Ordered, received, and assembled a TronXY X5S.

Ordered, received, and assembled a TronXY X5S. The printer comes as a kit, is quite cheap (got mine for $310 delivered), offers a largish ~300mm^3 build volume, and uses the CoreXY layout.

The kit as received has … issues. Knew this before ordering, as had checked before for what other folk had said online. This is my second printer. You would not(!) want this as a first printer. Bought this as an exercise.

Overall, this kit feels like some smart kid built a printer as a high school science project, and then someone decided to ship the same as a kit. Lots that is not well thought through … which is fine for a kid’s first project. :slight_smile:

First major issue spotted during assembly was the heated bed. The artwork suggests this is one of the http://reprap.org public designs. It is not. There are two different URLs printed, and both are bogus. Which is unfortunate, as if the maker had put this in front of the community, they might have got needed feedback.

The heated bed is a bit dangerous. The copper bus bars are too close to the edges, and much too close to the drilled mounting holes. The bed springs will eventually scrape through the thin insulating layer, and make electrical contact to the frame. This is not good.

Used a bit cut from a business card (heavy paper) as a temporary fix.

The thermistor is directly against the heating element. This is always going to read higher than the actual temperature of the print surface. Especially as the actual print surface is a ~2mm plastic(!) sheet. As plastic is a thermal insulator, likely more heat will escape through the bottom, than will in fact heat the bed.

Also this is a large print bed, made of quite thin material … and sags in the middle (and likely more when heated).

Have a sheet of cork to insulate the bottom. Need to find some appropriate-sized aluminum extrusion to brace the underside.

eabedfb457434a1c89a4b615b713b751.jpeg

https://github.com/johnsonm/Marlin/tree/tronxy-x5s-1.1.8 has mesh leveling enabled, and I definitely found that at least with glass on top it’s not flat; off by over .5mm. I don’t know how much was glass and how much was the bed though. Planning to upgrade from glass to mirror.
https://github.com/johnsonm/Marlin/tree/tronxy-x5s-1.1.8

u see the point is that the kit is so horrible that you will start to investigate y things wont work and in the end will learn a lot about electronics and mechanics so that you will become a expert in putting together and even designing your own printer !!! : ))…

for my first prusa i2 kit i got over 5 years ago the bed was a curved ( in z axis ) pcb and wont go beyond 40 degrees… from that experience i started building heated beds with silicone embedded heating wire ( for heating blankets ) and aluminium plates. I have yet to see a cheap pcb completely flat. Also using low voltage and high current for a heated bed doesn’t make sense to me.

My aluminum bed is not flat, but rather than sagging it bulges; it’s higher in the middle than at the edges. (Makes it hard to really take advantage of that large build area effectively.)

A correction: The build volume (according to the manufacturer) is 43560000 mm^3, or 43l, at 330x330x400mm.

@Lars_Clausen Just measured the plastic print surface at 330mm square. The underlying print bed is a bit under 329mm on each side, and the heated portion is inset by several millimeters. Taken together, we can likely expect the inner 310mm square portion of the bed to have fairly consistent bed temperatures.

(Measured as I was about to cut aluminum extrusion to brace and flatten the bed.)

So the manufacturer claim of 330mm is a bit(!) optimistic.

Conservatively, we can reasonably expect to use the inner 310mm. Might push to 325mm if the print is not sensitive to bed temperature.

Can the head also move to all parts of the bed without hitting things?

@Preston_Bannister I measured the edge of glass on top of the bed at about 5⁰ lower than the rest of the board at 60⁰ if I remember right. 310mm is probably about right.

@Lars_Clausen yes, it does, except for clips holding glass to the bed.

@Lars_Clausen Yeh, about that… :slight_smile:

The head can in fact move well past the edges, so with a different print bed, looks like we could add a few centimeters beyond 340mm.

However…

Remember those clips used to hold the bed? To clear the clips we have to inset the head about 15mm. Can fudge a bit by hanging the clips closer to the edge, and/or not putting the clips on two edges. Bit dubious, that. Inset by 15mm on each edge, and we are again down to ~310mm square, or perhaps a touch less.

BTW, the print bed in the picture has 1/8" of cork sheet glued to the bottom of the print bed, and some aluminum extrusion supports. Took out the visible sag of the bed. Looks flat(?) to eye. Also the cork is effective as thermal insulation. Took ~14 minutes to heat the bed from 18C to 50C, after which the cork bottom was still cool to touch.

@mcdanlj About your mention of using glass…

Glass sheet can serve as an inexpensive flat surface. If you have an unflat bed (as in this case), then slapping on a thick piece of glass is a quick solution. But … there are some large drawbacks.

Glass is a thermal insulator (about 200x less thermal conductivity compared to aluminum). If the bed under the glass is not level, then uneven contact with the aluminum will mean uneven heating. Taken together, the bed will take longer to heat, and heat unevenly. Also, I would guess that temperature on the top surface would be much less than registered by the thermistor taped (to the heater!).

We could slap a thick layer of thermal paste (as used with CPU coolers) between the glass and aluminum - but that makes removing the glass rather messy.

If we make the aluminum flat, then we get even and faster heating, plus we could use thinner glass. Then again, if the aluminum is flat, why are we using glass?

@Preston_Bannister https://plus.google.com/+MichaelKJohnson/posts/DcYfj6rDv7a is my recent post along those lines.
https://plus.google.com/+MichaelKJohnson/posts/DcYfj6rDv7a?fscid=z120ufh5uxuvd5h2f22uthua4wbjgh5m3.1521489073881802