View Full Version : Built on a foundation of past mistakes.

09-25-2004, 10:08 AM
This is the project log for my next machine.

This next machine will be built with the lessons of previous mistakes in mind.

While my first machine is/was successful there are a few annoyances that I wish to illiminate in the second machine.

The first machine does not take a standard sized sheet of MDF
It has rails on the sides that make putting the work onto the table a bit awkward. I want a table where I can slide a sheet on.
It does not have limits switches or zeroing swithches.
The X axis is belt driven from one side only. I plan to have a drive shaft through the gantry and drive it from both sides.
The gearing down is not enough to get maximum torque out of the servos. Even though the machine works fine the motors are only running at 20%. Instead of 3:1 ratio I will be trying double reduction to get 9:1. I figure this will give the machine greater torque and also greater control over the gantry movement at higher speed. At higher speed I noticed partial overrun of the gantry.
The visual appearance is poor.I am thinking of ways to hide as much of the working components as possible. If I can make it look like a block then this will be good. ( The volvo look )
It does not have a good hold down system. I plan to incorperate a vacuum hold down system. I have come up with an idea that will allow solid hold down no matter what the shape.
The spindle is not controlled by the computer. It was manually turned on and off.
There is not dust collection setup

If you are like me, you dive in and make many mistakes, disregarding much advice that others give, only to find that you may have a few regrets down the track. On the other hand, I hope this thread will assist future builders with their designs. From reading this post so far you might be thinking my first machine was a total failure, in actual fact I am quite pleased with it and make a living from its work. I just want to improve on its features and usability.

I took a few contraversial chances with my first design that I have proven to myself to be quite acceptable, and will be keeping with the new design.

Using timing belt and pulleys for movement, instead of ball screws.
Using V bearings on the edge of angle iron, instead of traditional THK rail or linear bearings.
Using allthread for my Z axis

All of these comprimises have worked well and have saved me a lot of money.
I will continue to use these methods.

I expect this next machine will take me 1 month to complete. I will post lots of pictures along the way.

This will be my mentor machine along with a few of my own modifications.

01-20-2005, 07:29 PM
Well if you read the blurb at the start of this thread you would expect a state of the art machine to be made. The fact is I comprimised nearly all my objectives. Oh well, I have finished the machine and it still has that ynneb piece of crap look about it. It works though, and thats the main thing.

It has a footprint of 2m X 4m. I made it this big so as to never get caught out with a job that is too big.

Here is a pic of the machine and a sample of its cutting ability.
The sample is a cheap test cut only. I will be making and selling these cabinets with black memamine and T moulding etc.

Sorry I didnt make progressive posts, the truth is I am a embarresed at its appearance. I am only posting the pic because another member has encouraged me to do so.

PS. I know I made a big thing about not posting in this forum anymore. I guess that was a bit ambitious too. Maybe the comprimise will be just a heavy reduction in posts :)

EDIT: that funny picture of the bearings is a ynneb poormans idler.

Oh, and if this thread encourages anyone, it can be that you can make a machine with no special tools other than an angle grinder, welder, and a drill plus a few hand tools. (no mill or lathe required, just be prepared to comprimise)

I have found that to cater for lack of precision, you can bolt everything together with over sized holes, and adust things to suit as you go. Once you are sure about the adjustments, you can then weld it into position.

<iframe title="YouTube video player" class="youtube-player" type="text/html" src="http://www.youtube.com/embed/1HxwgAV8On0" frameborder="0" height="510" width="640"></iframe>

01-20-2005, 07:32 PM


01-20-2005, 09:42 PM

Glad to see you!

01-25-2005, 02:41 PM
Benny, there's no need to be embarrased at how anything looks. All that is important is that it works. I'm impressed to see that I can cutout arcade cabinets and aquariums.

I know that you want to limit your time spent on the forums, and for a good reason, but will you post a few more pictures of the critical working components?

Your machine is a true inspiration to those who don't have access to machine shops, and want to build their own cnc router.

01-26-2005, 03:16 AM
Be warned, if you want to see fine engineering then dont look at this post.
If however you want to see engineering that achieves the desired result at a low skilled, low budget price, then look further.

01-26-2005, 03:25 AM
Here is the very first thing I cut with the machine.

It is a desktop arcade game control panel. It does not use the ipac converter, it uses a 10 dollar usb hack of a game pad.
This was another test cut. The finished product will also be made out of black melamine and t-moulding along the edge.
The only reason I have posted these pics is because some may doubt my machine could work or cut very well. It probably doesnt have terrific accuracy
but I guage the accuracy by seeing where the cut starts and where it finishes. If there is no noticible ridge then I think it is accurate for what I need it to do.

One day when I have enough time and some spare bucks I want to cut a Dalek and a tardis. What I will do with them once I have them I do not know.

01-26-2005, 05:32 AM

Fyi, I was impressed by what I saw when u invited me over during Xmas. I'm amazed how little time it has taken you.

Your common sense approach worked real well for me and has influenced my thinking. At the time, I had been labouring in my thoughts about all sorts of complicated things which I've now been able to disregard... Thanks.

Looking forward to showing you my machine which will have a number of ynneb features in it. (Yes... I will post pictures to the Zone!!)


Fyi, CK87028 was the chassis number of the bicycle I was given for my 11th birthday.

01-27-2005, 07:25 PM
Be warned, if you want to see fine engineering then dont look at this post.
If however you want to see engineering that achieves the desired result at a low skilled, low budget price, then look further.


What size and pitch belt are you using for the track and what is the diameter of the cog wheel? Just for a reference what size are the servo motors? Also where did you find the v-groove bearings?



01-27-2005, 11:58 PM
The belts are 16mm wide and are T5 ( I believe)
The big pulley is 60 tooth and the small one is 20 3:1 reduction. I should have it geared down more to get maximum out of my servos. But it has come down to econimics and time. I will do a further reduction later. All it means is I have the acceleration and decelleration truned down a bit. It all still works fine though.

I dunno what the servo ratings are other than they are driven at 60volts and pull 10 amps each at maximum speed. Does that mean they are 600 watts each? I dunno. Anyway they are an ebay special that were new , but excess stock for big real 2 real computers. They have 2000 step encoders on them.

Here is a video of the router in action. I was cutting a freebie job for the Salvos who are doing some kid related project. If you cant work out what I am cutting, they are meant to be large flowers.

01-28-2005, 02:36 AM
P = IV
60 X 10 = 600 watts
Good job on the volunteer work!

01-31-2005, 06:32 PM
Hi Benny.
Your work looks great to me and if it all works as planned then your on a winner. Don't worry how things look for now as you can always make it look better with a "coat of Jam" (paint) at a later date. I have an almost full workshop and it has taken me almost 2 years to get where I am now with my router and I'm having a few problems along the way with torsional twisting of the frames etc so I may have to rework the whole thing later. I have a small mill drill with a geared head that I use for all my machining but it has a horrible round column that bends over 15 thou when raised to the top so I can't mill anything square on the damned thing. Very frustrating. I plan to get a loan for a Bridgeport with DRO fitted, today so I CAN do some accurate work. A secondhand one costs $5000 to $8000 over here.
Just keep building and putting up your pics and we will keep drooling over someone who can accomplish something in a month instead of years.
I work 11 hour shifts 5 days a week so there isn't much time to keep the house clean and feed myself let alone get in the shed, but I do make time. I will try and add some more pics of my setup soon and you can all see where I went wrong. This helps others.............doesn't it ?
warm regards Richard.

01-31-2005, 09:08 PM
Hey Guru....(Benny).

it is amazing that you completed it so quickly. great achievement. Thanks for posting all of the details, warts and all.... we can all learn from our mistakes. Anyone who says that they didn't make a mistake is either a Liar.. or hasn't done anything. Hopefully when mine is finished I too can share my mistakes and reflect on how I would have done things differently.

Hopefully all of the 'New' buidlers can take inspiration and know it is all worth it in the end when you see your little 'baby' take its first steps....


10-23-2005, 08:19 PM
Enjoyed your post. I am getting ready to do my first machine and I had been thinking about using many of your techniques, so it was most reassuring (you know how some of your bright ideas just don't seem to work as planned!). While I found your discussion of the gear ratios interisting, I am curious to know what speed motors you used/would use next, and if your measured the breakaway torque on the axis'.

10-24-2005, 11:06 PM
Benny...that was pretty fast....great job!

I'm still working on mine....darn designing and machining ballscrew journals has really slowed progress. Got the machining under control after annealing the end of the ballscrew.

I finally broke down and ordered some bearing blocks from Mitsumi, so I can move forward.

Jason Marsha
05-03-2006, 08:12 AM
I am in the process of building a 5X10 router and I would like to know if you had any problems with the belt stretching.

Is there a special grade of belt to purchase?

Have you replaced your belt due to wear?



05-03-2006, 10:36 AM
Jason, my belts have not stretched yet and are still like guitar stings if plucked. The belts plastic with steel impregnated cables in them. They are perfect for cnc routers, but probably would stand up to machinists standards when working in the 100th on a mm catagory. Not only this the belts have zero backlash due to the nature of how they are etnsioned on the pulleys, and the way they interlock with the pully. I guess in real terms there is a minute momentary stretch when triyng to get the weight of the gantry moving. But in practical use this is immeasureable and make no difference to the quality of the cut.

Jason Marsha
05-03-2006, 12:06 PM
Thanks Benny,

It seems I have a lot of choices of belt material and configurations to choose from. I am partial to the GT2 series , 5 mm pitch belts from SDP/SI but these are neoprene - nylon covered fiberglass reinforced only. The HTD and T series are available with the additional option of polyurethahe-translucent reinforced with steel tensile cords which I assume is stronger but the minimum diameter of the pulley to be used is larger than that of the neoprene belt due to flexibility.

What is the diameter as well as the number of grooves used in the pulley that your long belt runs on?


05-03-2006, 12:17 PM
Hi Benny,

I'm curious if you would have an opinion about a light weight machine running belts in an axis length of say 24 ft? Could this be done or would it be better to rack and pinion?


05-05-2006, 08:15 AM
That is one cool machine!

05-05-2006, 08:50 AM
Mike, I really have no idea what would be better over 24 foot lenght, other than I selected belts for cost saving over Rack and pinion. I do expect you would need very wide belts in order to compensate momentary stretch. You wouldnt have that problem with R&P.

Thanks Netjams, you are easily impressed. I would rate my machine amongst the lowest compared to other members machines, in terms of build quality. But thanks anyhow :)

Jason Marsha
05-05-2006, 09:42 AM
Benny, I think you rate your machine to low as I have not seen many who have a working router that big complete with supporting table and is reliable enough for you to source outside jobs. I am sure your machine does what a lot of much more expensive machines do and just as well, shopbot,ezcnc etc. I know of an AXYZ machine that just cuts plastics and mdf for signs, I am sure if they had the knowledge they would have tried to build one themselves.

After Spalm linked me to your thread I decided to use belts for my long axis when I saw what you did with yours, which now makes my task a little more simple.
One great thing about designing and building a machine yourself, you know how to fix it yourself.


05-05-2006, 10:07 AM
think you rate your machine to low
Probably I do. But you know when you build something yourself, you know of all the mistakes and flaws. The other thing is, if you rag something, you dont dissapoint anyone when they see it. I recon its better to undervalue something with the chance that others think better of it, than overrate it and pople thinking what you have done is crap.

12-02-2007, 06:41 AM
ynneb, you have to give yourself credit for building a working machine at least,especially in the time that you completed it in, I'm building a 4 x 4 router at the moment and its work like yours that inspires and helps the rest of us. I'm considering using belt drive on my x axis now after finding out about the success youve had with your machine, keep up the good work.