The extrusions for my machine cost $460.99 to ship from northern Indiana to where I live in western Washington. CNCRP is located in North Bend, WA, which is a little over an hour north and east of me. For most of the country, shipping direct from 80/20 is by far the cheapest way to go. To me, it doesn't make sense to pay maybe $350 (their machines are significantly lighter than mine) to ship to North Bend only to pay a lot more to ship to a customer on the the east coast. Much less costly to just have the extrusions ship from Indiana to wherever the customer lives. One shot and it's done at the lowest cost. No repackaging and other handling at CNCRP's end.

Having variation across the width of the machine would not necessarily cause noticeable cutting/engraving irregularities. Even if the extrusions were to vary as much as .015" from one cross piece to the next, the transition would necessarily be gradual and split between the two sides (.0075 per side). The cross rails on the CNCRP Pro are separated 15-3/4" on center. If I have my math right, that works out to significantly less than .001" per inch of travel - gradually increasing or decreasing over that distance as the case may be. Not enough to notice, unless you were taking great pains to measure it. If you are working in wood, you would never notice it. In metal, you might see a tiny gap between what should be close fitting parts, but I'm guessing that most would write that off as what happens when using a machine designed primarily for wood - assuming they even noticed it. For most, less than .001" deviation when cutting aluminum on a wood router would probably seem like a dream come true.

I'm not convinced that any width deviation would even translate to a measurable error when cutting. If/when the width increases, as they say, "something's got to give." Something has to flex to accommodate it it, or bind to the point of lost steps. I would expect the flex to occur roughly the same at each side. If I'm correct, the Z-axis will stay true to position. Ditto with flexing inward. I wouldn't expect one side to bear it all, especially given that each side is constructed to the same same shape and dimensions with exactly the same material. So, if you divide the flex equally between the two sides, Z-axis will remain unphased by it all.

However you slice or dice it, I believe minor deviations are unlikely to translate into readily observable cut errors. So, I wouldn't expect customer complaints.

All this isn't to say that there aren't practical consequences associated with deviations of .015". That is out of spec for linear rails, and that can translate into premature wear. The hobby user may never use the machine enough to wear out the rails or bearing blocks. I question whether there are many who use CNCRP on a day in/day out production basis, but there probably some. For those folks, rails and bearing blocks may be viewed as a wear item needing replacement periodically. Again, probably not something to complain about unless they wore out very quickly. I doubt that most folks using an extrusion machine for heavy production have much experience with industrial machines and would really know what to expect in terms of wear.

Gary

[GME]

From my experience, they are more forgiving, if the bearings are the double sealed variety. We ran a machine for 10 years where the bearings were always covered in chips, and never had a problem with it. Just lube it regularly to flush any contaminants out.

Thank you, Gerry. Just what I wanted to hear.

[GME]

For GME:
On your post number 4 link for your picture are not good !

Sorry, but I don't understand. Your link took me to Ebay for something about suspension parts. I suspect the link was unintended. I've had CNC Zone add links to posts that I didn't input and made no sense. In fact, it happened to me with this post. I had edit to change the B in Ebay to lower case to make the link go away.

Are you indicating that you cannot see the photos in post #4? Please clarify.

Thank you,

Gary

[ger21]

I've had CNC Zone add links to posts that I didn't input and made no sense.

There's a setting for this, called "skim links" or something similar. Look in your user settings.

[wmgeorge]

I am going to say the CNCRP extrusions are cut in house to a much higher standard than 80\20. I know they drill and otherwise make it a no skill involved build. If 80\20 does supply some of the material its also possible they are cut to a closer standard.

I was very impressed by the online comprehensive instructions I read through.

I am also sure they do not purchase from a retailer like 80\20 but order the extrusions by the truckload direct. Most manufactors do the same to control the quality and cost of materials.

It would be interesting to hear from others who built machines from the 80\20 extrusions.

[davida1234]

I am also sure they do not purchase from a retailer like 80\20 but order the extrusions by the truckload direct.

I thought 80/20 ARE the manufacturer (or importers) of those 80/20 profiles as I have not seen anybody else offering those in the US. The T-Nutz profiles offered in the US have a slightly different look so they might be dodging some patent issues with 80/20.

[wmgeorge]

I thought 80/20 ARE the manufacturer (or importers) of those 80/20 profiles as I have not seen anybody else offering those in the US. The T-Nutz profiles offered in the US have a slightly different look so they might be dodging some patent issues with 80/20.

Even have milled surfaces for the rail mounting.


https://us.misumi-ec.com/vona2/detai...1503634&Inch=0

[davida1234]

Even have milled surfaces for the rail mounting.

Yes but, their profiles are different in sections.

[GME]

I am going to say the CNCRP extrusions are cut in house to a much higher standard than 80\20. I know they drill and otherwise make it a no skill involved build. If 80\20 does supply some of the material its also possible they are cut to a closer standard.

I was very impressed by the online comprehensive instructions I read through.

I am also sure they do not purchase from a retailer like 80\20 but order the extrusions by the truckload direct. Most manufactors do the same to control the quality and cost of materials.

It would be interesting to hear from others who built machines from the 80\20 extrusions.

I thought 80/20 ARE the manufacturer (or importers) of those 80/20 profiles as I have not seen anybody else offering those in the US. The T-Nutz profiles offered in the US have a slightly different look so they might be dodging some patent issues with 80/20.

David is correct about 80/20. They are the "mother company." They have a few distributors around the country who sell 80/20 products, along with other product lines.


Bill, I have never seen anything suggesting CNCRP or do anything else with the 80/20 extrusions. I'm curious where you found that information.

Here is a picture I found on the CNCRP subforum on a new build: https://www.cnczone.com/forums/cnc-r...896-build.html



As you can see, the most of the boxes are labeled from CNCRP, but four of them are labeled 80/20. Having bought more than my share of 80/20. The 80/20 boxes the same ones that have come with every order I've placed. See the staples in the end? They use a 1 by block of wood in the ends and staple the thick cardboard to it. I've never been able to open one of their shipping boxes without damaging it. I mention this, because it leads to the conclusion that 80/20 does not buy truckloads of extrusions, cut them, and ship them out. Even if they received them on an order-by-order basis, they wouldn't be able to open the boxes, make the cuts, drill holes, and thread ends and repackage in the 80/20 boxes. So, if they were processing the 80/20, the extrusions would come in CNCRP boxes. This was a 4 X 8 build.

There's another thread where the OP stood the extrusions on end and measured them for length. All were different, but within a few thousands. He commented that 80/20 had cut them and they were within 80/20's specs. He decided to use shims (like I did on my first machine). Remarkably, his ends were apparently cut square. I've not seen that. In my experience, they have all been and our beyond the maximum permissible tolerance. Maybe just my bad luck.

Long story short, it appears 80/20 does the processing.

Yes, I agree that the CNCRP instructions are very impressive. I would have put info on leveling and re-leveling at various stages of the build to ensure everything goes together nice and square. Put it together with one corner too high and your are going to have a baked in twist. It's very fixable but more work. That they don't is probably a reflection of the market they serve. If you've not seen it, Nate has a video for his 4 X 4 kit machine. Also very thorough and very nicely done. Anyone who couldn't follow his video instructions has no mechanical aptitude.

Gary