What's the Tiny Touch Plate Neodymium Magnet for?

For now, given both the affordability of the Tiny Touch Plate and its built in support in the V1 firmware, I’ll probably just order it and see how easy it is to have Z but keep doing what I’m doing for X and Y. If I later get into probing all three, I won’t be out much in the way of either money or time.

Funny! I was looking at this the other day. I was especially interested by the Estlcam video, for the alignment function shown at 1:06. Knowing myself that would be useful :slight_smile:

I am still struggling to figure out things, but I suspect regular probing all 3 will not be that difficult.
According to how I see things now you would do something like:

G38.2 Z0
→ probe Z axis and define it’s location as 0
G92 Z14.0000
→ set your probe thickness, in my case 14mm, so Z0 is actually -14mm
G00 Z19.000 F500
→ lift your probe 5mm above the sensor
G00 X-20.000
→ move x 20mm past the probe
G00 Z-0.5000
→ drop Z so it will touch the side when probing
G38.2 X0
→ probe X axis and set location to 0
G00 Z19.000 F500
→ lift your probe back 5mm above the sensor
G00 X20.000
→ bring probe back to starting point
G00 Y-20.000
→ move y 20mm past the probe
G00 Z-0.5000
→ drop Z so it will touch the side when probing
G38.2 Y0
→ probe Y axis and set location to 0

I am sure I will have made a noob mistake in my code, and values depend on probe location, but this my logic on how the Gcode works to probe 3 sides based on the G38.2 command.

When I find the time, I’ll try to simulate this using my regular probe, just to see if it responds to G38.2 Y or X. If that’s the case I suspect it might work, then I want to figure out a way to find that angle deviation.

One other thing I am looking for, is if I can permanently connect one wire of the probe (I use an alu base). But my test so far showed that the spindle shaft seems to be isolated on my Katsu:( although that makes sense somehow, as it’s some kind o a coil that makes the motor move :slight_smile:

2 Likes

Lol, already spotted a rookie mistake, if you would go back to the original starting point, to probe third axis, you’ll need to take into account thickness of the bit.
Although if you would move far enough, like 2cm in my example, this would be still ok.

1 Like

@Olivier

That’s some cool stuff in that Estlcam video! specially the auto angle detection.

Indeed!

@DougJoseph I can confirm that 3 axis probing works like a charm on SKR Pro using Z axis probe sensor.

As I do not have such touch plate I had to simulate it with my regular one.
The commands I provided earlier work just perfect using terminal, so they can be added at the beginning of the Gcode script.

Now, if only someone had ideas on calculating that angle deviation, I would really need to buy a new touchplate :smiley:

That’s great to hear.

Thank you. You are the man! One question though. You had mentioned:

The example you gave gets around that issue completely, right?

Yes!

Let me try to explain in my best English;
It’s a matter of placing the bit rather close to the edge of the 3 axes touch plate.
The script needs to move far enough on the y/ x axis.
Say, you place the bit 2 cm of the edge, and in the script you move it 1 cm to probe the side, this will obviously not work.

If you probe, and set this axis to 0, and want to go back above the plate you need to move it far enough in the script to reach that Point (including the bit thickness). I believe it’s better to move it further to probe and go back to it’s origin, then saving time and use short distances.

1 Like

Btw, I figured out that the angular deviation measurement would be rather easy to implement using G68 command, however I believe it is not available in stock Marlin and I’m not (yet) clever enough to know how to implement this. :joy:

Source
Quote:

– Align work that is not exactly aligned to the coordinate system. For example, suppose you wanted to run without tramming in a fixture. If you can probe the fixture to determine its angle, you can apply a G68 to “zero out” that angle and then run the part program. This can reduce setup time by reducing the need to be accurate and trammed in.

@Olivier

Ah, got it! Thank you!

1 Like

@Olivier

I had ordered a 3-axes touch plate, and it arrived today.

I’m looking at the sample code you supplied.

One question I have is: shouldn’t the x and y axes sections have their own versions of the G92 command, that takes into account the thickness of the lip on the touch plate (amount by which it overhangs the material to grip it), and also takes into account the diameter of the bit?

For example, on Z axis, you have:

G92 Z14.0000
→ set your probe thickness, in my case 14mm, so Z0 is actually -14mm

And on this website’s instructions for the Tiny Touch Plate, it shows:

G92 Z0.5

So, wouldn’t the X and Y need a similar approach?

1 Like

@Olivier

One other question, this time regarding G38.2. Your comment says,

However, in this YouTube video I’m watching, link below, the man says the number after the Z is setting a limit on how far down the probe should be willing to go while looking for the target.

Is he wrong? Or is your comment of “define it’s location as 0” not being understood in the right sense?

Thanks for any help or clarification.

Here’s the video link: https://www.youtube.com/watch?v=ZELGtCXdSso

It’s around the 4:55 mark.

PS: You have great English! One interesting tidbit: a really odd quirk of English is that “apostrophe s” ('s) shows “possessive” except in the case of “it.” In the case of “it” the apostrophe means you want a contraction of “it is” — and to show possession, there should be no apostrophe. So “its” means possession, while “it’s” means “it is.” So the comment would be “define its location as 0.”

1 Like

you are right!
In my example I forgot to set the offset (?) of the wall thicknes, on Y & X axes :slight_smile: If understanding Gcode correctly, the probing sets the 0 point, but you will always need to take into account the thickness of the probe. In all 3 directions.

Sorry for that mistake, I am still learning :slight_smile:

But, I believe there are multiple ways to achieve the same thing. I can be totaly wrong, but I would imagine that the following commands might result in the same thing, I should check it one day.

G38.2 Z14

or

G38.2 Z0
G92 Z14.0000

Oh, Here you can find the Linux Gcode details on 38.2, and here the Marlin documentation of it.

I am going to take time to watch that clip, it´s interesting.
In my tests, if I raise Z to the max, and issue the command G38.2 Z0 it keeps going the total distance I want it, until I trigger the probe. So it could be that “0” is just an “unlimited” command. But I am going to test to be sure.

I can say for sure now, that if I first set G92 X0 Y0 Z0 , my probing wont go further then let´s say 2-3cm.

Thanks Douglas, that made me just a bit smarter today :slight_smile:

Ps, in case any one looks for info on the G68 command, I found that Marlin does not support it. But you can actually add new G commands to the firmware. It would be a matter of implementing the G68 code, and add a way to set variables in the script, based on few probes. Then you could rotate the whole Gcode script based on the angle of your workpiece.

@ Olivier

Thank you! You are amazing.

I think if I get a working script that includes the x and y axes sections have their own versions of the G92 command, I would be golden.

So, did this morning a test, and it seems I am wrong :slight_smile:
Using the first command, you´re probe isn´t moving downwards as it needs to be:

G38.2 Z-14

But this would give another issue, it would only drop 14mm from current Z zero point (just like described in that YT video).

Next, I noticed that depending on the fact if I zero-ed the Z axis in the menu, following command would not always work. Didn´t find the reason for that just yet;

G38.2 Z0

So I believe, that for probing Z, it would be best to measure you maximum Z height, say 400mm, then use this command:

G38.2 Z-500.000
G92 Z14.000

That way your Z keeps going down, without giving a timeout, no matter how far it is raised. That would be important when using a homing Z call, as I believe Z will be at his max (in the example 400).

To complete the script to use 3 axes probe, something like this would work I think (without homing, or setting zero xyz!);

G38.2 Z-400.000 ; Probe Z from max
G92 Z14.000 ; Set Probe offset to 14mm, thickness of probe
G00 Z19.000 F500 ; Move 5mm above probe
G00 X-30.000 ; Move 30mm on X
G00 Z-9.000 ; Drop Z 5mm so X probe can make contact
G38.2 X0 ; Probe X axis, if zero´d X then cahnge this to for example - 30.000
G92 X10.000 ; Set Probe offset to 10mm, thickness of probe
G00 Z19.000 F500 ; Move 5mm above probe
G00 X20.000 ; Go back to original starting point
G00 Y-30.000 ; Move 30mm on Y
G00 Z-9.000 ; Drop Z 5mm so X probe can make contact
G38.2 Y0 ; Probe Y axis, if zero´d Y then change this to for example - 30.000
G92 Y10.000 ; Set Probe offset to 10mm, thickness of probe
G00 Z19.000 F500 ; Move 5mm above probe
G00 Y20.000 ; Go back to original starting point

Going to see if I can get a cheap 3 axes probe somewhere in Europe. Scienci Labs uses fair prices, but shipping and import tax make this extremely costly.

1 Like

ive been searching high and low for what the hell a touch plate does lmao… thanks for this explanation.
you at least got me on the right path. lol :slight_smile: thanks again

Hi Doug. You ever manage to get the 3 sided touch plate workign properly?

Well, couple thoughts. I am trying to remember… I definitely got Z height checking working with it, and I cannot remember if I got 3 point touch working or not. But most of the time since I have not used it. However I’m minded to get back to use of it at least for some tasks.

1 Like

I would too.

Please share your gcode when you do though.

2 Likes

I’ll try to remember to do that.