Printable View
The frame looks good.
The welds look fine considering you've not ground the mill scale of before hand. Although MIG welding will burn through the scale, it can lead to weld contamination as standard MIG welding can only burn of a limited amount of contamination.
Thanks for the heads up. next time will not only grind 1mm the corners but as you say.Quote:
Originally Posted by m_c
Here are 2 recent pictures. Just soldered the hardstops for the gantry from 100 UPN profile. My friend insisted on soldering 4 short legs, again from UPN100 profile. Strengthwise i would have soldered 6 legs instead of 4, or none at all and lay the machine and thin layer cement the machine on place but i was very curious of the deflection of my design.
Now just measured it so here is the result:
1. Applying 100kg static force/me/ on any one of the bead beams leads to 0.045 mm deflection in the middle
2. Applying the same 100kg to middle of the raised sides supported only at both sides by 9 cm legs from 100UPN profile leads to 0.03mm deflection on that side.
Conclusion: With gantry and bead mounted in real working conditions the overall vertical deflection of the machine will be unmeasurable or less than 0.01mm which is irrelevant especially with 0.8kw spindle. So first impressions are perfect.
Anyways, i will most possibly insert pads in the middle of the machine when placing it on its place, so it will stay rigid and quiet
PS. In fact using the excel file for calculating deflection on gantry and gantry side, i guess the real final value of deflection in all directions of this particular design of the finished machine will be 0.0035mm. Which i believe is great.
Attachment 10414Attachment 10415
it looks really warm there... ive just returned from holiday in turkey, to miserable cold england. :(
top work... keep the photos coming...
Your test shows that the deflection due to the frame will be very small (<0.01mm), but remember the linear guides and Z-axis especially will all lower the stiffness, so overall your deflection will be greater. Still, your gantry design looks reasonably strong so I expect it will be fine.Quote:
Originally Posted by silyavski
Thanks. I am aware of that. For me this build is a preliminary build to my real machine. A kind of experiment. Something like how far i could go at every step.Quote:
Originally Posted by Jonathan
As the machine we all know is a sum of its components.I constantly meditate on "the machine is strong as its weakest part". The Hiwin blocks are with A preload, the deflection calculator tells me the gantry sides deflection cutting aluminum will be lower than 0.01mm, so i guess the weakest part will be the Chinese spindle bearings and the bit. I will be extremely happy if i succeed in making a precise machine, just for the sake of it.
I need some help here with the Leadshine AM882 drivers and the SY60STH86-3008BF motors from Zappautomation.
-As far as i understand the auto finding of the parameters using the dip switch would be enough?
-However i wonder what step/micro-stepping should i use with the 1605/z/ and the 1610/x,y/ ballscrews? having in mind the way i build the machine.
-Also about the Switch 8 at the drive.
the manual says:
Pulse signal: In single pulse (pulse/direction) mode, this input represents pulse
signal, each rising or falling edge active (DIP switch configurable); 4-5V when
PUL-HIGH, 0-0.5V when PUL-LOW
So i asume rising is Pul High and falling edge is Pul Low. I wonder which one will work with my C23 board from CNC4PC
The microstepping setting is generally first limited by the output frequency of your parallel port. Using higher microstepping can help with reducing resonance problems, although those drivers have more advanced ways to deal with that so it's hard to say. Generally people don't use finer than 1/8th microstepping as going further can reduce performance, but if you can even use that will depend on the maximum frequency your parallel port will output and the feedrate you require.
For example, suppose you want a feedrate of 8m/min (sensible for cutting wood) with the 10mm pitch screw driven via a 1:1 ratio with a parallel port that can only output up to 25kHz (a lot will do more, but not all):
Screw angular speed: 8000/10=800rpm
Base frequency: 800/60*200=2667Hz
25000/2667Hz=9.4 ... so nearest is 1/8th microstepping with the above conditions.
http://www.mycncuk.com/forums/machin...-mach-3-a.html
From what I've read 1/8 is a good starting point
RE. switch 8 I guess if it won't work one way then try the other
Well depends a lot on your PC and how good the parallel port is and what speeds your looking to achieve. PP and PC speed to some degree will determine the Kernal speed in Mach3 and how many pulses you can put out which in turn will determine the speed you'll get from the machine. Higher micro steps require more pulses to achieve same speed has lower MS but they give smoother running motors and to a lesser extent some resolution.Quote:
Originally Posted by silyavski
Personally I don't go higher than 10x or 2000ms and this is my preferred choice if PC can handle it. Higher MS also helps with resonance but these drives are great with handling resonance so it won't be a problem and if it is you can change tune it out using the software.
I would agree but check with Arturo and I'm think it may even be possible to change a jumper on the board has well to suit but either way your covered because the drives allow it.Quote:
Originally Posted by silyavski
If you can't find out then you'll have to make some witness marks run some G-code and then check if it's lost position. If the pulse is on the wrong side it will drop a step on every direction change so won't take long to find.
Bloody heel all that typing and Jonathan had already said it all . . Lol
Not that simple Eddy it will work in either mode but if pulse is on wrong side then it will drop a step on direction changes.!Quote:
Originally Posted by EddyCurrent
Thanks a lot guys!
Yes 300ipm / 8mpm/ is the speed i aim considering the tooling/price and the spindle speeds.
8 microsteping was what i had in mind. I have tested the PC upto 100Mhz as far as i remeber and it was Ok. With other BOB, but i assume will run with this also.
Anyways i will make to files for Mach3, one for 100ipm and one for 300ipm, as the guy is new to this, so at first untill he is aware what he is doing...
On my small commercial machine the z steps in mach3 was set to 10000? this is quite high micro stepping, thats why i wondered, on the other axis is 8 microstepping. reverse calculating it the microstepping in z is 256, i am not quite sure now for the 10000 but it was ridiculously high.
Don't run that high thou otherwise high chance of troubles. Only set the Kernal speed high has needed and no more. 25Khz is the best and every increase above increases chances of trouble. Wouldn't go above 60khz no matter what driver test says or PC thinks it can do.!!Quote:
Originally Posted by silyavski
If you need that high pulse rate then buy motion control card, in fact I'd buy one anyway if you want the best performance from machine.!
Hi guys,
thanks to all for the valuable inputs.
For now i have to wait until they cut with laser the pieces, so i can proceed further.
Meanwhile as a thanks to all for the help :encouragement:, here is the Sketchup file/its in winrar format for easier transfer/. So anyone who wants can use it as it is, use parts of it or just use the components or table drawings. Before ordering the laser cut parts i revised it piece by piece and against the ballscrews, housings, etc. So all small errors are fixed. It cost me a whole day to rework it precisely. The empty spaces are for epoxy, there are some clearances where necessary. The machine has work area of 400x1000x200 . Any questions are welcome if you have something that is not clear. The main point being here is that the machine is completely scaleable up to the limit of the ballscrews or with minor adjustments even up to 4x8 , this without changing the type or thickness of the profile. Other main point is that the drawing is final drawing, no need for correction before real assembly, however there are some steps to be followed, but easy to figure when playing and simulating the actual assembly. If by channce there is a small error i will fix it right away in the future.
Attachment 10608
So enjoy it and thanks again. Will update when i have more to show.
Hi Silyavski,
Thank you very much for putting your sketchup files up for everyone. I am in the process of designing a new machine and learning sketchup so your files will be a big help! :)
What sort of rough price does the laser cutting cost?
Looking forward to seeing your machine come to life!
I dont understand something
The cncrouter will have legs down to ground or the cnc on the left of your scetchup will be put on the white table
Hi there,
The table on the right is if sb wants to upscale it to 1000x1500 or similar,with small adjustments, using the gantry from the left. On the left is the complete machine 400x1000 which i am building now.
Furthermore have in mind the gantry can be simplified removing the raise sides if intended mainly for aluminum.
I ordered them in my country/ Bulgaria/ and they cost there 250 euro. + 50 for a guy to bring them when he is coming back from there in 2 weeks.Quote:
Originally Posted by cncJim
Here in Spain, the price would have been outrageous.
Anyways, its a lot of metal. I don't have the means here, that's why i preferred laser cut. But in fact if you can cut and drill and have spare time, they are quite simple to produce at home.
Now i am designing my next build and definitely will try to lower the number of parts.
Let me ask something
I am planning my cnc 1500X1000 and i am very close to these plans
Is it better to weld everything together or make seperately the base and then the cnc machine that will "sit" on the table. I am asking because if i weld evertything together my construction will be very heavy ( difficult moved even by 3 or 4 people )
What is better
Solution 1
one unique construction ( both table and cnc together ) or
Solution 2
two seperate constructions( the cnc and the table). Will this second solution have stiffnees issues ?
Thanks for your time
By mistake i upload the same reply two times. Sorry
Hi,Quote:
Originally Posted by ba99297
I personally chose solution 1, welded together. I prefer playing with additional fixtures, instead raising, lowering the bed.
Other people here on the forum however choose solution 2 . There will not be stiffness issues.
At the end is up to you. Now i am building another 1250x2500 and i also have the same doubts as it will weight considerably. Its worth noting that if not welded together, possibly you will have to use some more beams and at the end it will weight even more. For my 1250x2500 i am contemplating making a hybrid, half of the bed welded, half removable.
It is possible i didnt make my self clear beacause of my english
Syliavski i will use your photo to explain
When i say cnc i mean the element A of your picture
and when i say table i mean the element B of your picture and not the cutting table
So the question is
Will i have any stiffness issues if make two different construction A and B and then let A rest on B
Or i must make one construction both A and B together in order to have more stiffness.
The problem is that if i make one unique construction it will be too heavy
Thanks for your time
Attachment 10954
Thanks for your time
Hi,
the element A is a benchtop cnc 400x1000x200. The element B is not a table. It is another CNC, 1000x1500x200. It is the same, just scale the gantry.
So A is strong enough even if you put it on a simple table/strong enough/ . There will be no compromise of stiffness. Same with B.
In other words just follow B design if you like it. It would weight 130kg if following the design. Look at this thread also: http://www.mycncuk.com/forums/router...ddy-cnc-2.html to see it in real life
Silyavski i know that element b is not only a table.
I just use it as an example of table in order to give you understand what i mean
So you tell me that if i make a benchtop cnc 1000X1500 and bolt it on a table ( whatever strong table that can handle such weight ) i wont have any problems right?
Ok, I get it. If you use the same design and make a simple rectangular frame with the legs and diagonals on the legs, it would be rigid enough. If you use profile smaller than 80x80 you should use other design.Quote:
Originally Posted by ba99297
Talking about the frame or any other frame for that matter,
If it's to be painted, which would be a good idea, do the rails sit on top of the paint ?
If epoxy is to be used under the rails or part of the associated support structure is the epoxy under the paint or on top of the paint ?
I'm asking because paint has thickness and is not rigid.
I am still waiting for a friend to bring me around new year the laser cut parts. Will solder, epoxy, mount rails and paint. In that order.Quote:
Originally Posted by EddyCurrent
Let me explain. This machine exactly i am making for a friend who is pro model maker / yachts/ . First i suggested him to spray with gun Henkel machine paint/ anti oxide and paint directly on metal/, which i use and is very good. But the guy is a maniac, and honestly i felt ashamed when i saw his perfect yacht models which any watch maker will envy. So lazy me, i pretended i don't know how to paint a car :-) and he will come to help me with the painting. It would be painted the way the cars are painted- perfectly with glossy double component car paint and all the stuff that comes before that. Will document with pictures the process.
At the end i thought that it is good that somebody raise my bar, when perfection is expected. I will do so with the other machine i am building. Yes, 2 days job but imagine the hours i will be happy looking at it.
Basically the plan is to base paint the metal, polish, thin base, polish, 3-4 hands, laquer 2-5 hands and polish after a week or so when dry. I will use Standox car paint or Duracoat. At the end, looking at all the time and money that will go there, why not make it beautifull.
Forgot to say that the epoxy can be colored black for example, a smallish bottle of like 20ml will color 1l. I doubt it will affect its viscosity, the small bottle i have at home looks quite liquid, like water. Its special epoxy color though, not general type. That would be great if its difficult or not possible to paint it later, or the frame is painted beforehand.
Well, the laser cut parts came. So, i finished soldering the parts to the frame. That includes the ball screw mounts. What i did was mount the ballscrews to the supports/in the middle so there is place left for adjustmennt/ and that greatly helped the correct placing. Carefully made some tacks with the Mig and waited all to cool, so checked again if the screws move smoothly. Have to note that the machine was welded with precision <1mm greatly helped.
Attachment 11369
Attachment 11370
Now its epoxy time. Several mistakes here. Just to fulfill my worries.
I used some aluminum angles, hot glue from outside, instant glue below, silicone at the angle inside.
the mistakes:
-the channels connecting both sides should be wider, say at least 3cm, not like mine-8mm :-). Epoxy has difficulty to flow in so narrow channels
-i am epoxy greedy, calculated the just amount. had to mix second time to fill the channels
-scratched the base below the epoxy, with the idea of better contact. Now i can see the scratches. nevertheless i would paint the machine, but is good to know that the epoxy is quite transparent yellowish color
The clever stuff:
-the epoxy sides will be 40mm wide, works great for leveling.
-it seems the way i fixed the channels holds pretty well , cust came from the garage, no pour for now, one hour later we will see.
-used bolts on the legs to level the machine perfectly
-as i wanted 5mm epoxy thickness, i found some magnets which together made 4.5mm and put then at the corners. This was extremely clever as right from the beginning i could see how much epoxy and where to pour. In fact i believe i poured it perfectly and did not need the channels, as there was a moment when the epoxy just covered the magnets.
-used small torch to help leveling and take away the bubbles.
Attachment 11371
Attachment 11372
Attachment 11373
Other clever stuff was that i welded 2 wheels at the back and fixed 2 handles at the front. That helps me greatly maneuver the 100kg around.
Very neat. Keep the piccies coming. G.
Good job! Thanks for taking photos, sometimes I find it hard to visualise what people have done. The epoxy looks very neat, I would love to see some pictures of it when it's cured. How do you pour it? Just on one side and let it flow down the channel to the other, or fill both sides at the same time?
Did you weld the frame yourself? Looks very good.
Jim
Thanks!
I filled both sides at the same time. The magnets at the corners helped as they were 4.25mm thick, as i could clearly see how much i should pour to achieve 5mm thickness.
As the channels that connect the sides were very thin and i have not mixed enough epoxy, had to mix second time and pour in the channels.
if i was to turn back time, i would have used the same aluminum angle profiles, fix them with scotch only from outside and use cheap epoxy with a brush to make a kind of base, which would close the distance from the channel to the profile against spill. When dry and so secured, pour over the expensive epoxy :-)
Yes, i welded the frame , the process is described at post #43. My second welding job in fact. Good that i was careful there, now things are much more easier that could have been with not so careful welding. used a cheap 250A Mig which i obtained from ebay+C02, hence the not so smooth welds.
Yes the magnets seem a good idea. If I was doing the epoxy I think I would have also used the thin channels, so I am sorry it didn't work out as well as you intended but I am pleased you shared the results so others (me!) can learn from it.
Sorry i thought I had seen all of the posts but I guess I have missed a lot!
As Jim says, the epoxy looks very good and there's no way that frame is going to move.
If using the self-leveling epoxy method, then the only way worth pursuing is to have a channel between the two rails (sides) and ensure sufficient depth of epoxy to flow between them. If they were poured separately, then the sides would end up at different heights, potentially resulting in significant stress on the bearings when the gantry is attached, unless shims are placed under the gantry on one side.Quote:
Originally Posted by cncJim
I know that I have posted these before but this is the way I have done it. The epoxy is the Wests system Very slow cure. ..Clive
Now something very important. I will not move the frame at all. After a week has passed will solder the gantry and sides. Then will turn upside down the gantry and lay the upper side on the level surface of the epoxy. Then will pour epoxy on the gantry side plates that step on the Hiwin bearing blocks, connecting them together. Also pour epoxy on the lower side of the gantry beam. When dry, will turn the nagtry to normal position , rest the leveled side plates / epoxy on epoxy/ and pour the final epoxy at the top side of the gantry beam.
So with these steps i will achieve all hiwin rails lay on epoxy surfaces that are paralel to each other. The only thing left will be to ensure squareness of Z. I hope to achieve a quite precise machine.
Well, that's an obvious way to set up the epoxy on the gantry. You know, the kind of "obvious" that's only obvious when someone else says it!
I've been kicking ideas around in my head for the last week or two about how to align the Y rails for "constant spacing" if I put them on the top and bottom of the gantry bars for my new design. Easy enough to lay the gantry on its back and use epoxy to level the Y rails on the front face, but I couldn't work out how to do it where the fixing faces for the rails were not in the same plane. Easy enough if you have an accurate level surface to put the gantry on and do one rail at the time but how do you generate that initial level surface in a home workshop? So use one of the level surfaces you created when you used epoxy to set up the X rails! As I say, it's obvious once someone else has pointed it out...
In practice, any idea of how much out of true a typical length of, say, 50x50 or 80x80 steel tube would be? Are we talking about a mm or two, or significantly less than that? How much does the epoxy thickness vary, in other words?
This should help http://www.parkersteel.co.uk/media/p.../JPSBS0010.pdf I was looking for it yesterday when I wanted to look at how "true" my beam might have been. Don't forget you're also at the mercy of the stockholder and transport to keep the steel flat and not to induce any additional bending/torsion etc...Quote:
Originally Posted by Neale
Very interesting build, and good research for me. A weld bead causes shrinkage where it is deposited, so weld induced distortion is not so much due to the 'heat' that goes into the part, but is due to the fact that a weld, when solidifying, pulls on the adjacent parent metal. Apart from looking nice, I'm not convinced these types of bed frames, with massive members, need to be fully welded at the joints, from a stress point of view anyway - stitch welding would be perfectly adequate - having said that, the effect of inducing bending due to weld shrinkage is possibly only noticeable on smaller dimension square hollow section, less than 70mm say. The effect of weld shrinkage induced distortion may be mitigated where possible by welding to the perpendicular plane of the material, if that plane isn't as critical for flatness, and the design configuration suits it - I don't think it applies in this particular build case though.
Sorry to theorize in your build thread - being a noob, I guess I'm eager and this seemed an appropriate, and reasonable relevant place.