View Full Version : Seating the Y Gantry on the X Rails - correcting a twisted gantry

Greg J
Thu 13 March 2008, 07:17
This has been bothering me for a long time, so I had better ask rather than remain silent.

My Y gantry has twist. Only 3 of the 4 rollers make contact with the X rails. When I assembled everything before painting, I placed shims (washers) on the X rails in different locations. The X rails are very flexible in the Z (vertical) direction and all 4 rollers make contact at the same time.

The issue in my head that won't go away is theres a wave patterns in the X rails. I place a washer (approx. 0.100 inch thick) [2.54 mm, duh :)] under a bolt that attaches the X rail on one side. The opposite bolt on the other rail is flush with the C channel. The next bolt on the first rail is flush with the C channel and it's opposite has a washer/shim.

With this wave pattern in the rails, I can get all 4 wheels to make contact. But is this the correct way to seat the Y gantry? What is going to happen at the cutter tip. Will I have a wave pattern in my work or no?

Is it better to shim the X rails or cut metal on one hole of the Y gantry where a V roller is attached so all 4 wheels make contact on an un-shimmed rail?

Gerald D
Thu 13 March 2008, 08:30
Greg, we are going to have some fun working through these issues with you ......all will come right in the end :)

First, a gantry twist is never fixed by shimming the x-rails. If all 4 wheels of the gantry don't contact the x-rails, then the cure must be in the gantry only. We will come back to the gantry twist later . . . .

First thing to do is to get the X-rails back to straight, level and parallel to each other. A shim for the the rails is a fairly big thing - it must support the rail over a big area. The x-rails must not be "flexible in the z-direction" they must be carried by the channels, or the shims - you should not see much light through between a rail and the bed it lies on.

Once you have the x-rails all parallel and straight, which may also entail adjusting the feet/legs to the floor to rake out twist of the whole table, lay the gantry upside-down across the x-rails. The gantry should rock (caused by its twist) very little when lying like this - you should be able to seat it with the pressure of one finger.

If the gantry is really twisted too much, then you have to twist it straight: Put about 1" thick wooden blocks under 3 corners of the gantry and clamp those 3 corners down to the heavy channels under the x-rails. The one corner without a wooden block is one of those that were high to start with. Now pull that high corner down with a 4th G-clamp, twisting the gantry for correction. You might have to pull it 1/2" down before it gives and corrects - pull down a bit, release, and check if it twisted. Small patient steps - every time a little more twist till it gives. You can do it scientifically with a digital vernier to see how far you want to go and how far it springs back. Eventually it will come right. Don't worry if your first attempt twists it too much - move one of the three blocks to the open side and twist in the other direction. This procedure does not damage or weaken anything in any way.

After the gantry is untwisted, flip it over and check if the 4 wheels ride right. If a particular wheel is consistently high, and you cannot seat it with the force of one finger, you need to file the holes in 1020452/1020451 to move the rollers slightly, until they seat.

The one finger spec is because the spring tension on the motor plates will have a much bigger effect than one finger, and they will do the final seating.

Why bother to twist the gantry if the roller's hole can be slotted? Well, we don't want the y-rails twisted relative to each other - that will cause another set of problems. The router might be plumb/perpendicular to the table at one end of the gantry, while lying skew at the other end of the gantry.

Greg J
Sun 20 April 2008, 17:28
Finally back working in the shop. You have no idea how great it feels to be working on the MM and listening to a baseball game in my shop. :D

Anyway, I've got the gantry trued and riding the rails very nice. All 4 V-rollers are in contact. Thanks Gerald, I couldn't have done it without your help.

I clamped 3 corners using C clamps and wood blocks. Since the Channel had angled edges, a small wood shim on the lower edge of the clamp made the clamps hold strong. On the corner requiring adjustment, I used another C clamp and made small adjustments. Overall, I move the corner 1.779" for a 0.253" adjustment. That was too much, so I just went to the corner needing adjustment and made the correction. Very simple and straight forward.

I had to add weight to the table because the table was flexing. I used weights from my tractor (Doctor said I can only lift 20 lbs, and the tractor weights have handles for the gantry crane). I added 200 lbs (90 kg) to the corner that I was making adjustments and its diagonal. 400 lbs total.

I hope that helps anyone that has twist in there gantry.


Gerald D
Sun 20 April 2008, 21:52
So it was the corner of the gantry nearest to the camera that needed to go down?

Nicely described and glad it worked out! I am sure that many others will also use this procedure.

You are a farmer, and should have a tractor . . . . . . :)

Might have to add weight to the table to keep all the feet on the ground, but I am fairly certain the gantry will give before that proves necessary.

Oh well, the tractor did end up being useful . . . .

Greg J
Thu 24 April 2008, 07:07
The same procedure goes for the Y car. The 4 V rollers were not making contact with the Y rails.

I turned the Y car upside down and clamped 3 corners (with wood blocks) to the cross member channels. The 4th corner that was out of alignment was easily adjusted and now the Y car rides great.

The whole process for the Y car took about 1 hour.

Gerald D
Thu 24 April 2008, 08:48
Greg, were the slotted holes in 10 30 450 P too short for you, or were you just keen to practise your twisting skills? (Or, didn't you notice the slotted holes?)

Greg J
Thu 24 April 2008, 22:00

My Y car did not have slotted holes for the V rollers. I checked the drawings, and they didn't call for slotted holes.

Just to refresh your memory, I'm working off the very first set of drawings. It's really not a big issue. Very easy to "true" the gantry and Y car.

I know I've been working longer on this project than anyone else, but ... Oh, who cares. I'm still having a blast.

I'm getting close to cutting wood and watching this beast come to life is more than words can describe. You really have to build one to appreciate the beauty of the design. A work of Art.

Gerald D
Fri 25 April 2008, 00:05
Those "slots" are ridiculously short (about 1mm if I remember correctly) and they are only in the y-car plates, in one of the V-roller holes. I had to file those holes slightly to seat the y-car properly, because the y-car is too stiff to seat down with a little finger pressure.

Putting pressure on the car with clamps to "bend" it right is also a sound engineering option.

Greg, you might be having a slow build, but it is an excellent one! This pioneering work on straightening the gantry & car, with photos, is extremely valuable for the MechMate builders. When you started this thread I was a happy man.....couldn't have wished for a better person to work on this one. The younger, high speed, builders don't have the patience to resolve problems and tell others how they did it. :)

J.R. Hatcher
Fri 25 April 2008, 06:14
First I would like to say I resent that remark about us younger builders :D:rolleyes:.
My parts were some of the first parts that Don sent out and I did notice the slot but only after first thinking it was a mistake. I'm holding 1 of those original cut parts and the 12 mm hole is exactly 1 mm slotted, but only 1 hole on each end plate.

Gerald D
Fri 25 April 2008, 06:55
I thought I had better put that note on 10 30 450 P before that JR kid came along and told me there was a mistake. :p

Greg J
Fri 25 April 2008, 07:25
wow, a slotted hole. Didn't even notice. Next time (if ever) I remove the V rollers, I'll inspect very closely.

I forget to mention that I used a good set of calipers (Starrett) when "truing" the gantry and Y car. Not a tape measure or metal rule.

Fri 25 April 2008, 07:52
Learning curve...
I realized that the better your set up on the x and y axis is, the easier it is to adjust and make perfect the z slide set up.
I don't know what the rest of the crew had to go through, but that z-slide fidgeting the 4 spider mounting bolts and then the eccentrics about drove me batty for a good hour or so getting all planes true. The eccentrics made it REALLY easy to get the y/z axis plane correct. That darn x/z plane was a little more challenging. But heck, getting a precision tool set up is never "that easy" :)

...the better the foundation, the easier the house is to finish - "Tom Silva"

Greg J
Fri 25 April 2008, 08:18
...the better the foundation, the easier the house is to finish - "Tom Silva"

He's one of the best. Really enjoy his work.

I'll be adjusting the Z slide tonight/tomorrow so thanks for the heads up.

One issue for me that will be interesting is one X rail has 0.050" height difference over the 8 feet. The other X rail rides +/- 0.015" over the 8 feet. Don't think this will be big issue since one surfaces the spoil board.

Correct me if I'm in error. :)

Gerald D
Fri 25 April 2008, 08:30
one X rail has 0.050" height difference over the 8 feet. The other X rail rides +/- 0.015" over the 8 feet.

Where do you measure those dimensions? Is it from the cross-supports up to the gantry? If so, then you are bragging and not complaining! :)

Greg J
Fri 25 April 2008, 09:21
I didn't measure from the cross member to the gantry. But I will tonight.

I used a dial indicator with magnetic base on the gantry. The tip of the dial indicator was in contact with the very top, flat edge of the rail. It doesn't make sense, because the gantry should follow the rail with zero difference.

Let me double check again tonight and report results. I've forgoten something and not reporting proper information. I've slept since then. :)

Sat 08 November 2008, 09:24
I will put the picture of the boss on the wall :cool:.

This is the distortion of my gantry :(.

The work I expected ;)




Gerald D
Sat 08 November 2008, 09:37
Sorry to see that twist Yuri. Let us know if we need to explain anything more in Portuguese/Brasilian for you. . . . . .

Sat 08 November 2008, 10:22
The topic is complete, easy to understand. :)

There is no need to explain anything more. Its design is perfect, I error the welding. :o

Sat 08 November 2008, 15:39
Would that kind of twist that Yuri and Greg had be typical for us non welder type people? Do you have any tips to minimize this phenomenon for the non welder people here?

Greg J
Sat 08 November 2008, 17:00
Its design is perfect, I error the welding. :o


I know how you feel. :)

The correction is easy and my gantry rides beautiful today.

Gerald D
Sat 08 November 2008, 20:52
Heath, the non-welder types have a better chance of getting it straight - they might read the tips (http://www.mechmate.com/forums/showthread.php?t=230) you asked for . . . :)

Basically, start with small tacks, follow a sequence that involves a lot of walking , turning, checking and keep checking flatness all the time. The key is the sequence - when one weld distorts in one direction, go to the opposite weld to distort it back again.

Sat 08 November 2008, 21:53
Thanks for reminding me of the weld order. :o

J.R. Hatcher
Sun 09 November 2008, 05:26
One other tip .............. make sure you start out with it right, straight, flat, and square. I used 2 lengths of inexpensive 1/2" all thread rod, that went all the way through the center of the tubes and sandwiched everything together. This way you can put a lot of holding force right where it's needed.

Tue 11 November 2008, 20:02
Success ... the principle :rolleyes:

I followed the method of Gerald and Greg_J.

The first picture is from yesterday ... little result ... but some blisters on hands. :eek:

The second image is of today, using staples G (or C) at the second attempt and the result was positive. :D

The second time, we used 02 MDF (50mm). She was lowered 32 millimeters (approximately) or 0.7 between the sides.

We will see the result after the installation of rails (V-Cap SAE 1045 without tempering) :cool:

I followed the whole sequence of welding. Perhaps missed in the control of distortion.

I thought my solder would not be very strong ... while using the staples G, it was possible to note the strength of the project. Thanks Gerald

I thank all who helped to overcome this obstacle ... we will solve other now.:D

Gerald D
Tue 11 November 2008, 21:11
I almost wish that I had a twisted gantry so that I could do this myself - it looks like fun ! :)

What happens is that 2 corners, diagonally opposite, are pushed up with the wood blocks and the other 2 corners are pulled down with the c-clamps. You actually only need 2 big c-clamps. But it makes the measuring easier if 3 corners are firmly supported on wood, as per the pictures above.

Greg J
Tue 11 November 2008, 22:22
Nice job Yuri. :)

I had the same experience. The first attempts were not so good. Once you get the setup figured out, "truing" the gantry is very easy and fun.

Gerald D
Sat 23 May 2009, 11:18
I almost wish that I had a twisted gantry so that I could do this myself - it looks like fun ! :)

Well, it wasn't as much fun as I thought, it was over too quickly! Used two clamps and a baby trolley jack.

Funny thing was, the gantry was nice and flat before fitting the y-rails. Once the y-rails were screwed tight, the gantry rocked. Didn't expect the rail fixing to pust a twist on the gantry, but it sure did - still trying to understand that one :confused:.

Tue 13 April 2010, 20:47
Hi All,

I might be missing something on the drawings regarding positioning of the laser cut gantry ends, and I am now a little confused looking at the photos above.
My question is are the two gantry ends without the additional slots fixed to the near side of the gantry??
and the two gantry ends with the additional slots, are they both fixed to the far side of the gantry (E-Chain side) as seen in the photos above??


Gerald D
Wed 14 April 2010, 00:46
You cannot use the photos above for identifying which end goes where. . . . .during the "untwisting" it doesn't matter at all.

Wed 14 April 2010, 03:06
Gerald and others,

I was working out what bracket went where on each end of the gantry from drawings 10 20 451 and 10 20 452 however when I was looking through this particular thread I saw in the photos that the brackets were installed differently to the drawings. My question is which is correct? I would presume the drawings are correct not the photos but do not want to presume and take the chance in welding the gantry incorrectly.
If my posts are in the wrong thread please feel free to move them to the appropriate thread.

Cheers & Thanks.

Wed 14 April 2010, 03:55

I think I found my answers here http://www.mechmate.com/forums/showthread.php?t=1295