RepRap: Blog

The mystery image three posts back was a lame attempt at the Mona Lisa. The whirring noise you hear in the background is Leonardo di ser Piero da Vinci revolving in his grave at high RPM. I tried printing a 43x64 icon with 4μm pixel spacing on the grounds that I could definitely keep the probe level. It came out a smidgen better but still awful:

43x64 dithered icon of Mona Lisa, with adjusted contrast

Barely recognisable even if you know what it is. The whole thing is 1/4mm tall and I stood no chance of tracking the print's progress at that resolution, so I was surprised anything came out at all. Even the original icon wasn't that flash:

Anyhow, got that out of my head. Back to more serious work. I've fitted a larger probe point and photographed the Touch Probe assembly process for documentation.

After Jon Oxer's Livestream https://www.youtube.com/watch?v=DZTzVepHpmA last weekend a few things cropped up. The major one is that the hole spacing on the Stage was wrong, so if you have printed one prior to last weekend you'll need to redo the stage on pika_sheet_2.stl . I have updated github and printables.

I thought I'd better print the latest one and test assembly, as I'm running a hybrid of V0.01 and V0.02 parts on my bench. In doing so I noticed that the left-hand Stage mounting hole on the XY Table is a bit shallow, only allowing 2mm of vertical adjustment screw to enter. I have fixed this in the model, but if you have already printed the XY Table you can slowly run a 3mm drill into the hole about 10mm and that'll give you all the adjustment clearance you'll need. There's nothing terribly important under the hole, so don't worry about breaking it.

While I was there, I took out the captive lugs on the inside of all the nut holes. As the mechanism is saner to assemble now, there's no need to retain the nuts and nut insertion is much, much easier. 

It still looks the same and assembles in much the same way, so no need for a documentation update:

 

Flushed with success, I thought I'd try the new probe out over a large area - 2mm x 1.3mm - and this did not look so flash. But we learn from our mistakes so here is one. 15μm spacing but a stippled image (figure it out, heh):

Click for the full-size image

Careful examination shows that the probe contacts as a point nicely across the print area along X=0, and going up to I dunno 0.7mm or so? Then the probe starts smacking into the slide too much for the flexure to cope with and it starts skidding off sideways.

Something is off somewhere, because beyond that area there are some dots, and some skids. I do not know why.

The tip fitted to the current Touch/Probe head is far too fine for me to practically work with on the 10's of microns scale that my hands are capable of. But before I retired it, I just had to find out what it would do to "Hello World". We really are bumping into the limits of what one can do with a pointy thing, Sharpie marker, and a Konus binocular microscope here, and I had to drag out the Awful (but cleaned) Biolux to get better shots. So here's what I can see with an 8μm spacing using the Konus:

Few things to note. First off, that's 56μm tall. This about the size of the spacing between dots in previous Hello World's (50μm). Second off, the amount the probe skids sideways on contact is far, far smaller than previous attempts. I attribute this to the more precise probe height positioning possible with the Touch/Probe, and the fact that the Probe Tip moves vertically when pushed rather than diving off at an angle.

Time then to zoom in on a gory detail shot with the Biolux:

 

Woah. I think this illustrates pretty well that PIKA has sufficient control to work on single digit micron scales.

Which begs the question, what if I tried 4μm spacing? OK then.

At this point the debris kicked up by poking the point into the Sharpie is becoming a limiting factor. Here's a shot through the Biolux:

Yep, definitely pushing it there. Limits seem to be in the substrate/tip, not the μRepRap.

So, takeaways from all this. PIKA with the Touch/Probe is going to be up to positioning the probe as accurately as is needed for sub-10μm features. That means I need to get the Touch/Probe design put in the repository, and put a wider tip on it to begin resin deposition on the 10-20μm scale once more.

FYI This is the current probe configuration:

 

I took the pointy contact probe from the previous blog entry and smacked it into the glass slide a few times to Z-10μm and it seemed consistent, though the Z axis seemed to pick up unwanted vibration. Will have to look into that. Anyway, it didn't seem to be flattening so I put some resin on the slide and picked some up with the probe point.

Now this point is a scrap one I over-etched, and it looked a bit too pointy. I figured a finer point would be more likely to exhibit damage, and I didn't have a use for it, so it got sacrificed. I got confused because the contact indicator was showing contact with the slide, and as far as I could tell the Z axis wasn't going down any further, but I could not see any resin deposited through the USB microscope.

So I took the slide to the Konus trinocular and looked at it under 90x. Couldn't see any dots. But I had a feeling, and so I brought out the not-so-brilliant Biolux microscope and had a good look at 160x (10x objective, 16x eyepiece) with some lighting and focus plane tweaks that I can't do on the Konus:

This was taken through the eyepiece with my phone (5x optical zoom), and the Biolux optics are gritty with age*, but I think you can see a horizontal trail of dots across the middle. So I was depositing resin, and deposited in excess of 20 dots on one dip of the probe.

The dots are spaced 20μm apart, and measuring the image pixel by pixel in the enlarged section suggests a diameter of approximately 3 microns. That's smaller than half a human red blood cell.

That's pretty impractical for me to work with at this point, and that line is not as horizontal as I would like it to be. Not that the wavyness matters so much if I'm drawing lines 20μm wide. But it's nice to know that in theory I can make features that small.

Anyone want to chip in for a 2x Barlow lens for enhancing the Konus? 

 

* That image was so horrible it actually drove me to take the Biolux eyepiece apart and clean each lens individually with IPA. Not a huge improvement, but a bit better. Do I dare fiddle with the objective lens and risk ducking it all up?

Having made a Touch Probe that works give or take a micron, I wanted to start printing in resin again. But I became troubled by the issue of zeroing the probe's point on the slide to the stage where it was causing prevarication. Thus prevaricating, I decided to stick a Probe Tip on the Touch Probe:

I turned the probe beam section into an L-shaped trough and just glued the probe wire in with copious cyanoacrylate adhesive (wire was held in place with Blue Tape while that cured). The side wall of the L prevents adhesive from seeping into the flexures. There is no electrical probe connection - I'll be levelling it manually, so no need for a Z Touch capability right now.

The main concern is that I'll knacker the probe tip by banging it on the glass slide before I figure the height out. These tips seem fairly robust though, so I might get away with it. I'll test on aluminium first, which is a bit softer.

A minor concern is that the probe beam might vibrate too much. I didn't see that happening during touch tests. Limitation: The whole purpose of is is because I can't see micron movements... 

The main operational disadvantage is that I can no longer view the Stage from directly above. The Touch Probe flexures get in the way. The new microscope pole clamps are working pretty well, so I hope I'll still be able to get an informative angle on what's happening at the pointy end. But if I place a microscope to view from the "front" then I can't twist the pole to move both microscopes out of the way as easily. We'll see if that's an actual problem.

If the new touch/tip probe does anything useful I'll upload the CAD files. 

I've made a small test flag and printed it with a lug to mount it on the Z Axis with. The "flag" was flipped around so that it was near the contact point - much handier when you're trying to watch everything through a microscope.

I can certainly see movement that happens over a micron or less. This is done by bouncing the probe 10μm. If it twitches as it gets to the bottom of the movement, then you're contacting the slide. If it doesn't, drop it by a micron and bounce again. It makes for a boring video.

There's a bit of tidying up to do, making things align in the microscope's field of view, making sure there are more straight edges when you're looking for motion etc. But I might be on to something.

Options might include making it into a proper indicator gauge, and being clever with capacitors or inductors to give electrical feedback. For now, just watching the indicator arm twitch gets me +/- 1 micron.