Dial Gauge Holder

Many 3D printing forums and blogs lecture on the quintessential importance of a perfectly level print bed. Achieving this state of printing nirvana is however a tedious and difficult task. Checking the height of the bed at various points requires you to move the print head to that position and then carefully lower it until it is almost touching the bed. You can then note the position by sending an M114 command to the printer.

Deciding whether it is “almost touching” involves a bit of judgement and the print head only moves in 0.1mm steps anyway, so the measurement is never really that accurate. Also this method means that for every point you measure you have to raise the print head, move it to the right position, lower it a step at a time, get the position from the console, move print head up again, expire of boredom etc.

My dad observed this process during a visit and used his powers of old-school engineering to suggest using a dial gauge instead. With a dial gauge attached to the x-carriage you could whizz the print head around the bed and just read off the bed height at each position. He then managed to pick up a dial gauge with an accuracy of 0.01mm, now we’re talking! All that was needed was a way of mounting the dial gauge to the print head.

A quick look on thingiverse showed that many other people have had the same idea. However they were all either for different printers or different dial gauges, so we set out to make our own. My dad drew out a scale design (on paper, like with a pencil and everything), then I put it into OpenSCAD and we ran off a few prototypes. Originally the plan was to glue the holder to the x carriage on the Huxley and use a screw to secure the dial gauge, but once we had the first prototype we thought we could play around with making everything snap fit.

It seems to work very well. Based on the measurements I made just now one end of my bed is around 0.3mm lower than the other. I knew it was bad, but I didn’t realise it was that bad – I think I have got away with it so far because most of my prints have been quite small and positioned in the centre of the bed. Now that I have a way of measuring the bed position I am going to try adding some thumb screws to make adjusting the bed easier, then finally I may be able to attain the impossible dream of a level bed (I will also have to add new tape to my print bed, it has got quite torn up recently through over zealous hairspray use and semi-drunken part removals).

If you want to make your own the .stl and OpenSCAD can both be downloaded from thingiverse

Things that went wrong this time

It seemed like everything that could go wrong did go wrong while making this part

  • First of all the belt slipped off the y cog during a print. It has clearly been slowly working its way further down the motor shaft for the last year and chose this moment to wreck the print
  • Then the threaded rods that give movement on the z axis fell out of their flexible couplings (again during a print). There was much cursing
  • The most terrifying failure came when I started up after replacing the z axis. I started to heat up the print head and the Wife pointed out that smoke was coming off the print head. I looked at the pronterface screen and saw that the printer thought the temperature of the head was -237 degrees C, so it was merrily pumping every amp it could into the superpowered heater cartridge that is installed in the printhead. I hastily shut down the printer but not before the leads for the cartridge started to char and the nozzle got clogged up with melted teflon. Clearly the thermistor contact had broken momentarily and the printer had interpreted open circuit as a very low temperature. I often leave the printer going overnight or when I am out of the house so this really scared me – if I hadn’t been there to switch it off it could have caused a fire. I had read that this issue existed in old versions of the firmware but never got round to checking whether my firmware had the problem or not. Public Service Announcement: upgrade your firmware to avoid burning down your house. Aside from saving lives and property it will also stop you from having to unblock your nozzle which is a pain even when using the fire method
  • Then the x axis belt came out of the x carriage, making the x axis the 3rd axis to fail in 48 hours
  • Then the other side of the belt came out of the x carriage, making me kick myself for not securing that one at the same time as the first one
  • Then the first side of the belt came out again, leading me to reevaluate my method for holding and tightening the x belt
  • …and then it worked

12 thoughts on “Dial Gauge Holder

  1. Good work on the epic failures. It’s as if Squirty hates you, had you been ignoring him? Perhaps he’s jealous of GIB’s new home?

    How did you design the “snap fit” bits? I’ve only just started experimenting with thin-walled sections to allow some flexibility in PLA parts, but I can see that there’s going to be a fine balance between enough flexibility to allow movement, but still retain enough rigidity to give a decent “hold”. Any clever techniques for dialling in this part of the design?

    1. Squirty has nothing to be jealous of. His home is complete, soon I shall post a picture and GIB will demand an upgrade.
      For snap fit stuff the section I used on the inside corners for the dial gauge holder seems to work quite well for adding flex (you can see it in the picture where I’m holding the gauge). Cutting a bit out of the inside corner has the advantage that you no longer have to have a perfectly sharp corner for it to fit over the x carriage.

  2. You will need to keep an eye on the Z axis couplings.
    They will continue to work off. It’s a weakness in the implementation. However bless their cotton socks an improvement is available…

    Not sure what gear type you have on the Y axis but they distort if its the plastic cast variant. This makes them out of round slightly and then they work their way off the motor shaft. Keep a close eye on this and after a while it seems to stop. I intend to replace them with Aluminium ones at some point.

    Don’t recall any issues with the X belt.

    See you have shielding on the limit switches, I simply re-routed these. If shielding, you might consider shielding the motor cables to keep the EMC in so protecting everything on the machine.

    I found some high temperature aluminium coated adhesive tape for the bottom of the X axis print head support. To reflect bed heat away from the bearings and plastic bracket.

    In the first hundred hours of bedding in, its good to keep an eye on the machine as this is the first part of the bath tub failure curve.

    A smoke alarm would be s small investment and I never leave mine switched on when I am out. See http://northernhope.blogspot.co.uk/2012/08/maintenace-weekend.html

    You might want to look at getting some borosilicate glass for the table.

    I now have over 600m of printing done. So its a good little beastie when you have the wrinkles worked out.

    1. Thanks for the suggestions. Based on the emptiness of the various spools I have lying around I think Squirty is now on about 1kg of PLA which should be about 300m. Glad to hear yours is still going long after that point. What is likely to fail next? I need to print myself a full set of spares so I’m not caught with a broken part that I can no longer replace…

      I already have the upgraded z couplings. I might just have not tightened them enough. I have the printed y axis gears rather than the cast version – will keep an eye on them. The x axis problem was my mistake.

      I do have a smoke alarm. After reading your (quite alarming!) post I will stop leaving Squirty on when out at work, at least until I can put some failsafe on it to stop the power if there is smoke.

      The borosilicate looks very nice. I noticed you used the heated bed with it. Does it work without heating? I find the heated bed bit of pain and would like to avoid it if possible. I will have to pick up some of this no additive nail varnish remover from Tesco next time I am in the UK.

      Why is your printer called Huxley #710? Also have you found any thumb screws that work nicely for bed leveling? All the ones I’ve seen so far don’t fit because the screw heads are right up against the heated bed.

      1. Hi,
        Borosilicate glass needs the heater on to get the surface to to at least 60 C. What you will love is that when it cools to less than 55 C parts fall off! Unlike Kapton tape. With glass like surface finishes on my parts, It’s simply the best improvement I have made. This will save you more time than the bed ever takes to heat up.

        I am sure any ‘no additive’ nail varnish remover will do. Tesco brand lists in its ingredients Acetone, Aqua, Propylene Glycol, Glycerin, Denatonium Benzoate, Cl 17200 and Cl 42090.

        My wife once changed the tissue paper I use to wipe the acetone with to clean the borosilicate glass. The new stuff had Aloe Vera in the sheets. We nearly had a divorce when I fount out the root cause of my extrusion not sticking!

        Huxley #710 is named after the number on the Micro controller label. I didn’t have the imagination for a super name like ‘Squirty’. I assume they have sold at least 709 printers before mine.

        Once I went with glass, the lack of handling to put tape on means I rarely have to touch the bed leveling now. So have not found any thumbscrews.

        I think the hot end Fan is likely to fail next. I notice there only seems to be one LED on mine now! Could just be the LED’s but I suspect that its 12v fan running at 19v so with the higher rotational speed the bearings may well suffer as well. I may well replace it proactively as part of a preventative maintenance programme.

        1. Thanks, it sounds great! I’m going to see if I can get hold of some borosilicate here to try it out. I had a similar experience to yours with the aloe vera additive. I was going nuts trying to work out why my prints were coming loose even when the bed was sparkling clean. In the end I noticed that the nail varnish remover I was using had palm oil in to help nourish your nails…
          I’ll have to check my Melzi board to see what # Huxley I have!

          1. I found some borosilicate in Singapore ($40 SGD from ). I cut it to size (surprisingly straightforward) and have tested it out. First results were very disappointing – the filament didn’t stick at all even at 60 or 70 C bed temp. Instead it just stuck to the print head. However once I applied a little bit of hairspray it worked really well – sticks well during print then snaps off easily once cool. It’s also much easier to level and clean. I do have to get used to it though. Every time I look at the printer I keep thinking the print has started to peel off the print bed, then I realise it is just the effect of having the transparent substrate.

            I’ve read that glass’ surface properties can vary a lot, so maybe you were lucky to get a piece that works without any adhesive (or I was unlucky!). It could also be that I am using paint thinner of unknown composition rather than acetone or no additive nail varnish remover to clean it (still haven’t found somewhere to get it)

    2. Yikes – this is why you shouldn’t leave a printer running unattended. Looks like the plastic x carriage burns very nicely
      Burnt printer

  3. If your setting the bed temperature to 60 ish the fact that glass is a poor conductor will mean the surface will not be at 60C.

    Suggest you check the surface temperature or raise it another 20C or so. As I found its the surface of the glass temperature that is important.

    1. I will try this, thanks! As a safety measure I want to add some additional thermistors to the printer controlled by an arduino that can shut off the power if anything gets too hot. First thing I will do is measure the temperature of the top surface of the print bed and see how much cooler it is compared to the reference temp.

    2. The borosilicate started working really well- sticks perfectly at 70C without hairspray, then pops off when cold just like you said. I think the issue was the dodgy paint stripper I was using to clean the plate. I need to find some cheap nail polish remover instead.

  4. Hi,
    after all the work on Kapton tape, it was like magic. Glad it worked for you.

    The cleaning agent and technique is really very important, once my wife changed our toilet paper. I use it to apply TESCO nail varnish (acetone) to clean the glass. After a number of prints there is some kind of build up that prevents adhesion that needs cleaning. Unknown to me she had purchased some paper with Aloe Vera.

    It nearly led to a divorce! One thing I can tell you; PLA does not stick to Aloe Vera residue.

Leave a Reply

Your email address will not be published. Required fields are marked *