IR board fix

I noticed that all of my IR boards had a longer distance from bed to nozzle than 3 mm. I found that my surface mount IR board would change from 10 mm to 30 mm depending on bed surface (reflectivity.). I posited that there was light leak from the transmitters and receiver that was transmitted/received down. I took black RTV and put it on the bottom of the transmitters/receivers and I put the exact right size diameter heat shrink tubing over the capacitor. My IR board went from 30 mm down to the 3 mm it should be.

 

Hi,

Would this fix the nozzle crashing into bed issue some of us have?

Sent from my SM-N970F using Tapatalk

 

I doubt this is the fix for your problem. This prevents early trigger resulting in a larger gap. It is more likely that you are not heating the bed before you do the auto level I have found that the bed changes a lot from 25C to 70C. This is my start up Gcode. It heats the bed before it does anything and keeps the print head cold until the last thing so it can not drip on the bed. It also make sure that the steppers are on, because bed heating can be long enough for the steppers to turn off and lose an accurate home position. This is my GODE Scripts.

I have also gone to 25 points in the auto leveling (5x5)


;heatup of bed
M190 S65
; set to millimeters
G21
; home
G28
; set absolute mode
G90
; turn off fans
M107
; move bed down 230mm
G1 Z230 F300
; Home
G28
; zero extruder
G92 E0
; Robo Leveling
G36
; home XY axis
G28 X0 Y0
; zero extruder
G92 E0
; move bed to printing position
G1 Z0.2
;non blocking heatup of bed
M140 S60
;heatup of extruder
M109 S220
; print 150mm priming line
G1 Y-150 E10.0 F500
; Retract
G10

 

Cool, thanks for that. My machine is currently stripped down, my nephew is putting it back together, am experimenting with vesconite bearings I bought for it a while ago... unfortunately life gets in the way of the fun stuff.

Sent from my SM-N970F using Tapatalk

 

I tried those bearings, they are too tight and the machince can not drive the head with the bearings in place.

 

I designed a cover that works very well and the surface mount IR boards now work great with the cover on.

 

It is so easy to fool oneself. I now "know" what the problems are. I got a hold of the source code for the last Robo release and modified it so that it did not do a check as to whether the board was analog or digital. It now jumps straight to digital. All of the Robo boards I had tested had the exact same frequency trim value programmed into the eeprom. This is a by micro value and should have be calibrated before programming.

So the two problems are: the board may come up in analog mode and operate really poorly. Another problem is the board could be running too fast and flaky or too slow to make accurate measurements.

Partsbuilt sent me some boards to program and test and what a difference when programmed correctly and when forced to digital. The software has the board come up fast and double blink to indicate digital mode. There are also some fuse settings that were not correctly programmed.

I noticed right off that now it was running at the optimum speed and settings that the led ( I drilled a hole through the shroud to observe the led) now had a solid pulse at the transition point instead of the bouncy led at the transition.

It seems to give me perfect results right now. I am using a 25 point bi-linear mesh measured by the ir board with a 10mm fade. The initial layer is perfectly flat and the initial line down the left is perfect. This despite the fact that I am using a garolite board that has been sanded out many times.

These boards should be available from Partsbuilt soon and are worth the upgrade.

I do not get any compensation from Partsbuilt and purchase anything I use. I help out so that Robo users can keep their machines operational and because the Robo community helped me so much when I started.

 

I have now verified that the new firmware changes to the ir boards work well. If you are not satisfied with your IR board, and this applies to old and new, PM me and I will send you my address. Send me your board with return postage and I will program it and return it to you.

 

Reflectivity matters. I have 3 machines where the priming line is perfect the entire bed length. I have one machine that it was not. I print on the front of the bed. It has a different material micro porous glass. I rotated the the plate 90 degrees clockwise and the priming line was prefect....for awhile. I realized that using hairspray I had to coat the whole bed, even what I was not printing on to get a perfect priming line. Once the hairspray was evenly applied to the whole plate the priming line was good.

On this R2 clone I use a 12x12 bed, a 6x6 probe matrix using bi-linear ( I have posted instructions on how to do this).

 

The firmware change on the IR board made it work much better. I would slow down the z axis speed on the bed on unmodified boards for best results. I would guesstimate that having reflectivity differences gave me a priming line thickness change of about .05 mm on one out of 4 machines. I would guesstimate that the modified IR board is repeatable to .01 mm. I am no longer observing room temperature drift changing the first layer. It is pretty impressive.