Differential IR height sensor, Attemptin' Clone....ain't sure of success...now stage 10

Thanks for the input, Mark...............my robo version implements M565 in the g start code, however I don't revise there.....right now, awareness may be all that is needed....appreciate

 

Probe Trigger Height varies from machine to machine, Why?

I have been thinking about this 'issue' this morning...........

The 'issue' may not be an issue, It is important that the sensors trigger height is consistent, within reason, on your machine.

There are only 3 components, on the sensor, that will have a significant influence on the trigger height, (LED illuminated).

The photo-transistor, (receives the IR light) and the two LED IR emitters, (sends the IR).
The lesser influence would be the actual variance of the the two part numbers intrinsic parameters.
The greatest influence would be the physical assembly of the 3 parts on the PCB.
It is a 'sorta' standard with PCB designers to create thru-hole leaded component holes in the range of 0.035" to 0.038".
components leads average ~0.020", this allows a comfortable annular ring void for solder to flow. Because of this slop,
consistent angular placement of those three components is next to impossible. Calibrating, matching the IRs,
and figuring the ideal mounting height will be the next item on the agenda.......why not?

 

Thanks for your expertise waldo..........

I received the the definition from that wiki post.....ie: M565 command
Set the offset from the extruder tip to the probe position. The X, Y and Z values are the delta between the extruder and the actual trigger position of the probe. If the probe trigger point is below the extruder (typical) the Z offset will be negative. This just records the point offset; it does not actually do the probing................posted reply #140

I did notice and was 'scratching' my head' ....Marlin did not appear in the choices for M565.....I noticed it was in my start code...but never revised it.

Hey Waldo, Thanks for 'clearing that up' for me, always a pleasure

 

Actually the IR is a more typical method of autoleveling. Where the sense point aka trigger point is when the nozzle is above the bed. If you imagine a swivel probe, it will put the switch below the nozzle otherwise the nozzle would hit the bed first and the switch wouldn't.

Thus an M565 value for this would positive. If you define this offset using Z_Probe_Offset, then it's saying "this is where the probe triggers relative to the nozzle" Which would be negative, since it's below the nozzle

The robo uses the nozzle as the probe and the trigger point is after the nozzle has touched the bed. From the printers perspective the nozzle is inside the bed. This would give you a M565 value that is negative. And the Z_probe_offset would be positive.


It's confusing I know but that's how it works.

 

Thanks Mike for the M565 explanation......With my limited experience in the 3D world............I never dealt with M565...or at least was aware of. I figured out early the nozzle is the Z probe & must "bury itself" in the bed until it lifts the X carriage thus deactivating the Z switches to achieve the Z trigger point. I knew recording the distance the nozzle snugs down on the paper is what you want....then in cough, choke, matter control, add the distant movement after that where the limits deactivate. Never needed to adjust offset anywhere else.

However, with z offsets tweaked here or there & with different firm-wares, slicers, and mods..........It's hard for all to be on the same page.

The robo forum is a great source of knowledge, ideas, and a willingness to help. Okay, soon time to watch cinci lose there first....Pittsburgh 6-4 now.........Your Peyton is in a slight slump.....Just as Rodgers......Thanks again.....

 

Love this thing. 7 for 7 on prints after I got the offset set. Thanks again Jim

 

Having a small issue with consistency. Trying to post a video although the forums is blocking it because of post count.

 

I'm running into repeatability issues on various platform materials. (Aluminium, Carbon Fiber, PEI, Borosilicate glass)

The printer is obviously a delta style in which the Z-height is set from the top mounted limit switches. Any variation in repeatability causes huge discrepancies in calibration. Ideally the z-probe would be installed closer to the nozzle, although this was just a quick test.

Am I correct to assume that the LED light should be solid during the first detection of the surface? Right now, I'm getting random Z-height values when running a calibration scripts and the LED blinks.

Would you suggest tweaking the sensor/emitters to adjust the angles?

 

Kas,

Thanks for the video....you are the 1st to install..using the sensor in the 'analog' mode. Be certain to do initial power-up with the sensor far from the bed. Move the z axis well above the bed...unplug the connector, (removing 5 volts), wait 10sec, then plug back in. Wait about 30sec. the sensor will flash 4 times. You may have already done that properly.......just need to mention. Sometimes power switch doesn't remove arduino power in certain machines, ie: robo.

Bed changes vary greatly in their reflection properties, hence different trigger heights. Robos bor. glass with its black base color works very well...reflections from below the glass.....not really there. I also see that you are aware of the differences...changing the bed surfaces.

In the analog mode...the sensor.....should first trigger...'I'm getting close'.....the z servo speed will decrease....then Z will continue.....to the final trigger point, LED solid on. That is accomplished by the Atmel monitoring the voltage from the photo-transistor. You can experiment with long nose pliers, carefully twisting the 'Far' led, (far right one, close to the vertical board edge). That will change the reflection angle. The sensors photo-transistor first looks for the far LEDs ir light...At this point, I am thinking the far led may require tweaking.......if you break it, tweaking, I'll send you another sensor.....in the mean time, I will pull my white wire, on the robo, tricking the sensor it's a rambo board, (high impedance)......and do tests.......I will post results later today, stay in touch........Take Care

Revision: You are the first to install in a Delta.....

 

Thank you for the quick reply. I'm using the IR Sensor with a Smoothieboard controller. The IR sensor flashes two times after about 10 seconds when I power it up.

When I get home tonight, I'll try to tweak it and see what happens. I'll also remove the plug and follow your instruction to see if I get the 4 light flashes.

 

Okay Kas, That gives me a good starting point......the flashing, it's in the analog mode but, the two flashes is digital mode....I'll do some research on the smoothie....gut feeling, I think its an analog input...........What I have seen...most deltas are analog inputs.............I'm sure someone out there will offer some insight......I'll be testing late afternoon.......too much to do in the present....later.....

 

Kas, post reply #154 is spot on.......I had two sensors assembled. I tested first one, no blinking......worked well triggered about 4 or 5mm from board edge.......shipped that one today to WI. Second one flickered LED, believe me about 4 inches from bottom of board. Triggered a few inches from board. I grabbed "Far Led" from rear with long-nose pliers & twisted clockwise...distance increased, not good. Twisted CCW, distance decreased until I was no longer flickering and triggering about 4-5mm as the first board. Having said that I was rushing, a ton of stuff to do this morning, so be careful on my memory of cw/ccw direction......I'll be testing more later today......Also, I planted by design the led emitters low, to keep them from being inadvertently misaligned. That low height makes the far led tricky to adjust.....again, if you break it......probably not your fault. I believe I spoke earlier in this post about the board really needs to be placed in a fixture aligned, then soldered...

I'll be back at it in an hour or so..........I'll address your smoothie issue......Later....................

 

Kas,
Got some time to work on the Smoothie Board issue...Here we go!

No physical tests yet but theory for now.......Looking at the Smoothie Schematic, it uses a LPC1769 micro...the 4 A/D analog inputs are Port 0 (23,24,25,26) and are used by four thermistors. The Zmin. input is Port 1(28) which is digital. That suggests the IR sensor should power up in Digital Mode...that would be true for all Smoothie operators. Two LED blinks after power up.

Now, the Smoothie implements a 1K pull-up Resistor shown as R100, (the component designator on the PCB). The IR sensor cannot pull down this low value part thus, the smoothie input to port 1 (28) is stuck at an high. Note, this has nothing to do with your blinking problem, that's a separate issue.

The problem may be addressed with any one of these three suggestions:
Replace R100 (1K) on the Smoothie with a 10K resistor.........That will allow the ~10K "pull-down resistor" on the sensor to create a digital low to the Smoothies Micro input. If 10K doesn't work increase upward.

OR

Replace R4 (10K) resistor on the IR board with a short.

OR

Connect a 470 ohm resistor from the white wire, (output) to ground, may work....worth trying cause it's easy.

First, get the blinkin' stopped........tweak that far LED........IR should trigger 4-5mm from bed (I'm still working on the best# with the Emitters I chose).......project still fluid......

Second, Try shorting out R4 on the IR....30awg solid, wire-wrap wire to two points.

I'll be outta town till tomorrow nite...........I'll wait to see how you make out.............Take Care & Later

 

Kas,

leaving shortly.....Forgot to mention...You may want to experiment & paint that aluminum bed.......a flat black...will for sure bring down that trigger point.....if the glass is transparent/lucent.....you may try to sandwich a cutout of black construction paper....Have great weekend............jimmy

 

Jimmy the miracle maker, just wanted to say, we are working great here, thinking it might be a nice sensor for the laser cutter. Rock on and have a great weekend.