welding steel base or just getting aluminium extrusion

This is how i was thinking until Dean mentioned that its all good in a cad program but in real world it does not work that way. i was making my machine as tight and solid as possible in solidworks and had in mind that i will order these parts the same dimensions as solidworks and all will be ok, kjn have a tolerance of +/- 0.5mm now my gantry will have 2x 80x160mm in a l shape but how much of a problem will i have if one of those extrusions was - 0.5mm and the other was + 0.5mm.

edit.. more then likely i am way overthinking lol but thats my mind being ocd and following how lovely and neat my cad drawing is :)

Quote:

---

Originally Posted by reefy86

Don't want to be insulting to anyone regarding price but personally i think this is disgusting but then i am no expert on whats worth what when it comes to stuff like this. i have been quoted just under £2k just for 4 pieces and 2 of those pieces are only small plates for the rails

Attachment 23427

---

I feel I have said this before, but Reefy, there may be two factors at work here :-
1. The company has enough on their order books to put them in a position where they can issue F**k Off quotes to one off machining.
2. They are cynically set up to cream off the desparate.

The price per unit if you were ordering 1000 off each would be a lot lot lower.

I would advise you to keep looking:nightmare:

Edit: Having looked at their website, I go for option 2.

True but me personally would not be issuing F**k Off quotes when that person would potentially spread the work and bring me more custom regardless if i have mass companies ordering from me, I would just simply say to them i can but its months waiting list because obviously bigger orders are a priority in business :)

What F**ks me off the most though is they buy materials in bulk and have the machine to make these parts as if they are butter so more then likely it cost them hardly nothing for the material due to bulk buy and charge me almost £2k and thats not even half of the plates i need doing. on their behalf i bet its a good thing but for poor old me it bloody aint lol

They won't save that much in material by buying bulk.

People think that just because it's a flat bit plate, that it's easy to machine, but it's rarely the case. Taking your end plates for example, how would you hold them down?
On a router table, you could just screw the plate to the bed, or rely on a vacuum table, but I'd doubt they'd be using a router to machine them.
On a mill, you could clamp to the bed, but then you have to work around the clamps. You could use some form of low profile vice fixture, but you're still looking at two setups. Other option is to make a fixture plate, but that's additional cost and machine time.
Even your small plates, if you've specified tight tolerances around the full perimeter, instead of just using a bit suitably wide bar in a vice, milling both ends flat, maybe surfacing the top face, and drilling/tapping the holes, you now need to have two setups so all edges can be machined. Every time you have to do a setup, is time that has to be paid for by someone.

Plus you've specified a blasted finish, so that's more cost.

The quotes do seem high, but when you actually breakdown the time involved, they're probably not that high for a professional machine shop doing a one off job.

Could be wrong but i think you will find the +/- .5mm will be on cut length,so all you have to do is group your sections together with all of one end flush lightly clamp them together, call the longest your master rail and measure the difference of the others (shims/vernier) if you recess your end plates these can be done to the same depth and on assembly put appropriate packing into the end plate(simples)

Quote:

---

Originally Posted by reefy86

This is how i was thinking until Dean mentioned that its all good in a cad program but in real world it does not work that way. i was making my machine as tight and solid as possible in solidworks and had in mind that i will order these parts the same dimensions as solidworks and all will be ok, kjn have a tolerance of +/- 0.5mm now my gantry will have 2x 80x160mm in a l shape but how much of a problem will i have if one of those extrusions was - 0.5mm and the other was + 0.5mm.

edit.. more then likely i am way overthinking lol but thats my mind being ocd and following how lovely and neat my cad drawing is :)

---

This is very important! If using KJN aluminium to cut your extrusions you must have the pieces cut on the same day at the same time. They can offer a very high degree of repeatability but not if they have to reset the machine.

How accurate are KJN cuts I just had cut my MDF bed cut to 135.1cm *100cm based on my cad models and it dropped into place perfectly so I'd say at least for me there cuts were very accurate. I also had them cnc drill holes in my frame so I could assemble it with m12's instead of corners.

I did say to myself i would stay away from this thread.

That said, I do not like to see people getting quotes like you have for your plates. I have not read all the new posts ive only glanced over them. Plus i have no looked at those quotes in detail but its not right. I know roughly what you require as i have made parts for several forum members router builds in the past.

As mentioned they simply do not want to do the job or your drawings are asking for a lot and thats reflected in the price. Small jobs like this are agro for most company's and they charge you for that agro. Your drawing and the details on them will have a large bearing influencing the price. Some of the comments people have raised, i have said and others have said before and they are totally valid. I think its something most new people do not appreciate and it only sinks in with experience.

As the guys have said, you can do your plates long hand and of course has worked for many on here, in the same breath many choose to get them done outside.

I do this kind of works for people at a fair rate because i know the problem you can encounter with pricing. I like to see people doing these projects because it keeps Engineering alive and the knowledge base alive, if you guys cant do these projects, it dies.

If you want me to look at your plates please send me the drawings in either basic 2d dxf or dwg of if 3d in stp/step format. Be prepared to have your drawings scrutinised and alterations asked to be made. But please remember Anything i have said in this thread to you is for your benefit, even though you may see it as criticism, thats probably true for many other that have made comments also. Just to be clear I do not need the work, I have more work than i know what to do with at the moment, so work like this is done when my normal workload is reduced, so you have to wait until i have the time, but this means you get a professional job by someone who has been in the engineering sector (hand on) for over 35 years.

The offer is there. I'm not an ogre and i do not bite, but realise Engineering is a very straight talking trade. It has to be or we would never get anything done.

Quote:

---

Originally Posted by spluppit

I did say to myself i would stay away from this thread.

That said, I do not like to see people getting quotes like you have for your plates. I have not read all the new posts ive only glanced over them. Plus i have no looked at those quotes in detail but its not right. I know roughly what you require as i have made parts for several forum members router builds in the past.

As mentioned they simply do not want to do the job or your drawings are asking for a lot and thats reflected in the price. Small jobs like this are agro for most company's and they charge you for that agro. Your drawing and the details on them will have a large bearing influencing the price. Some of the comments people have raised, i have said and others have said before and they are totally valid. I think its something most new people do not appreciate and it only sinks in with experience.

As the guys have said, you can do your plates long hand and of course has worked for many on here, in the same breath many choose to get them done outside.

I do this kind of works for people at a fair rate because i know the problem you can encounter with pricing. I like to see people doing these projects because it keeps Engineering alive and the knowledge base alive, if you guys cant do these projects, it dies.

If you want me to look at your plates please send me the drawings in either basic 2d dxf or dwg of if 3d in stp/step format. Be prepared to have your drawings scrutinised and alterations asked to be made. But please remember Anything i have said in this thread to you is for your benefit, even though you may see it as criticism, thats probably true for many other that have made comments also. Just to be clear I do not need the work, I have more work than i know what to do with at the moment, so work like this is done when my normal workload is reduced, so you have to wait until i have the time, but this means you get a professional job by someone who has been in the engineering sector (hand on) for over 35 years.

The offer is there. I'm not an ogre and i do not bite, but realise Engineering is a very straight talking trade. It has to be or we would never get anything done.

---

Thats kind of you mate and you were never in the wrong in the first place or did i ever discourage any of your advice i was just simply trying to avoid welding and go with aluminium but that soon changed lol. I have a few alterations of my own yet so ill sort those out and send you the files if thats ok with you.

Been thinking about how i am going to square the gantry up and thought of an idea using the method used below (watch from 2:20)

i know that method won't work on mild steel box section but what if i used epoxy resin on the side and wait for that to fully cure which will then give the side of the steel box section a really flat surface and then do the method in the video? i would need to do this before welding the top rail beam to the frame off course.

what do you think is it over the top?

https://www.youtube.com/watch?v=YOVHtjXi07Q

Quote:

---

Been thinking about how i am going to square the gantry up and thought of an idea using the method used below (watch from 2:20)

---

The way I see it is that all he is doing is aligning the rail to the same as the ali profile so if the ali is bent so is his rail:confusion:

It sounds like you need to invest in some tools such as a straight edge, engineers square.
I attached one of the long X rails using a straight edge to get it straight, then it was easy to space the second rail from it. Once you have the bearings fitted and a gantry cross beam attached you can feel if the bearings are tight at any point in the travel and make adjustments to the second rail.
Here's how I squared the rest up; http://www.mycncuk.com/threads/6565-...5217#post55217

That is pretty much how I did my y axis with the rails onto extrusion.


Sent from my iPhone using Tapatalk

I am getting the impression that you haven't done your basic research (ie wadding through the vast amount of information in the build logs) the epoxy leveling procedure and rail installation has all been well covered, and Boyan went into this when he started his build. you can't build an accurate machine without certain items(ie precision straight edge and square) no offense is intended by my comments.As Clive has said if you proceed as outlined in the video you will just duplicate any errors in the ally rail.
Regards
Mike

By the same token, the guy bolts his rails straight to the top of the extrusion. No particular reason to believe that this is any straighter than the side of the extrusion - the surface will be influenced by errors in the supporting components, etc. However, errors in work on the machine will be in depth rather than profile accuracy, so perhaps up-and-down errors are less important for a lot of work.

For a router intended for mostly work in wood or plastics, though, will the errors be significant? If you are going to be making sign boards, doing lettering, cutting out stand-alone shapes which do not have tight manufacturing tolerances, is it worth chasing the last micron? I'm not saying you should ignore any errors, but you need a sense of proportion here. If your ballscrew/ballnut/bearing assemblies give, say, 0.05mm backlash, how much better than that do you need to be in the supporting structure?

Steel is different - my own 1.8m support rails dipped by 1.5-2mm in the centre, which is why I used epoxy - but in practice, how accurate are these aluminum extrusions?

Quote:

---

Steel is different - my own 1.8m support rails dipped by 1.5-2mm in the centre, which is why I used epoxy - but in practice, how accurate are these aluminum extrusions?

---

I have built several router type machines with the heavy duty profile 90x45 with the L type gantry layout with the rails top and bottom with no issues. So in my conclusion they are fine for the majority of diy builds and work very well.

Was just a thought after watching that video, i know i will be needing straight edges and engineer squares but most of these videos on youtube are just some guy pointing his camera at an engineer square and going up and down which lasts 20 seconds long lol. i am still reading on here and watching other people tips and tricks but all i can picture in my head is getting one rail on and have it as straight as possible on one side and bolt it down and then use the gantry to slide up and down to get the second rail inline and then bolt that down and then just use engineer squares against the aluminium beam and side plates to make sure thats square

Quote:

---

Originally Posted by EddyCurrent

It sounds like you need to invest in some tools such as a straight edge, engineers square.
I attached one of the long X rails using a straight edge to get it straight, then it was easy to space the second rail from it. Once you have the bearings fitted and a gantry cross beam attached you can feel if the bearings are tight at any point in the travel and make adjustments to the second rail.
Here's how I squared the rest up; http://www.mycncuk.com/threads/6565-...5217#post55217

---

read through your thread mate and i have a better picture in mind now so thank you, after reading about your problems with the bk12 bearings is this something i am going to run into when buying from zappautomation? mine will be the bk15 and bk17 not the bk12.

Quote:

---

Originally Posted by reefy86

Was just a thought after watching that video, i know i will be needing straight edges and engineer squares but most of these videos on youtube are just some guy pointing his camera at an engineer square and going up and down which lasts 20 seconds long lol. i am still reading on here and watching other people tips and tricks but all i can picture in my head is getting one rail on and have it as straight as possible on one side and bolt it down and then use the gantry to slide up and down to get the second rail inline and then bolt that down and then just use engineer squares against the aluminium beam and side plates to make sure thats square

---

You are going to need to be able to square the gantry once the machine is up and running, by adjusting home positions for each end of the gantry. Whether you do that via software or manually (home one end as part of the usual homing sequence and then adjust the second end manually) doesn't really matter but you do need to be able to do this. Probably the easiest and most accurate way to check that the gantry is square is to use the machine to make four holes in a square pattern in a piece of suitable material (I use MDF and make the holes a firm fit for the shanks of a set of identical drills). Measure the diagonals and adjust the home positions for the gantry until those measurements are as close to each other as you can make them. Any machine which uses two motors/ballscrews on a single axis will need to be able to do this.

Hope you all have a good Christmas guy's

is it possible to combine a 4th axis capable of cutting up to 200x200mm to my design? i am struggling to figure out how i would without having to adjust the z height, my plan was to adjust the width of the machine to allow me do add a t track to the side of the machine which does not affect the spoil board so i can adjust the 4th axis length so maximum cutting for the 4th axis would be 2400mmx200mmx200mm. So far my z height already has a cutting height of 200mm so if i was to somehow add the 4th axis i feel as if i would now need to adjust my z height to around 300mm and i have no idea how weak it would make my z axis doing this.

Attachment 23493

http://www.chinasavvy.com/tolerances...tnessroughness

What about this idea; https://www.cnccookbook.com/wp-conte...atheAttach.jpg
or put it at the end in a smiliar manner.

Thanks eddy :)

Just responded you to the PM, but now i see here where you are going with that. There is absolutely NO NEED for the spindle to extend below the table bed in order to achieve good 4rth axis

Meaning that the centerline of the round object could be level on with your bed. And the Z gantry clearance will be your object max radius. Add 10cm for the bit and chuck. There should be your spindle in upmost position. i recommend boxing the spindle like on mine machine and making tha gantry a bit wider in vertical. So more or less a normal machine without too much mods except for strong Z. thats it

I recall a build by KingCreaky that had a removeable bed that allowed a 4th axis to bolt in (can't find the thread though).

been over thinking again lol but that's the fun of reading tons of posts on this forum and coming up with new idea's (stolen idea's). just out of interest whats the price difference between good stepper motors and servo motors now? because of my high z axis i need to make the gantry very beefy but after reading some more for 3d work the z front plate needs to be light so i will update soon on an updated version and see what you think.

What i think for sure is this:
Do not buy anything more until you have definitely finished your drawing and have it all clear 100%. That includes rails, screws, motors, board, etc. It will be just throwing money in the bin. And yes, we have done it all.

One more thing to note: Everyone here thinks he is a top machine designer. Nothing wrong with that. And in reality here we could see most of the best build ever worldwide. But what i learned from experience is that even if you overthink your machine, you always mess something, even small. Once you make the machine though, use the machine 1 year 8h per day, maybe make another one later, revise a couple of times the design customising it for clients, then this machine becomes better and better.
What i am saying is: assume you will make errors, if someone has made a nice working machine and you like the design, ask, listen and copy shamesly. Do not over think and over invent. look for good working designs, easy to make at home.

Thanks boyan, i have not ordered anything apart from the steel for the frame and while spending the whole of january welding it together i will wait until that is completed before confirming other parts to order. i have seen loads of build logs on here from people who have rushed in and bought the parts before asking and i will not be making that mistake. i am not expecting this machine to turn out how it looks in cad software or in my head but what i am expecting is knowledge from trial and error. that is the way i always like to learn because if you can fix an error then thats a new skill you get to keep :)

Morning, unsure what you decided for drivers but if going the Leadshine AM882 route I'd order them now took about a month for mine to arrive from Hong Kong.

You can buy EU equivalent Leadshine EM806's but they're twice the price and the specifications seem identical both made by leadshine. AM882's you will not find available from UK seller if you do link me to him please lol so i can buy another 4. The only difference I can see between the EM806 and AM882 is they have different dip switch settings relating to auto motor setup but both support autotuning making this an non issue.

Don't buy 2nd hand like I did didn't save enough money to pay for the non working one I got I ended up buying a new one for £60 delivered, quick tip reduce the chances of important tax by making 4 separate orders to 2 different address's.

Personally I'd order the Hiwin's and ballscrew in at same time and from personal experience this time I would make the Hiwin's 20cm longer than the ballscrews. I'd want the Hiwin's and ballscrews before I started to make the frame after all the frame has to fit the Hiwin's and ballscrews not the other way round. You can't practically cut ballscrews down I rang a company in Leicester that specialise in ballscrew repairs to see how much to end machine a ballscrew was. It was about twice the price of buying the ballscrew from China with BK/BF fittings for the machining only.

If I had to order the Hiwin rail in UK I would use cncforyou which sells original carriages but a clone rail but I would just order from China/HK as my experiences so far have been mostly positive.

Quote:

---

Originally Posted by Desertboy

Morning, unsure what you decided for drivers but if going the Leadshine AM882 route I'd order them now took about a month for mine to arrive from Hong Kong.

You can buy EU equivalent Leadshine EM806's but they're twice the price and the specifications seem identical both made by leadshine. AM882's you will not find available from UK seller if you do link me to him please lol so i can buy another 4. The only difference I can see between the EM806 and AM882 is they have different dip switch settings relating to auto motor setup but both support autotuning making this an non issue.

---

I would only ever buy Leadshine drives from a reputable source, as there are a lot of counterfeit drives about.

Personally, I've often wondered why the AM series got discontinued so quickly, but I suspect it was to do with forgeries. IIRC Leadshine superseded to the EM series in well under 2 years, but by that point you could pick up the AM drives cheap in lots of places. So either there was some major internal redesign, or they needed to get a step ahead of forgeries. I suspect the latter, and the AM firmware was easy to retrieve (the Chinese are renowned for their ability to extract such things, using methods you wouldn't think were possible), and the EM not so, which is why it's not yet been copied.
Most older digital drives are copies of Leadshine, which in turn were re-designed copies of Geckos, so you can't say Leadshine are just a victim in the whole counterfeit drive area, but they at least build reliable drives, and didn't deny for years they had an inherent design flaw which meant they could randomly blow up during power up (it's a shame JohnS isn't around anymore, as it's what ultimately forced him to use Chinese drives, and he liked a good rant on the subject).

I don't know if anyone has done a controlled side-by-side comparison of the AM882 and the EM806. I bought the EM drives for my machine as I was fed up with buying things that seemed to go obsolete almost as soon as they were delivered and the EM series was still fairly new on the street, with supposedly better anti-resonance algorithms, etc. That was around 3 years ago but the AM drives are still going strong! However, I can say that I am very happy with my EM drives - no resonances ever been apparent, and the anti-stall is very useful if for some reason the gantry has gone out of square and loads get too high on rapid feed. Machine stops quickly without tearing itself apart. Would the AM drives have worked as well? Dunno!

One day I'll make up a serial cable and try to set up the autotune properly. Still slightly confused about how that's supposed to work with a dual-motor setup as you clearly can't do it with the motors connected to the ballscrews, but it doesn't seem to make sense to try to tune the motors when they're unloaded.

I have an Am and two EM's on my milling machine and I can't see much of a difference to be honest.

However, for my new router I am using four EM's. As you say, Neale, the stall function really works, I had my milling machine travel out of bounds without a limit switch and as soon as it hit the end, the driver stopped it preventing any damage. Also good for when you make a mistake and go too severe with your cut, it soon stalls the movement and prevents further damage.

My thought is, if you can afford it, buy EM's from Zapp, they are actually quite reasonably priced and it's just a one off expense anyway and it supports a UK firm. If you really are on a very tight budget, buy AM's from China, but expect customs duty, some wait and at the end of the day, you are not saving much more than a round of beers, if that.

Quote:

---

Originally Posted by Neale

I don't know if anyone has done a controlled side-by-side comparison of the AM882 and the EM806. I bought the EM drives for my machine as I was fed up with buying things that seemed to go obsolete almost as soon as they were delivered and the EM series was still fairly new on the street, with supposedly better anti-resonance algorithms, etc. That was around 3 years ago but the AM drives are still going strong! However, I can say that I am very happy with my EM drives - no resonances ever been apparent, and the anti-stall is very useful if for some reason the gantry has gone out of square and loads get too high on rapid feed. Machine stops quickly without tearing itself apart. Would the AM drives have worked as well? Dunno!

One day I'll make up a serial cable and try to set up the autotune properly. Still slightly confused about how that's supposed to work with a dual-motor setup as you clearly can't do it with the motors connected to the ballscrews, but it doesn't seem to make sense to try to tune the motors when they're unloaded.

---

That's what we need is a side by side comparison, when I build my next router I might go with EM806's (Although planning the easy servo drives with matching motors) just to compare with same toroidal PSU setup. They are plug compatible so could relatively easy swap them over just for testing.

They don't sell EM806's on alixpress and the suppliers I contacted said they don't sell them in Asia they are only sold in US and Europe but I always take what suppliers say with a pinch of salt lol.

On the clone issue I think it's how good a clone is it? If the clone performs and works exactly the same then the issue is slight, if it's a poor clone then problems occur.

I have no idea if my AM's are clones but the internals look identical when I stripped them down and they all seem to perform the same. Almost silent performance if I set the microsteps to 32.

https://www.youtube.com/watch?v=O4_u5ov-9cc

I'm going to set them to 8 for now which makes them sound like normal whiney steppers but more torque of course.

Customs duty is fickle but split your purchases and send to friends to eliminate them for stuff like this. I paid £15 on my first 2 and nothing on the next 3 and the one that was faulty I probably blew up as I also blew up a toroidal transformer so that's down to me being crap lol.

I can understand what your saying but you can buy 4 New AM882's for £240 delivered vs £440 for the EM806's that's quite a few rounds of beers in Leicester lol. If I was on a budget it's here I'd save my £200 to spend elsewhere say on PSU or upgrade the ballscrews/Hiwin sizes.

The limit is £40 from China I've found split the deliveries and anything under £100 you don't get charged. I spent less than £70 on tax on my router build and over £1200 from China and most my tax was the ballscrews.

I think the £200 buys you a replacement EM806 if one goes wrong which is a good warranty for me ;)

I too have crashed 2 out of 3 axis on my router and stall protection kicked straight in ;) so if I have clones I have working clones.

There is after all 2 ways to clone this drive
1. Make an exact clone with same chips or equiv and use the same firmware, done right the difference between the original and clone is minimal and sometimes clones are actually better.
2. Put a cheap driver in a copy AM882 box and the stall protection ports are not connected to anything lol. Hack the firmware so the dip switches make sense.

This machine is crying out loud for servos. That's my opinion.

Quote:

---

Originally Posted by Boyan Silyavski

This machine is crying out loud for servos. That's my opinion.

---

What's your opinion on
http://www.leadshineusa.com/productt...y-servo-motors

If I understand it these are not real servos but somewhere in between

Quote:

---

Originally Posted by Desertboy

What's your opinion on
http://www.leadshineusa.com/productt...y-servo-motors

If I understand it these are not real servos but somewhere in between

---

I decided long ago - No to closed loop steppers. Too close to real servos price. 750w servos from BST will much better. Mine machine is all over 400W Samsung servos and i could not be more than happy with them. I doubt this machine will be heavier or much different than mine anyway. If i have to repeat, i will go the same way, like new used second hand brand servos from reputable seller, together with cables and so. If i am making the machine for a client, then new 750w chinese servos. But that's me.

PS. Just as a side note, brand servos like mine and the panasonic will spin to 6k rpm happily, as opposed to cheap servos. So actually you will never use them at 100%.

Cant decide what i will need until i finish my gantry design and report back as the way i am going like boyan said is i may need servos which i was trying to avoid, what price range am i looking at per axis setup for servos? the hiwin rails and ball screws are whats taking up a lot but i am in no rush

edit.. just going to point out just in case someone can rescue me from over designing which eventually is going to push me to servos lol, the only thing i will be cutting is wood or up to 10mm acrylic and thats it but it does need to be a solid gantry for 3d carving. i will however want to make a small machine to mill aluminium and steel eventually :)

can you mix steppers and servos? was thinking off having just servos on the long axis that moves the gantry

Quote:

---

Originally Posted by reefy86

can you mix steppers and servos? was thinking off having just servos on the long axis that moves the gantry

---

In linuxcnc seems easy enough

https://forum.linuxcnc.org/10-advanc...stepper-system

Not sure about Mach 3 but if I'm going servo go all the way all 3 axis ;)

If you're using standard ballscrew at the lengths you want you'll probably see screw whip with normal steppers never mind servos so you might need rotating ballnuts or R&P to see proper performance from servos.

Google screw whip calculator to get a good idea of what a normal ballscrew can do but it's nowhere near 6k rpm Boyan quoted that servos will do.