[GeckoSub]

Hi,
so long as the current rating of the filter exceeds the load your fine.

I believe vector drive is perfectly doable at higher frequencies but requires more processor power. If you want that sort of thing check out Yaskawa.

Craig

So, went into the Yaskawa rabbit hole for a bit. They don't seem to have that much info readily available. And at least for the V1000 series, the 1000Hz version seems to be SW upgrade.
But from what I can tell from the one document I could find is that even for the Yaskawa V1000, you loose vector control in high frequency mode. At least that's how I read it - they seem to call it OLV - Open Loop Vector which is disabled when running the 1000Hz SW. But they still have something called Torque Compensation. Anyhow, perhaps they have more advanced VFDs that can do it.



The data sheet is here for anyone interested.
https://www.yaskawa.com/pycsearch?ke...=Entire%20Site

[joeavaerage]

Hi,
if I'm not mistaken Torque compensation operates like this:

A VFD drives and asynchronous motor, also called an induction motor. An induction motor rotor always is a little slower than the rotating magnetic field, this is called slip.
As the mechanical load comes on the rotor slow somewhat, increasing the slip, and that increased slip cause the motor toque to increase to match the load.

Lets say you have a 2P spindle motor. At 400Hz applied the synchronous speed will be 24000 rpm. With slip however the actual rotor speed might be 23700rpm. Now if you
start cutting the rotor speed might slow to 23500rpm. The increased slip will cause an increase in induced rotor current and thereby increase the torque.

A VFD with Torque compensation will increase the frequency of its output so that the speed of the spindle does not decrease or sag as load is applied. A small but useful wrinkle.

Craig

[GeckoSub]

Hi Gecko - The "mill" does not have to be any more accurate than the router. Just make it as accurate as you can. It would be a shame to be close to what you originally wanted (unless the want has changed) because it will be years before you get around to modifying it. May as well go that far now. HNY Peter

Heya Peter,
Just in case you have nothing better to do... please check your inbox

[jackjr-123]

So, went into the Yaskawa rabbit hole for a bit. They don't seem to have that much info readily available. And at least for the V1000 series, the 1000Hz version seems to be SW upgrade.
But from what I can tell from the one document I could find is that even for the Yaskawa V1000, you loose vector control in high frequency mode. At least that's how I read it - they seem to call it OLV - Open Loop Vector which is disabled when running the 1000Hz SW. But they still have something called Torque Compensation. Anyhow, perhaps they have more advanced VFDs that can do it.

VFD that can handle more than 600 Hz with precise speed regulation (vector control) are "military use" since a few years ago, and you won't find them easily. There are plenty of them, but you need to talk directly to the manufacturer to get them.

[joeavaerage]

Hi,
I've used V/F control for years, I can't really be bothered with vector control. V/F gives me perfectly good enough speed control, within a few percent....what more do you need?
If you want to do rigid tapping then get an AC servo and be done with it. Even a vector control VFD still does not give you position control in anything like the manner of an AC servo.

Craig

[joeavaerage]

Hi,
I've used V/F control for years, I can't really be bothered with vector control. V/F gives me perfectly good enough speed control, within a few percent....what more do you need?
If you want to do rigid tapping then get an AC servo and be done with it. Even a vector control VFD still does not give you position control in anything like the manner of an AC servo.

Craig

[GeckoSub]

VFD that can handle more than 600 Hz with precise speed regulation (vector control) are "military use" since a few years ago, and you won't find them easily. There are plenty of them, but you need to talk directly to the manufacturer to get them.

Yeah, I saw someone allude to that and enriching uranium/plutonium

Hi,
I've used V/F control for years, I can't really be bothered with vector control. V/F gives me perfectly good enough speed control, within a few percent....what more do you need?
If you want to do rigid tapping then get an AC servo and be done with it. Even a vector control VFD still does not give you position control in anything like the manner of an AC servo.

Craig

Yep, I'll prolly be fine. Either, way I wouldn't really know since this is still my first machine

[GeckoSub]

Back to Bases - Part 1
And An Important Question...
Hmmm... I am gonna have to ask a stupid question here, so pardon me in advance.
But I was really set on the moving gantry design from early in this thread. The smaller space needs was - and still is - a major advantage for me. But also, I could easily add end plates with gaskets to make a water bath out of the whole bed. Since I think I will cutting carbon fiber, too supposedly that's a nice way to do it. (Flood coolant is for later down the road once I start having fun with titanium). Also, maybe in the first few months of cutting alu with mist coolant, the tall side walls combined with end plates would help contain the chips enough that I can postpone having to build an enclosure.

Here's an early rendering (design is beefier now but Z-assembly and gantry side plates still not beefed up) and also a pic of a Chinese made Fixed Gantry Granite machine:




Disregard that the Fixed Gantry design in the pic above is obviously beefier but instead assume they both had equal girth where it matters. I know everyone says fixed gantry is inherently stronger. So, here's the stupid, basic question: Why is that?
I mean my design has really tall, sturdy frame walls, so the gantry side plates themselves (which needs beefing up) are not those tall flimsy ones you know from Chinese extrusion machines.
I can't wrap my head around why Fixed Gantry is supposedly a lot stronger. Some people say it has to with not having all the moving parts on the same assembly. But if it has to do with give in the motion parts, then doesn't that add up to the same amount whether they are all mounted on the same assembly/axis or separated out more?

Next post will be about why this matter now...

[GeckoSub]

Back to Bases - Part 2
The Mashup Option
The reason I am asking these questions right now is three-fold. We have Peteeng's nudge about making sure the machine is capable, other statements I have read in the past about the superiority of fixed gantrys plus I found a deal on a used fixed gantry machine that would be interesting. (I was hoping the high side walls in my moving gantry design would go a long way towards mitigating the supposed difference.)

Anyhow, I have spent some days looking for Fixed Gantry machines for sale. The somewhat affordable ones come in two different flavors:

- A lot of cast iron ones of questionable quality especially in the smaller sizes.
I know you shouldn't put too much into how they look on the surface (many look terrible), but I am befriending a young, Chinese machine builder who says many of them has flaws, threads pull out of bad castings and while the ways sometimes may be ground OK, oftentimes the support unit faces are not. It's quite the game of luck he says and not one that I am ready to play now. Which leaves us with:

- A few granite ones, though almost all of them are much bigger than I need or want. As for those, the same machine builder said that they are generally very precise, though of course not as stiff as a well made iron one.

Now, I came across a deal on a Fixed Gantry Machine in granite in the right size. 400x300x200mm travels which is almost exactly what I was aiming for bar a bit on the Z. It comes with ground PFI ballscrews and rails, has servo mounts, too and a table. So, if there's life left in hardware, I save a lot right there. (Besides the rails and ballscrews, I've been pricing out C5-C3 support units and motor brackets and it's not cheap...).


The problem is that the base of the machine is also the stand and in total, it's prohibitively heavy at 600kg... No way I want to have that shipped here or have to deal with it once here. In the pic, because of the angle, the stand looks small, but it's a solid, hefty block of granitw which follows the shape of the gantry all the way to the ground.
But... I can get the gantry itself and the rest of the metal hardware. I would then need to have a base made, but I have already gotten quotes for that, and it can be done affordably, too.

So, that's why I kinda need to know why it is that Fixed Gantry is so superior?
Because in terms of space needs and enclosing it, I will be at a disadvantage if I switch to this design. But yes, I do want the frame to be able to handle titanium prototyping of small parts down the line - not heavy hogging production stuff, mind you so I can make the compromise if needed.

Here are some very simple renders, used for getting quotes.
My original Moving Gantry design now has 75mm thick walls. The Mashup Option with the 2nd hand Fixed Gantry seems to have pillars of ~90x120mm (I don't have exact measurements, yet). And then the base design would be up to me in terms of how beefy, I'd wanna go. In this render it varies from 120mm where the gantry would sit to 100mm at the rail ways to 80mm for the rest of the base, more or less.


The third option would be to have a Fixed Gantry Granite frame made from scratch to my design...

[joeavaerage]

Hi,

So, here's the stupid, basic question: Why is that?

It is not true. If fixed gantry were stiffer than other designs for the same amount of material the EVERY machine would be built that way....but there is no one perfect design or layout.
Every design and layout can be made as stiff as any other, all it requires is enough material with enough cross section in the right places. It may be that one design has an advantage over another by your
choice of material or construction method.

Many criticize a column design....its not as stiff as....design 'x' or design 'y' or whatever design is popular today. That is complete rubbish, a column design can be as stiff as any other....if you
design it that way. It will certainly require a column of large dimensions to achieve that stiffness, and you might decide you can do better by using a different design...so go right ahead.
I chose a column design because it gave me the best flexibility (and I don't mean wobbly) to design and build a machine that suited the materials I intended to use and the techniques I could use.

If you choose to use granite then you would probably choose either a fixed gantry/moving table design or a moving gantry design. Both have advantages and disadvantages, but both are identical stiffness wise if
you design them so.

Having said that the pic you posted of a fixed gantry looks good. Is that something you can buy ready made?

My biggest bug-bear with granite is that I cannot work it myself. Even drilling an tapping a hole requires you to drill with diamond tools an bond a thread insert in there. Cast iron or steel, just drill a hole and
tap it. Could not be easier. If I need to cut steel, than a hacksaw, or oxy-acet or plasma....no trouble. Granite requires a diamond saw, and while there are a few in the city, I'd have to pay someone to do it,
whereas steel I would do it myself.

Craig

[joeavaerage]

Hi,

So, here's the stupid, basic question: Why is that?

It is not true. If fixed gantry were stiffer than other designs for the same amount of material the EVERY machine would be built that way....but there is no one perfect design or layout.
Every design and layout can be made as stiff as any other, all it requires is enough material with enough cross section in the right places. It may be that one design has an advantage over another by your
choice of material or construction method.

Many criticize a column design....its not as stiff as....design 'x' or design 'y' or whatever design is popular today. That is complete rubbish, a column design can be as stiff as any other....if you
design it that way. It will certainly require a column of large dimensions to achieve that stiffness, and you might decide you can do better by using a different design...so go right ahead.
I chose a column design because it gave me the best flexibility (and I don't mean wobbly) to design and build a machine that suited the materials I intended to use and the techniques I could use.

If you choose to use granite then you would probably choose either a fixed gantry/moving table design or a moving gantry design. Both have advantages and disadvantages, but both are identical stiffness wise if
you design them so.

Having said that the pic you posted of a fixed gantry looks good. Is that something you can buy ready made?

My biggest bug-bear with granite is that I cannot work it myself. Even drilling an tapping a hole requires you to drill with diamond tools an bond a thread insert in there. Cast iron or steel, just drill a hole and
tap it. Could not be easier. If I need to cut steel, than a hacksaw, or oxy-acet or plasma....no trouble. Granite requires a diamond saw, and while there are a few in the city, I'd have to pay someone to do it,
whereas steel I would do it myself.

Craig

[joeavaerage]

Hi,

The problem is that the base of the machine is also the stand and in total, it's prohibitively heavy at 600kg... No way I want to have that shipped here or have to deal with it once here.

My machine is 800kg, in fact more now with the trunnion table. Rigid machines tend to be heavy, fact of life.

Craig

[peteeng]

Hi Gecko - Every configuration can be made as stiff as required. But unfortunately this usually means mass and lots of it for commercial machines. The high rail design or moving column designs are used so the job does not move. In this way the motion systems and bed can be designed just to move the machine around not the payload which is a large variable. Moving table designs are used but then they have a bigger footprint then gantry or moving column designs nearly twice the footprint... Every config has a niche and you have a max footprint and max weight limit that will determine what's "best". So start with your required working envelope and work outwards. Start with a seriously stiff Z axis, then work out. If you work from the footprint in generally you run out of room for a stiff Z axis assembly. Ignore the footprint requirement and build out to see how you go. Maximise the size of everything as geometry is your friend and use the stiffest material that you can work with or afford to get machined. Your aspiration of titanium etc sets the bar very high. I suggest you look at steel plate designs, bolted not welded. I've built routers not mills (yet) but the high rail designs have been superior in stiffness to the column/gantry designs and they contain muck much better and the mechanics are out of the muck. Write a pro/con list for your designs and I'm sure the answer will pop out... Time to commit I think. Building machines is not cheap. It is not possible to build a cheap machine that does what you describe...so since design time is free I suggest you design the best amazing machine you can (ignoring cost for the moment) then review... Peter

[GeckoSub]

Hi Gecko - Every configuration can be made as stiff as required. But unfortunately this usually means mass and lots of it for commercial machines. The high rail design or moving column designs are used so the job does not move. In this way the motion systems and bed can be designed just to move the machine around not the payload which is a large variable. Moving table designs are used but then they have a bigger footprint then gantry or moving column designs nearly twice the footprint... Every config has a niche and you have a max footprint and max weight limit that will determine what's "best". So start with your required working envelope and work outwards. Start with a seriously stiff Z axis, then work out. If you work from the footprint in generally you run out of room for a stiff Z axis assembly. Ignore the footprint requirement and build out to see how you go. Maximise the size of everything as geometry is your friend and use the stiffest material that you can work with or afford to get machined. Your aspiration of titanium etc sets the bar very high. I suggest you look at steel plate designs, bolted not welded. I've built routers not mills (yet) but the high rail designs have been superior in stiffness to the column/gantry designs and they contain muck much better and the mechanics are out of the muck. Write a pro/con list for your designs and I'm sure the answer will pop out... Time to commit I think. Building machines is not cheap. It is not possible to build a cheap machine that does what you describe...so since design time is free I suggest you design the best amazing machine you can (ignoring cost for the moment) then review... Peter

A simple question for clarification's sake to start with: 'High Rails' - is that the same as what I have described as high side walls?

You're absolutely right about the Z needing to be stiffer in my moving gantry design. It took a while for me to see how obviously skinny it does look now but have known for a while. Same with the side and bottom plates for the gantry.

As for titanium, I can dial that down a fair bit on the wish list. If if I have to go slow and not get super finishes that's OK. This machine will predominantly make alu parts, small batches of them and basically prototype my product and a handful of other stuff, none of which needs to be machined with great MMRs. Once the product is dialled in in alu, the CAD file will go to a factory that can spit them out. Just judging from Fox's machine and mine already being 50% beefier (smaller rails and ball screws, tho), I should be fine for alu by now if I go with the moving gantry (with beefier Z and gantry side plates). As for other anecdotal "evidence" I have seen a few people make it work in Ti on maller Tormachs and not sure they are much stiffer, either. But that's guesswork on my behalf.

While I like the idea of hashing it out the design over a few months, I also really can't wait too long. And right now, I don't have the skills to really compare each design to each other - though my next post will be on that.

Also, if the machine works for alu, it will pay for itself by the small batches in the matter of months. So, even if it turns out I can't mill titanium, at that point I can either invest in a new machine - with all the learnings from the first one. Or "just" use the same design for my product in alu to have a factory make them in titanium.

While I have been hesitant to share numbers on some of these deals, I think it's worth sharing that I can get the top part of the fixed gantry granite machine and have a custom base made and have it all shipped to my door for less than USD 1000... That's including the used ballscrews, rails, table from the original machine. I have the spindle, VFD and servos ordered and incoming as well already - though the same parts will work well on a 100% custom design if I go back to that. Anyhow, we aren't talking huge amounts of money for something that should at least get me going in alu nicely.

[GeckoSub]

Comparing Designs - How To Do It In Fusion360 Simulations?
I actually finally had to start the full trial of Fusion a little while back so now, I have access to simulation for a few weeks longer.

But can any of you share a few pointers on how to compare various designs? I guess it would be ok for comparisons sake to substitute rails, cars, etc with solid material? But where do you add the load? You model a simple spindle and set a load 50mm under it? Again, I guess it can be even simpler than that if we are just comparig various machine bases.

That said, that will "just" tell me if my moving gantry design is equal to the fixed gantry mashup. (And will be a good tool for beefing up the Z if I go with a custom design). But it wont tell me how stiff they actually will be as I don't know how much load and where'd you set the load. And even if I did, I wouldn't know how stiff is stiff enough to have an attempt at titanium at hobby MMR.

Pete, sorry for bugging you again, but I think you have mentioned his before, but can't find it now. A simplified way of doing this. And an approximation of how much stronger a machine is with "solid" attachments instead of the real motion hardware.

Any screenshots or links or anything to get me started would be gratefully appreciated

[GeckoSub]

Hi,



It is not true. If fixed gantry were stiffer than other designs for the same amount of material the EVERY machine would be built that way....but there is no one perfect design or layout.
Every design and layout can be made as stiff as any other, all it requires is enough material with enough cross section in the right places. It may be that one design has an advantage over another by your
choice of material or construction method.

Many criticize a column design....its not as stiff as....design 'x' or design 'y' or whatever design is popular today. That is complete rubbish, a column design can be as stiff as any other....if you
design it that way. It will certainly require a column of large dimensions to achieve that stiffness, and you might decide you can do better by using a different design...so go right ahead.
I chose a column design because it gave me the best flexibility (and I don't mean wobbly) to design and build a machine that suited the materials I intended to use and the techniques I could use.

If you choose to use granite then you would probably choose either a fixed gantry/moving table design or a moving gantry design. Both have advantages and disadvantages, but both are identical stiffness wise if
you design them so.

Craig

Thanks for the sanity check.

Hi,
Having said that the pic you posted of a fixed gantry looks good. Is that something you can buy ready made?

Craig

The pic of the spanking new granite machine - I have asked around for off the shelf providers but haven't found one. The five I have asked so far say they are all custom made and none have a smaller size in the files already.
The used one in the pics; that's for sale right now and cheap. And the 600kg wouldn't have scared me if I had the space and knew I could keep said space for years. But I don't. But I can get the salvage guy to take the base-stand off of leaving me the gantry and all the hardware from the base. And then I can have a new base made - which is not a stand, just a lighter base - and the machine would end up looking like the brand new one in the earlier pic.


[EDIT]
Ever since the forum was down for a day (maintenance?) a few days back, it's been wonky. Almost all my posts end up repeating everything within the same post or double posting. I guess you guess may have the same issue.

[GeckoSub]

Hi,



My machine is 800kg, in fact more now with the trunnion table. Rigid machines tend to be heavy, fact of life.

Craig

Totally with you.
It's just that I don't have my own place here, nor a workshop I am sure I can keep. Nor am I sure I am gonna be on the island for that many more years. And also, when you need heavy stuff lifted here, the solution is island style - you ask more people to come carry. So, I just have a lot of reasons to keep things a bit more manageable. Even with a hit in stiffness/dampening.

[joeavaerage]

Hi,
given that the machine has to be mobile then you should look to machine that can be disassembled that you might move it more easily.

Conventional machines typically the column unbolts from the base. The weight of the machine halves is about 50/50.
Fixed gantry where the uprights bolt to the base/moving table also break the machine down into two more manageable parts.
High rail/moving gantry style is less amenable as the base and the sides are very heavy even if you were to able to remove the gantry.

I don't know how titanium crept into the conversation, but you are not being realistic. You need a machine of say max 500kg that you can dissemble into two pieces of 250kg each,
you have already decided on the size that you want.....so the granite has to cover that area. That will in turn limit the stiffness....and it is stiffness or lack thereof, that determines what the
machine can do....not your aspiration. I can well believe 500kg of granite in the size you want would make a very good machine for aluminum, possibly a little steel done gently, but titanium???

You want a machine to do all things, on all materials, be lightweight and be cheap. I think the Tooth Fairy was trying to sell such a machine. Last I heard he was sharing a flat with the Easter Bunny,
they were smoking a lot of weed and bonking all the girls.

Craig

[peteeng]

Hi Gecko - What you are describing is a machine that can be pulled apart, that is why I suggest a steel or aluminium bolted plate machine. You can break it down into big bits or pull it entirely apart if you can't get enough neighbors organised. If you read the Milli thread it goes through so much FE stuff it will sort you out. Maybe build a smaller machine first to get various things figured out. Initial FE models need to be big block geometry that captures the configuration so you can see the differences in the config. Generally a spindle nose load of 100kgf or 1000N is used. A dummy vice is built on the machine base and this is restrained. The load loop needs to be from the tool to the fixture, not the tool to the machines feet which is often done. The issues with granite and cast UHPC machines is that they are thick which makes parts very heavy, yet stiff. If you build up parts from say 12-16-20mm aluminium they can be ribbed efficiently hence they will be stiff yet "light". Peter

Heres Frankie - its made from 16mm F17 plywood. Cuts aluminium easily and uses an 800W spindle. Its a test build for a plate aluminium machine. This is the sort of config I suggest you go for. Bolted together it can be pulled apart upscaled, moded easily. You could build a ply machine then use it to build your aluminium machine... Thats been done here as well and I'm going down that path after mucking around with concrete, ceramics and various materials I've come to this conclusion...

Franky - is a high rail design. Also you say you can cut Ti by slowing down. Many people say this also about steel. But if the machine is not stiff enough then it just won't work. Going slow, going fast if the tool is being pushed aside then no difference... Seems you will make the stiffest machine in the budget and space you have, like others. Myself included...

[joeavaerage]

Hi,
start with a list of the things you want, like size and stiffness, with stiffness describing the machines capability.

Then write a list of constraints. The most glaring is mass, even 500kg is going to be a struggle unless you can disassemble.
You seem to be wedded to granite, and that poses certain limitations that cannot be overcome short of ditching granite altogether.
The next one is cost. What point in designing a machine, no matter how good, unless you can afford to build it?.

The constraints are effectively immovable, and they will set the best that can be achieved.......stiffness for a given mass and cost.....and that determines the best you can do within those
constraints. If that maximum capability is within the bounds of the things that you want....then you have a viable path to building a machine. If however the maximum capability is less
than want you want, then you either have to reduce the scope of what it is that you want OR relax some of your constraints, say cost or material or weight or all three.

I suspect that like most hobbyists you'll go ahead and build the best you can within the constraints that you face, and then discover how capable (or not) the combination is.
That is very much how I built my first mini-mill, and notwithstanding its limitations and outright mistakes it was not too bad. I found, like many before me, I had hoped
for capability level 10....but in practice had to settle for capability level 3. My next mill, my current one, benefited from my first experience and I decided on and attempted to design
a machine of 50N/um. When it was finished and within the constraints of cost, I have to settle for what it actually is, namely 20-25N/um. I had hoped for capability level 10 but got capability level 5.
Does that sound like a recurring theme?

Peteeng in his Milli thread has taken the reverse approach to most hobbyists, namely specifying minimum stiffness of 20N/um in all axes and a preferred machine envelope. He has if you like
specified the capability of the machine. Then he has proceeded to analyse materials, designs and techniques to match that basic stiffness within the bounds of cost, weight and practicality.
His search has been fairly exhaustive, and still there is no clear and winning strategy. He seems fairly well settled on cast UHPC and high rail/moving gantry style.
Just how well he has been able to live within his constraints of weight, size and cost is yet to be determined. I rather suspect that some of the constraints will need be relaxed somewhat
and yet achieve stiffness/size goals. The point is that the capability/constraint argument is always and ever present. Peteeng is a professional engineer and he is very disciplined about
sticking to constraints, whereas I tend to get as close as I can and then throw up my hands and declare 'close enough is good enough'. Where are you on that continuum?

Craig