[boof]

HI everybody , I m in designing stage of a new cnc machine , never built one this size , it has about 950mmx700mmx300mm working area (for now but gonna change it to 1200x900x300 later) it will be used for mostly wood/acrylic/plastics and aluminum ,and maybe some very thin steel sheets , anyway i ll be grateful if anybody point me any suggestion or errors in this design , the specification is :
X axis :
20mm supported linear shafts with 4 blocks and rm1605 c7 ballscrew with anti backlash ball nut
Y axis :
16mm linear shaft ,supported by shaft support , 4 shaft blocks as well , and rm1605 c7 ballscrew with anti backlash ball nut
Z axis :
same as Y axis

body made of 3/8" aluminum , it s some how a minimalist design no need for any machine cut , just screw holes , the other specifications :

steppers are nema 23 425oz no decision on controller yet and finally a 1.5kw water cooled spindle

so what do you think ? nothing is final yet so I m open to any suggestion

Regards

[adprinter]

Judging from your picture, it looks like a pretty good design. One suggestion: Begin NOW writing your "Owner's Manual", include as much information as possible for your own future reference. Things like CAD drawings of each of the various components, part numbers for things like bearings and belts (as well as the suppliers' contact info where you purchased the parts from). Wiring schematics, and diagrams which include wire color code translations for the various connections needed. Also any software purchased, and screen captures of the various program settings which are saved to your Owner's Manual will no doubt be an easy reference to lookup, years after you have completed your machine (and FORGOTTEN what the settings were in programs like Mach3, etc.). Like any machine your might purchase from a store, or dealer, the Owner's Manual is an essential item when it comes to adjusting, or repairing the machine (or ordering replacement parts). You will probably find that this book will grow to a rather large size, in a very short time. A word of advice, it is exciting when you are nearing completion of your machine, and the temptation to skip the information gathering of things like distances between bolt holes, etc. Thickness, and type of materials used to construct the various components, etc. RESIST THE TEMPTATION TO SKIP THIS INFORMATION! For it is only during the assembly phase of constructing your machine, that some of this information can be accurately gathered. To later realize this fact, may require you to disassemble you machine, simply to be able to record this information. RECORD EVERYTHING YOU CAN THINK OF, especially measurements, and dimensions of the various parts. You won't regret it, when you later decide to upgrade this, or add that, and wonder if it can be done. The more work you put into the book, the less work it will involve later when those inspirations hit you.

[wizard]

A few quick comments.

HI everybody , I m in designing stage of a new cnc machine , never built one this size , it has about 950mmx700mmx300mm working area (for now but gonna change it to 1200x900x300 later)

I'm not sure what your motivation here is but if you need 1200 x 900 it would be better to build it once. You will end up with lots ow waste otherwise.

it will be used for mostly wood/acrylic/plastics and aluminum ,and maybe some very thin steel sheets , anyway i ll be grateful if anybody point me any suggestion or errors in this design , the specification is :
X axis :
20mm supported linear shafts with 4 blocks and rm1605 c7 ballscrew with anti backlash ball nut

Supported by what? By this I mean what is under the supports that support the rails, I see nothing substantial. That could be due to the drawing but supported rails still need to be bolted to something stiff, flat and ideally with a bit of weight behind the structure.

Y axis :
16mm linear shaft ,supported by shaft support , 4 shaft blocks as well , and rm1605 c7 ballscrew with anti backlash ball nut
Z axis :
same as Y axis

The same issues here, you need to make sure what you bolt too is substantial

body made of 3/8" aluminum , it s some how a minimalist design no need for any machine cut , just screw holes , the other specifications :

3/8" isn't much over one meter. I'd seriously consider either thicker material for the base plate or a frame underneath. Mainly because of the distances you are suggesting for the x & y.

steppers are nema 23 425oz no decision on controller yet and finally a 1.5kw water cooled spindle

so what do you think ? nothing is final yet so I m open to any suggestion

Regards

You certainly have a nice start! I'm just not real excited about the idea of the long axis being bolted to 3/8" plate. You might want to consider backing up the plates with some extrusions or other structures especially on the X.

By the way the suggestion about putting together a manual is a good one. It helps with repairs and upgrades down the road.

[boof]

Thank you guys for replies

Judging from your picture, it looks like a pretty good design. One suggestion: Begin NOW writing your "Owner's Manual", include as much information as possible for your own future reference. Things like CAD drawings of each of the various components...

I know what you mean , been there , even a few month later after I finish something I have to take my time to understand what did I done ! but don't worry there , I have a excel sheet for parts info (Qs , price , where to buy , links , alternatives and ...) from the spindle to bolts and so on , and a notepad to write down everything I think I m gonna need to remember

A few quick comments....Supported by what? By this I mean what is under the supports that support the rails, I see nothing substantial

you right I 've certainly skipped some infos here , first of all the X shafts are bolted entirely on machine's bed (3/8 aluminum) supported by 1"*1" aluminum extrusions (underneath the bed that shown in the pic) , the Y shafts are bolted to back plate aluminum at the ends (donno if for this size it s a good idea or not)

but the good (or maybe bad) thing is because of my situation I can't start building this right now, so I m spending this forced time on reading and changing the design ! I was thinking to change the gantry's arms shape to get a better balance and i thought since i m changing the sizes and shapes and since I have the free time why not start over and change the whole design into a fully supported bed with 2 ballscrews (1605) and motors (2x400oz/in) for X axis (left and right) , changing my shafts to tbr20 for x axis and tbr16 for y , and hiwin 15mm rails for z axis also using 1604 ballscrew on Z and 1/2" aluminum for body parts and so on ...
I know it s gonna cost a bit more but I think it may worth the sacrifices !
so what do you think is it overkill or not ? (keep in mind that precision and accuracy is very important for me)

[wizard]

Thank you guys for replies



I know what you mean , been there , even a few month later after I finish something I have to take my time to understand what did I done ! but don't worry there , I have a excel sheet for parts info (Qs , price , where to buy , links , alternatives and ...) from the spindle to bolts and so on , and a notepad to write down everything I think I m gonna need to remember

Welcome to old age. At work I carry an iPhone around just so I can have a pocket full of notes.

you right I 've certainly skipped some infos here , first of all the X shafts are bolted entirely on machine's bed (3/8 aluminum) supported by 1"*1" aluminum extrusions (underneath the bed that shown in the pic) , the Y shafts are bolted to back plate aluminum at the ends (donno if for this size it s a good idea or not)

That is good to hear, 3/8" plate alone might make a nice small machine but I'd really worry with axis over a meter long. Even 1" square extrusions might be pushing it a bit as suitable reinforcements.

but the good (or maybe bad) thing is because of my situation I can't start building this right now, so I m spending this forced time on reading and changing the design !

Sounds familiar. It is amazing what can get in the way of fun.

I was thinking to change the gantry's arms shape to get a better balance and i thought since i m changing the sizes and shapes and since I have the free time why not start over and change the whole design into a fully supported bed with 2 ballscrews (1605) and motors (2x400oz/in) for X axis (left and right) , changing my shafts to tbr20 for x axis and tbr16 for y , and hiwin 15mm rails for z axis also using 1604 ballscrew on Z and 1/2" aluminum for body parts and so on ...
I know it s gonna cost a bit more but I think it may worth the sacrifices !
so what do you think is it overkill or not ? (keep in mind that precision and accuracy is very important for me)

You need to define precision and accuracy better because that means different things to different people. That being said stiffness is important to minimize deflection under load. I can't really say if your design will be good enough for your needs. If you have the time you might want to research the engineering calculations required to find out if the design is stiff enough, maybe do some finite element method. Im not an engineer so I can't help with that realistically.

However I look at or try to imagine what 1.2 meters of lightly supported aluminum would do and really have to question 1" square extrusions over that distance. What you might want to do is to experiment a bit if you already have extrusions. On top of simple deflection you also need to be concerned about vibration. I'm sure others will chime in here but I think you may need to consider a stronger solution especially for the longest axis. Again though that depends very much upon what you expect accuracy wise and surface finish wise.

[boof]

thnx wizard for reply ,...

Welcome to old age...

guess the age doesn't matter ) i m 30 and i ve been this way for years now ) anyway ,don lemme start on that ...

lemme explain what im gonna do , I mainly wanna build custom computer cases parts , so I need this to be accurate so I don have to do a lot of finishing on parts to make them fit , my test parts have 0.01 mm accuracy , it s may be abit over kill , but better safe than sorry , any way I redesigned the whole thing , the machine bed is now 1.5x3 extrusions laid on 3/4" mdf and reinforced across the length , all aluminum parts are 1/2" thick now and Y axis back plate is supported by another 1.5x3 extrusion , x axis guides are tbr20 , y axis guides are tbr16 and still didn't decide on z , may go for 16mm shafts with end supports or may go for tbr12 (or maybe some rails if I find them cheap enough)... anyway the gantry has a angle now so it has better balance than the previous design ...
the other thing is as i was searching , I found some really nice (at least they seem so) linear rails/blocks with reasonable prices on aliexpress.com but I donno how much they are reliable , they have some god reviews and some bad , but they have some buyer protection thingy now... anyway if I start to trust them i m gonna change all my shafts to rails :-) (do you think it s a overkill or not ?)

If you have the time you might want to research the engineering calculations required to find out if the design is stiff enough, maybe do some finite element method. Im not an engineer so I can't help with that realistically.

I did read some infos and use some online calculators as well (I hate those formulas !) and everything seems fine in theory (i mean for shaft and ballscrews diameter and weights and ...)

anyhow , here s the new design , the 3 1x1 extrusions on top are just there to support the top bed not the machine bed ...

[wizard]

thnx wizard for reply ,...

guess the age doesn't matter ) i m 30 and i ve been this way for years now ) anyway ,don lemme start on that ...

lemme explain what im gonna do , I mainly wanna build custom computer cases parts , so I need this to be accurate so I don have to do a lot of finishing on parts to make them fit , my test parts have 0.01 mm accuracy ,

0.01 mm is about 4 ten thousands of an inch and may be unrealistic in a Homebrew router. As for accuracy, if you need tolerances that tight for a computer case you are doing it wrong.
[/Quote]
it s may be abit over kill , but better safe than sorry , any way I redesigned the whole thing , the machine bed is now 1.5x3 extrusions laid on 3/4" mdf and reinforced across the length , all aluminum parts are 1/2" thick now and Y axis back plate is supported by another 1.5x3 extrusion , x axis guides are tbr20 , y axis guides are tbr16 and still didn't decide on z , may go for 16mm shafts with end supports or may go for tbr12 (or maybe some rails if I find them cheap enough)... anyway the gantry has a angle now so it has better balance than the previous design ...
the other thing is as i was searching , I found some really nice (at least they seem so) linear rails/blocks with reasonable prices on aliexpress.com but I donno how much they are reliable , they have some god reviews and some bad , but they have some buyer protection thingy now... anyway if I start to trust them i m gonna change all my shafts to rails :-) (do you think it s a overkill or not ?)
[/Quote]
If you can afford them I'd go for it. However like before, the best rails in the world will not do anything for you if the structure supporting them is flimsy.

I did read some infos and use some online calculators as well (I hate those formulas !) and everything seems fine in theory (i mean for shaft and ballscrews diameter and weights and ...)

That is only part of the equation. You want a nice stiff machine.

anyhow , here s the new design , the 3 1x1 extrusions on top are just there to support the top bed not the machine bed ...

[boof]

thanks for reply

0.01 mm is about 4 ten thousands of an inch and may be unrealistic in a Homebrew router.

I know 0.01mm is too much for ask , it s almost impossible , I meant my goal is to build it as accurate as I can afford to reduce hand working

As for accuracy, if you need tolerances that tight for a computer case you are doing it wrong.

I don understand , where did I go wrong ? what is your suggestion

the best rails in the world will not do anything for you if the structure supporting them is flimsy.

what would you do if you wanna build a machine this size , I mean for frame to make it steady , larger extrusions perhaps ?

I'd still worry about those thin side plates holding up the Y axis.

as I said those arms are 1/2" thick , however I could reinforce them or go for a thicker plate but I don think more weight is gonna be easy on x axis 400oz motor .... so what do you suggest here ?

[wizard]

thanks for reply



I know 0.01mm is too much for ask , it s almost impossible , I meant my goal is to build it as accurate as I can afford to reduce hand working

It isn't almost impossible it is impossible with conventional bolt together routers. Just buying ground lead screws with that sort of accuracy would move your budget into a different category altogether. We are talking order of magnitude difference in costs. That is for the lead screws and doesn't even get into what is required to get straight and square axis to that sort of accuracy.

I don understand , where did I go wrong ? what is your suggestion

It is a computer case not an effort at tool and die making. Excessively tight tolerance just make for lots of frustration. I actually struggle with this at work when an engineer comes to the shop with a print for a bracket that is toleranced to one thousands of an inch for no reason.

what would you do if you wanna build a machine this size , I mean for frame to make it steady , larger extrusions perhaps ?

Well first I'd get my head around what is really needed tolerance wise to make sure I can build it in my home shop! If not I'd have to consider spending the money to have a pro shop do the work.

As for the frame that is a more difficult discussion but you are talking about a rather large machine here to machine metals. As such I'd carefully consider how the design impacts stiffness. Since I'm assuming this is mostly sheetmetal you don't need to go overboard but need to look carefully at known weak points in other designs. For your needs it will cost you almost nothing to make a stiffer design as opposed to a wobbly one.

as I said those arms are 1/2" thick , however I could reinforce them or go for a thicker plate but I don think more weight is gonna be easy on x axis 400oz motor .... so what do you suggest here ?

Seriously an extrusion or box section or possibly a casting if you are into that. Real thick plate will do the trick too, but unless you come up with a fantastic deal heavy plate is generally pretty expensive.

The extrusion approach is probably the easiest for a limited home shop build but whatever method you use make sure you incorporate gussets to prevent pivoting at the end joints.

[boof]

Hi, with both designs, and the work envelope you indicate, you will struggle to cut aluminium even with a shallow depth of cut...
Ian.

It might help tto look on the net at various CNC machine builds. Here are a few examples: Joe's Hybrid 4x4 CNC Router, BUILD LOG: Aluminium Extrusion CNC Router Build, K2CNC - Table Top Router - KG3925, http://paul-flores.com/, Building a CNC Router Table - ArtofMyState, Homemade cnc router. Most of these designs address the issue of stiffness in the uprights in one manner or another. Realize that no design is perfect but you can tailor your machine to your needs fairly easily. Further I only am stressing the design here because of your stated tolerance levels even if they are a bit unrealistic. Before jumping into this build I'd do a lot more reading to broaden your design horizons. There is almost infinite design possibilities here.

One final thing; machining thin gage sheet metal isn't a bowl of cherries. You need to look into how you will keep things clamped in place or otherwise held down.

It is a computer case not an effort at tool and die making. Excessively tight tolerance just make for lots of frustration. I actually struggle with this at work when an engineer comes to the shop with a print for a bracket that is toleranced to one thousands of an inch for no reason.

thanks guys , it really opened my eyes , it seems i have to make some sacrifices here , if I want this machine as steady as it should be it s gonna be outta my price range ,... I wanted to do some large wood working and some aluminum cutting on same machine , but if I have to I can forget about those dimensions for now , and go for a smaller machine to do my aluminum working but it must has at least a 26x26 work envelope
so what do you think ? I know at the end it may costs more to build 2 machines but the aluminum working is more vital for me now and it s easier to make a smaller machine steady for that matter .... so tell me how s that ? what would you suggest for linear guides , motors and ball screws ?

[boof]

another view

[wizard]

another view

I'd still worry about those thin side plates holding up the Y axis. However it does look to be improved.

[pippin88]

Find a deflection calculator online and start putting some numbers in. You will be amazed at the difference certain changes make.

If I was to start from scratch I would use large square extrusion for my uprights rather than plate. Unless you find cheap scrap plate, it'll often be cheaper to buy a length of extrusion than the large plate need for gantry sides. Have a look at the moment of inertia for various extrusions.
See cnc router parts or kronos designs for examples. I think the most frequent weak point in most diy designs is the gantry uprights, usually plate with minimal bracing. These can easily bend or if using round rail allow pivoting.

[boof]

Find a deflection calculator online and start putting some numbers in. You will be amazed at the difference certain changes make.

If I was to start from scratch I would use large square extrusion for my uprights rather than plate. Unless you find cheap scrap plate, it'll often be cheaper to buy a length of extrusion than the large plate need for gantry sides. Have a look at the moment of inertia for various extrusions.
See cnc router parts or kronos designs for examples. I think the most frequent weak point in most diy designs is the gantry uprights, usually plate with minimal bracing. These can easily bend or if using round rail allow pivoting.

mmm , i donno if i m doing something wrong here , or how much this numbers are accurate ,... but I found a beam deflection calculator for solid rectangular shapes , and my gantry arm dimensions are 7"x16"x1/2" aluminum , and under 1000 pounds of force it only has 0.0000319" of deflection ...as mentioned by the site : "This Mechanical Engineering Calculator is to compute the defection of simple solid rectangular beams. It assumes the beam is supported on one end and the force is applied to the other end perpendicular to the width. Bending stress is also calculated."

so I think it s good enough (or am i wrong?) and it doesn't need any reinforcements ,... what do you think ?

[wizard]

mmm , i donno if i m doing something wrong here , or how much this numbers are accurate ,... but I found a beam deflection calculator for solid rectangular shapes , and my gantry arm dimensions are 7"x16"x1/2" aluminum , and under 1000 pounds of force it only has 0.0000319" of deflection ...as mentioned by the site :

We are talking about the uprights for the gantry arm not the gantry arm itself. Theses parts, the uprights and even the gantry arm itself are also springs and can be subject to oscillations or vibration problems.

"This Mechanical Engineering Calculator is to compute the defection of simple solid rectangular beams. It assumes the beam is supported on one end and the force is applied to the other end perpendicular to the width. Bending stress is also calculated."

The value they are returning seems oddly small but I could be missing something or they could be computing for a different orientation of the beam. You should try all three orientations of the beam as deflection ought to vary significantly depending upon how the forces are applied.

so I think it s good enough (or am i wrong?) and it doesn't need any reinforcements ,... what do you think ?

Truth is in the details! In this case your goal is to machine sheet metals which honestly isn't a huge load on the system. So the first thing is to avoid over doing it unless you have other machining goals in mind. However in this case I think the important thing to realize is that the cost difference between a good build and a so so build isn't that great and maybe non existent. It is how the parts that are put together that makes all the difference.

It might help tto look on the net at various CNC machine builds. Here are a few examples: Joe's Hybrid 4x4 CNC Router, BUILD LOG: Aluminium Extrusion CNC Router Build, K2CNC - Table Top Router - KG3925, http://paul-flores.com/, Building a CNC Router Table - ArtofMyState, Homemade cnc router. Most of these designs address the issue of stiffness in the uprights in one manner or another. Realize that no design is perfect but you can tailor your machine to your needs fairly easily. Further I only am stressing the design here because of your stated tolerance levels even if they are a bit unrealistic. Before jumping into this build I'd do a lot more reading to broaden your design horizons. There is almost infinite design possibilities here.

One final thing; machining thin gage sheet metal isn't a bowl of cherries. You need to look into how you will keep things clamped in place or otherwise held down.

[boof]

Thank you wizard for the time and effort you put to answer my questions lemme go check and will get back to you

[handlewanker]

Hi, with both designs, and the work envelope you indicate, you will struggle to cut aluminium even with a shallow depth of cut........the forces of the cutter will shake the gantry from side to side.....don't even think about steel........a depth of cut of .2mm on thin steel with a cutter of 6mm diam is the same as .2mm on thick steel.......the cutter pressure will flex the gantry and make the cutter want to ride up over the metal......it's too tough to cut with a moving gantry router, and you won't hold a thin steel plate down to cut it anyway.

With any steel cutting in mind, you would be better to consider a fixed gantry design, where the columns and crossbeam can be quite massive and braced to resist the torsional stresses applied to the crossbeam as they do not have to move so inertia with mass is not a problem.

A moving table, although taking up more length, can be more closely and rigidly coupled to the linear slides and are not subject to the slewing affect that a moving gantry router has when the spindle is to one side of the table and the single drive screw is in the middle of the table.

In a moving gantry, when the spindle is under load, it exerts a forward and backward torsional twisting force to the crossbeam, especially when it is in the middle of the X axis, and this is transmitted to the gantry side plates.
Ian.

[handlewanker]

Hi, just my 2 pennoth worth, on EBAY they have a 3 axis 6040 router, all the way from jolly China, which gives you approx 2 feet X 14 inches X 3 1/2" of working capacity and will cut aluminium as well as the usual blah blah etc.

The ones on EBAY are around $1,700 depending on where you are, but the control box needs upgrading to a gecko G540 as the one supplied, by all accounts, has been found to be total crap.

By the time you design and build one you will be out of pocket more than the buy and mend one.

Being able to have a working model straight out of the box almost, means you can dabble with it and build a bigger one if you want to, just using the design features as a starting point for the proportions as to what will work and what won't.

You haven't stated a budget or your machining capabilities......design is totally dependent on both options.

BTW, you could glue the sheet metal to a sacrificial MDF board, using a glue that is heat sensitive and melt the glue off after you finish.......something like candle grease etc.....LOL......area makes for holding strength, one reason why a vacuum table has such a great holding force.

If the need is strongly slanted for metal working, then I would consider, at the design stage, a fixed gantry model.
Ian.

[boof]

on EBAY they have a 3 axis 6040 router, all the way from jolly China, which gives you approx 2 feet X 14 inches X 3 1/2" of working capacity and will cut aluminium as well as the usual blah blah etc.

you probably right , it s wiser to start with buying such a machine , but I can not give up on learning to design one that easily even if I eventually ended up buying one .

You haven't stated a budget or your machining capabilities......design is totally dependent on both options.

I donno what u mean by my machining abilities but I m good with measuring and cutting if it s what you mean , I can weld too but I don have any welding equipment , If it makes such a huge different I may buy a cheap one for this job , doesn't hurt to have one around , and as for my budget I m thinking of around 4 to 5k and maybe a little more to expand .

you could glue the sheet metal to a sacrificial MDF board, using a glue that is heat sensitive and melt the glue off after you finish.

thanks for the tip

anyhow I was very busy but I ended up with this one , I tried to explain some of the materials on the pic , as you see I tried to go with a full supported base and minimize the gantry arms , I ended up with putting the Y axis rails on left and right sides instead of top and bottom config , I donno if it has a weak point that I couldn't think of , I didn't see anyone design it this way .
I used 1.5x1.5 steel square tubes at bottom , My biggest concern is the 1/4" brackets I used for ballscrews end supports , I donno if they are strong enough to hold them or not .
this is what I ended up for now , I learned lots of my mistakes from your replies , so don look at this as a final design just be kind and try to point me my mistakes or what I could have done better .

Regards

[handlewanker]

Hi, just been looking at your final design concept.

I like the way you have "eliminated" the gantry side supports and made the Y axis slides directly under the X axis crossbeam.

In my opinion this will stop the side flex that normal gantry design all have.

Dropping the table down below the Y axis rails will give a close structure, and should be a big improvement on the gantry side plate current designs.

I think this opens a whole new concept and a solution to the moving gantry rigidity problem.

I have to wonder if this design is current practice, but if not it needs a more closer evaluation.

One of the problems with a wide X axis is the slewing affect of a central ballscrew, but there are also potential problems with the twin screw design if one stepper fails and the other is unable to drive against the other stationary screw.

That is, unless the method of coupling the two screws to a toothed belt and driving with one central stepper to guarantee drive under all circumstances.

Having two steppers and coupling with a toothed belt will work but if one stepper fails the other is at twice load.

I think the long toothed belt will have problems too from stretching, and I'd think that a cross drive shaft with two mitre gears to the screws will eliminate the need for a long belt, still using two steppers but making the system fail safe.....backlash in the gears won't be a problem as they are only a synchroniser in case of one stepper failing or missing steps.

Looking at the first drawing in post #21, I would make the X axis crossbeam a bit more of a box section to resist torsional deflection from the cutter under load and keep the configuration of the linear rails on either side of the beam at the top as it now is, but possibly add a third one at the lowest point on the front face of the beam.

You show the side rails for the Y axis travel as 3"X 6" extrusions, so for the crossbeam, if this extrusion is doubled to give a 6"X 6" section that would make the X axis stiffer and less likely to bow or twist under load.

Without the rail on the front face bottom the Z axis will be too unsupported in its lowest position.

This means you will now have two rails on the front face, one above the other as is normal configuration and another on the back face at the top.

It's possible that having only two rails, one above the other on the front face, will be a better design, but you'd need to make the crossbeam a square box section to do that, which is OK too.

In the design you show in post #21, the X axis crossbeam is the main strong point as you now have no gantry side to worry about.

I would have to think that with this design concept you could possibly mill steel if the material sections themselves were strong enough.

I would even suggest that the total build be made in steel, not only from the economy, but also from the ability to weld it and the greater strength achieved.
Ian.