Posts made by kolja

@escknx have a look here: https://wiki.e3d-online.com/E3D-v6_Troubleshooting

"An easy way to check that your heatsink has sufficient air flow is to simply feel the heatsink with a finger after a print has been running for some time. (Be careful not to touch the heater-block, it will burn you instantly) The entirety of the heatsink should be cool to the touch, including the bottom fins closest to the hot parts."

But take care! This text's for the e3d v6 heatbreak. Might differ quite a bit from mosquito.

My v6 heatbreak is easily touchable. Only the lowest fin - directly above the heatblock - becomes warm. With my duct it's practically impossible to cool the lowest fin from below - but it doesn't affect the print. When i decreased the voltage of the fan below 20V, the fins became feelable warmer to the touch.

Be aware, that using an infrared thermometer might give wrong readings as it probably measures the temps in the middle of the heatbreak - not at the top of the fins. In that regard the finger-test - carefully - is more reliable.

Your shroud might fit better to the 40mm fan. If it properly directs the airflow to the middle of the heatsink, everything might be fine. With my delta's comparably small hotend there's not much space to mount something airflow-optimized. And the 40mm fans diameter versus the heatbreak diameter is quite different - which creates quite a bit of an airflow-obstacle working against the fan.

https://www.thingiverse.com/thing:2084339 With that highly optimized (filament-cooling) duct one can feel, that almost as much air is coming out of the duct as when the hand is held directly in front of the fan. Take other ducts and almost no air's coming out.

running 3x 25x25x10 gostime 24v fans on my smarteffector with e3d v6. during last weeks tried several things to calm it down. long story short:

• 25x25x8 ebm pabst 252 n/h (quiet expensive - 22EUR, absolutely quiet, 12v with buck-converter) don't deliver enough pressure to cool heatbreak enough (v6 fins mean a lot of static counter pressure - might be different with mosquito)

for better comparison:

• 2 gostimes at 24v for filament-cooling press 2-3mm deep holes into water surface (glass below hotend)
• 2 gostimes at 12v are much quieter and less turbulent, but only scratch the water surface
• 25x25 pabsts just scratch surface
• 40x40x10/20 noctua just scratch surface

my solution right now: got a cheap buck converter via amazon bringing my voltage down from 24 to 20v. decreases airflow to a just acceptable level while decreasing noise and vibrations quite a bit.

airflow is one thing. getting that lot of air through a tightening channel or along obstacles in the way requires air-pressure (the anti-pressure to overcome is a reason for vibrations and noise). axial-fans are not good in that regard. pressure increases with revs. the ebm pabsts turn with 9(N - absolutely quiet) to 12(H - less quiet) thousand rpms. the gostimes in comparison can do up to 24.000 rpm (quiet noisy - but lot of flow in comparison).

https://www.ebmpapst.com/de/products/compact-fans/axial-compact-fans/axial_compact_fans_detail.php?pID=53709

http://gdstime.com/product/?99_471.html

the gostimes - even on full speed - can run comparably quiet. high pitch like a turbine. but as soon as they blow against an obstacle that blocks their way, the sound changes from a steady, quiet, high-turning turbine to a vibrating rattler.

might at least to some new ideas.

using smarteffector with 25mm fans on my delta. did a LOT of fiddling through the last weeks to somehow decrease noise- and vibration-level. different shrouds, fans and the like.

usually i had such a lot of noise and vibrations, i always had to bring my hotend-temp below 45°C to get reliable probing.

what i tried and dismissed during the last weeks:

• most thingiverse-shrouds don't take cfm/proper air-flow into account - they brake the flow that much down it doesn't make sense to mount them
• cfm-optimized shrouds for 40mm fans are to big for smarteffector
• 40mm fans seem generally to big for smarteffector (the difference 40mm vs diameter of e3d v6 heatbreak is to big - to much static pressure)
• 40mm noctuas (20 & 10 width) don't deliver enough pressure
• 40mm ebm pabst 40x40x10 at 24v doesn't deliver enough pressure
• non-direct, but remote air via silicone-tube: axial and radial fans don't deliver enough pressure - tried several versions (the fan's outlet-diameter has to be pretty close to the tube-diameter and also outlet - which renders fans useless unless one wants to use 25-30mm tubes)

so i stepped back to the 25mm ones - using 3 25x25x10 gostimes delivered with my printer-kit since about a year

• 25x25x8 ebm pabst are totally silent, but don't create enough pressure
• 25x25x10 gostimes don't deliver enough at 12v
• they certainly deliver enough at 24v (tried filament-fans with watertest - they stamped 2 holes 2-3mm deep into water-surface; hotend with finger on fins of heatbreak)
• combined with a stepdown - buck-converter - one can limit noise and vibrations by going down from 24 to about 20v - with lot less airpressure

effective against vibrations/noise:

• good bearings
• have fan blow freely (cfm-optimized shrouds; putting the fan under to much static-pressure stress introduces force, vibrations, noise)
• soft flex-filament spacers - where applicable - might help a little bit (but fixing the cause - non-free airflow - has a much bigger effect)

what i also found out: my magballs interfere quite a lot with the heatbreak-fan-motor. infact they can brake him down to a stall. burned one fan this morning. with quite a bit of fiddling found out that the rods magnets can brake down the fan significantly on wrong polarity. one can check it by bringing the magnet from the distance towards the ball while looking at the fan. if the fan's visually and noise-wise braking, turn the rod around and try with the other end. with 2 rods mounted in one corner - 1 being wrong polarity - my fan came to a halt.

that might not directly answer the above question, but perhaps give some ideas.

@dragonn i thought to go down this road.

@briskspirit hell! yes - you're right. my printer came with 24. will have to check that.

As already mentioned less torque from motors. On the other hand try to get your hands on 24V fans... With 12 you could just go and buy whatever is out there - eg. superquiet Noctua-fans.

do you have an led indicating probe-contact? if so, i could imagine, you'll see it lighting up very often. for example due to a vibrating fan.

the end-script of S3D did put the M226 into the GCode-file (checked it), but as said, it didn't do anything from withing the gcode.

the M25 - triggered via the DWC - worked as expected (enough for me to solve my problem).

if you need more informations, i'd be happy to help.

Dave, if that might be of interest: The M226 showed no reaction

• as a "end"-script in S3D
• fired as gcode from the DWC

running 2nd-last stable release (from about 4 weeks ago)

Checked while printing:

• M600 - unknown
• M226 - should be known, no reaction
• M25 - storing status for powerfail & moving to pause-position

Would have had the possibility to format with 64K in Win7, but wanted to give the formatter a try. Results are at least amazing. Factor 100!

Gents, just for the statistics: Max block time uSD

Card delivered with Duet3D

• 360-400ms

SanDisk uSD Ultra SDHC I 8GB

• formatted 4K (Win 7): 800ms!
• formatted 32K (SD-Formatter) : 7.9ms, 8.6 ms

I'd say, big thank you to @Phaedrux for pointing out https://www.sdcard.org/downloads/formatter_4/

I tried to get a filament change in mid-print going (as of example). With S3D i defined 2 different processes, the 1st starting @ height 0 and ending @ 8mm. The 2nd starting @ 8mm and then up. The first one got a (M600 first - doesn't work with Duet) M226 in the end-script tab.

What i observed using that M226:
While manually pausing and resuming a print with DWC my delta moves the nozzle out of the way as defined in pause.g and resumes perfectly as in resume.g

Once hitting the M226, it only switches off the filament-fans (what accoustically gives me a signal at least), but moves on without any other fuss.

I reckon, i'd have to investigate the resulting gcode-file lateron to see whether the slicer might have buggered it up, but generally i'd assume that others have properly working M226-commands with their duets?

Looking @ your M122-results further up, your max network-loops are just a minimum longer than your max write time for the uSD-card. Sounds to me as if the card's write-time would be the main-limiting factor.

As for my system, it's purely the uSD-card delivered with the duet. your 112ms max block write are a ferrari compared to my max 400ms i've had so far.

For negotiating the network-transfer without to much hassles i'd resort to (ignoring much more time-consuming methods):

• M122 SD max write and Network max loop
• perhaps Chrome Devtools (Ctrl + Shift + I) - network tab

Generally i'd try to avoid all unnecessary network-problems by taking the router and the printer alone without any other obstacles. If they run together properly, you can investigate to more complicated steps. If not - righdo - than at least you've got less complexity at hand.

Just had a go to get another result additional to the already seen one above. Seems the micro-sd delivered with the duet is the limiting factor.

SD card 0 detected, interface speed: 20.0MBytes/sec
SD card longest block write time: 392.1ms, max retries 0

=== Network ===
Slowest loop: 393.80ms; fastest: 0.01ms
WiFi signal strength -48dBm, reconnections 0, sleep mode modem

I've got 6 different routers lying around together with 4 different types of interchangeable antenna-types. Tried to get a line of sight connection of about 20 meters through 2 window-glasses / 1 glass + 1 transporter-metal-wall. The results are not to fine - to be diplomatic.

5 meters behind some obstacles + wrong antenna-type and wrong height-level can already become quite problematic.

In my case the router's a meter away on the opposite site of the printer's antenna (behind printer) and 40cm above. In my case the connection theoretically shouldn't be a problem.

Just reassembled it and about to print. i'll have a look for chrome's devtools. The long time for upload's bearable, but it would be nice to have it sorted out for everybody.

WEB-Upload: I had several such situations, too. Experienced very quick upload (with progress-bar visibly increasing) several times. And then it started to take ages (progress-bar no progress visible for minutes).

Latest stable firmware.

Yes. But probably firmware couldn't be gone, popup again once and be gone again from then on - although not being impossible. And that would also not account for the temp-error troubling the whole system once touch-screen is hooked up to controller.

Thanks Dave for your help. Really much appreciated! I reckon that it helped Roland, too.

@dc42 The usb-psu is working properly. i'd say, the touch-screen just doesn't say anything anymore.

Re-checked just a minute ago. The display is not showing anything no matter where or what power it's getting (printer on alone; usb powering the duet-board with leds properly on).

Powering the duet-board by USB always switched on the touch-screen before the problems occured. Right now nothing's happening anymore except that the 2.000°C-error + fan blowing is happening when the full connection duet-board - paneldue-controller-board - touch-screen is established and printer's powered.

My guess would be that the touch-screen has an error/is broken. Do you think the same?

• powering the PanelDue via a strong USB (capable 2A+)
  • USB + Printer on: no change, display off, temp-error, fan blowing
  • USB alone (Printer off: display off)
• INT_5V_EN: enabled
• U3
  • touch connected: U3 cold
  • touch dis-connected: U3 cold, errors disappear