Z Carriage Accuracy

[Steve Tribe]

Introduce myself.

I am Steve, I am a Automotive Design Engineer by trade and have been for many Motor vehicle companies around the world. I know a lot about certain high end cad packages, so I am able to design good 3d models for printing.
Oh I might add I have had no problems with my Mod-t so far except this one.

This is where my story starts with this problem. Having designed my own part and uploaded it to the newmatter store, it was printing with no problems. Except I did not account for how much filament it would need. 6 layers from the top it ran out of filament. I changed filament no problems, then started the print again from where it left off.

As it printed I noticed the layers did not line up anymore. I checked the machine after it had finished and noticed the Z carriage had a lot of side to side movement in it. I partly dismantled it to see what was happening in the carriage and these was my findings.

As you can see from the photo the underneath has 2 aluminium guides to the steel rods.

After removing the top plate there is only 1 aluminium guide at the top on the right.

This is giving a side to side movement of atleast 1 mm at the steel rod on that side.

I was wondering if anyone else could confirm if this by looking at there machine, if this is a design flaw or quality control issue?

Thanks for your help and regards

Steve

[Steve Tribe]

After looking at the pictures in this thread.

https://groups.google.com/d/topic/mod-t/o1mLY268Sik/discussion

I have come to the conclusion, this is a absolute rubbish design for guide rails. I will be at some time redesigning the Z Carriage for myself. To be honest there is too much slop in the system.
Especially as part of there advertisement is a 3d printer you can have in every home. Thankfully my child is grown up and won't be running around, but for people who have children or pets this is a nightmare design. One little knock and you will have to print it again.

[Robert Villegas]

Steve,

Your printer doesn't have any assembly flaws as I can see from your pictures.

The vertical guide rod bearings are not made of aluminum but I get what you're saying. You see only three, not 4, bearings. Two on the bottom, and one on the top of the carriage. The set-up is not symmetric. Given that the vertical guide rods are fixed, to properly constrain (and not kinematically over-constrain) the carriage, only three bearings should be used. These three bearings in conjunction with the fixed vertical rods constrain the carriage in 5 of the 6 kinematic degrees of freedom.

Unfortunately what you have experienced though is likely lash from unavoidable manufacturing clearance of components (unless one of the vertical rods is loose). Keep in mind that to achieve the costs required to sell a 3D printer for $399 the use of more precise and expensive linear bearing or a stone rigid rod mount was not feasible. If you believe your machine is suffering from excessive lash or "bearing play" please contact the new Matter support team.

Thank you very much for supporting us and buying a MOD-t. Let me know if you have any other questions. It's really awesome to read about users' experiences with our printer.

Best regards,

-Rob

[Steve Tribe]

Hi Rob,

I have contacted Newmatter about this and have got no reply. I can understand what you are saying, but there is way too much play in those bearings. I can move that bearing from said to side, it has +0.5mm of clearance to the shaft. Veical rods are not loose. Given the accuracy of the XY carriage and then the lateral play in the Z carriage, that is too much. That is more than 6 degrees of threedom.

I will try and get a pic which shows the oversized bearing on the rod.


Steve

[Steve Tribe]

Just checked the lateral movemement in the Z Carriage it is between 1.5mm and 2mm at the filament feed point.

That is too much lateral movement on the Z axis to give anything near any type of accuracy that this machine should be doing. Especially as it only prints small parts.

Linear guides are not expensive. I know what things cost to manufacture in China, I am an Automotive Design Engineer. This machine does not cost anything near 400 dollars to manufacture, probably 100 - 150 per unit max. So putting better guides in the Z rails is not a problem of cost.

[ajf]

Hi Rob,

First, let me say that I really appreciate you posting to this little forum I started. And, having been involved in the development of more than a few consumer electronics products, I too take pleasure in getting direct feedback from users, even if it's not always positive. 

Like Steve, I have extensive experience in sourcing and manufacturing in Asia and have to agree that the production cost argument falls a bit flat. But, I think the seminal cause of many of the issues people are experiencing is the choice to use servos rather than steppers.  Yes, on first blush, a savings of ~75% on the cost of the motors seems attractive, however, once everything else is factored in the savings vanish. The choice of servos informed all the other design/engineering choices negatively and, frankly, resulted in a less than reliable/robust machine.  I'd be happy to be convinced otherwise, but believe that that single decision to use servos has led to increased cost in terms of software development, support, and goodwill.

regards,

aj

P.S.: I think some, if not most, of the issues people are having could be mitigated by opensourcing the firmware and letting the community of makers develop software fixes. I have personally tried to reach out to Steve Schell to make the case for opening the firmware, but have not received a response as yet...

[Robert Villegas]

It does sound like the bearings you have on your machine have too much clearance. There really should be little to no radial play between the bearing and rod.

Support might be a little swamped right now or backed up due to the holidays. I'll follow up with them to address your issue. For what it's worth when the unit has correctly sized bearings the position of the carriage should be very repeatable. I have a 20 month old and he's always bumping my two printers at home and besides pausing jobs by pressing the front panel button he hasn't messed up any of them.

If you are printing really small parts, be aware that the machine's reliable minimum feature size for sharp edges is about 2mm. For example, I have found that printing small cubes smaller than 2mm results in rather rounded off corners.

[Steve Tribe]

Rob,

When I say small parts I do not mean that small luckily. The part I was printing when this happened the last 6 to 8 layers was of the main part anything from 0.5mm to 1 mm and this stopped correct cohesion to the lower layers. Which in turn did not take much for those layers come of the part. The part itself was about 40mm x 40 mm x 40mm.

Thanks Rob for your help, I hope something can be done about this. Lateral movement in the Z carriage is very important within any cnc type machine. Otherwise my printer is a very good machine, no problems with Wifi (except the setup software does not show it is connected to Wifi), No firmware problems, No crashing into bed and had filament breakages but no problem in changing or clearing them.

What would of been a nice touch, would of been a light under the Z carriage. To be able to see what is happening (Also because here in the UK we have dismal light at this time of year)

[Robert Villegas]

Steve,

I'm with you on the lateral stiffness, repeatability and accuracy. Whether it's a milling machine or a 3D Printer. It's super important that the X,Y and Z positions are true and given a reasonable external force to the machine, the hot end nozzle should return to the original position. The MOD-t as designed will do that.

What you are reporting sounds like a pathological issue.

I'm out of the office right now but when I get back next week I'll pick this back up and we'll investigate and make it right.

If I bought a 3D printer that could not place its print nozzle in the right spot I'd be upset too.

Regards

[Robert Villegas]

I like your idea of a light under the carriage!