Teacup Firmware replicatorG & Skeinforge

Architect

Thanks for those tips I will try them.

My baud is currently set at the usual 115200 so dropping it a bit may help. never hurts with comms issues.

Have'nt adjusted current yet. next stop though.

Endstops I will check too. Question ? Did triffid get around to using a separate loop for homing or if end-stops are disabled does this turn them off completely ??

I would like it to home at least.

I need to run up some decent thermistor values too. My thermometer claims the temps are out a bit. Just easing it up 10 degrees a time to 250'ish to check out the ceramic glue used to fasten the thermistors in with and proove the power resistor heater.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

One more thing as i was thinking about the symptoms you described, what is your steps_per_mm for the xyz axis and speed settings?
if these are off and the stepper drivers are not adjusted then you might see one work until it gets to low of power and freezes up then the firm ware waits till it sends the next move. this might also happen if your speed in repG is too low and the steppers are not getting a good enough signal to move them.
as for Glue on a heater i fond that silicon tape rated @450C works much better then glue and kapton to hold my thermistor in. plus side its an insulator. i think i posted a pict on this thread a little bit ago of the heater.

I use cerastril ceramic glue stuff. nophead recommended it years ago and I have a small sample.

Looks like white powder, mixes with water to make a thick sludge and sets like ceramic.

With anything new I like to ease the temperature up by stages and burn it in.

Gives me confidence that it is working about right and the thermistor is reading. The PID needs a tune up as it overshoots badly, easy enough to do though.

Step etc are all as out of the box in the Ramps config.h so I am guessing they are what other Teacup RAMPS users have tried out. But worth reviewing.

From what you describe I would expect it to miss steps rather than freeze up. If it was the steps setting. I have not got as far as the fine tuning yet, just working on does it sort of do what it should ish.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

OK #teacup on chat.freenode.net was already taken.

But #ProjectTeacup is good to go. Just opened the channel.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

cool, I'll see you on there. :-) as for the glue... I'm not very good at measuring twice,cutting once. ;-)
i like to be able to take stuff apart completely. plus the silicone tape actually melds together to form a tight sealant and insulator around my heater. i can touch it @ 200C and not get burned. yet cut it off to remove the thermistor.
it also looks rather nice and comes in various colors for those who might be keen to that.

also, i do believe Nop said in that post that the glue degenerates over time and usage.
could be wrong about that though... here's the post where they talked about it in the comments section

I like the idea of covering the hot bits with silicone as an insulator. Thats clever.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Just had a rummage through the back posts here looking for piccys of your extruder hot end but couldn't find one. lots of electronics bits though.

Any chance you could post one up when you get a spare mo. I would like to see it.

Cheers

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

sure here ya go

Chased all over the place locally today looking for a tape supplier only to find that they are a wholesale supplier fo the most common stuff and Kapton/Polyamide is a special. Ugg They could have said on their web site instead of trying to convey the impression they had it over the counter.

Architect, nice picky thanks, much appreciated. I also hoped the same tape supplier would do silicone tape but they are probably the most difficult company to do business with that I have ever encountered.

Looking onwards and hopefully upwards.....

I think there is some mileage in skinning off the hot bits with some form of insulation, somehow.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

So, I've got Teacup running on my handy 328 and finally got repG to communicate with it. I can even load g-codes and get xyz automated axis movements.

The realtime control feature is working fine for the xyz axis but I can't seem to figure out how to make it work with the stepper extruder for some reason. I'm assuming you should be able to input a pmw or rpm into the box and it should start moving? I have the extruder moving somewhat using the reprap host, so I'm fairly certain the hardware is all functioning.

I'm running the ramps for mega profile at 112500 baud.

Any help/ideas would be greatly appreciated.

I don't believe Rep G has stepper controls in the interface for the extruder. I do my testing with either the command line or gCode (or sometimes ill use repsnapper). Skienforge has a control tab labeled Dimension this controls your stepper federate when carving an stl into gCode.

Nearest I worked out with RepG24 is that you set an RPM, make sure the enable stepper button has been selected and then select the direction. (Or other way around for the last two, but it does bring about a change in running state for the extruder stepper, and you should see the control G-code/s go in the sent window under RepG)

This seems to make my stepper do something, but it doesn't run. I am of the opinion at the moment that it not running is electronic (part of the fault finding and correcting thing that is commissioning).

My extruder stepper seems to judder rather than spin after being told to do something until I tell it to stop.

So I think the RepG controls are sensible enough and the issues are currently with firmware/electronics for this issue. Not entirely sure which. Gut feel at the moment is the Stepper driver electronics. Just wired up another stepper today that I am confident of the wiring for so will test tomorrow. (metered the other one out and it looks to be the same but never entirely sure the phases of coils are correct to the colour coding between model numbers, some are even wrong within model numbers)

So a certain amount of hit, miss and empirical learning going off.

Hope this helps.

Cheers

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

I believe most of the remaining problems with ReplicatorG and Teacup involve flow control. I think there's no concept of timeouts or retries, so it just keeps crashing if it either doesn't see something it expects, or gets something it doesn't expect. I noticed that the frequency of freezups increases with CPU load. The reprap5D driver may just be incorrect. I still haven't found very much from users indicating that ReplicatorG works flawlessly with other reprap firmware, except maybe ultimaker?

interesting i was/am using 23 right now and never actually used the extruder controllers in the window. not even really sure they are controls. it crashes every time I try to do anything. might look at 24 or the latest build.

my copy of 24 had the same problem. just did a git pull for the latest. doubt I will be able to find the issue tonight as I have an extruder issue in my hardware to solve but I haven't looked at the host software in awhile so I thought I'd update it.
as for my own issues, what is everyones working feedrate for the stepper extruder? I'm using wades extruder and have it set at
STEPS_PER_MM_E 36
MAXIMUM_FEEDRATE_E 4500
SEARCH_FEEDRATE_E 2000


and what about temps for PLA? I'm melting at 230C which seems high from what I've read.

anyways finished my LCD circuit board. its a 128x64 pixel lcd running off a Arduino Nano with a joystick input all connected up to teacup over I2c.
pretty cool. as of now I only have a temp graph running and input from the joystick on the display but I have plenty of room for more displays.

OK, onward with testing.

Also finished my first plug and play SMD Heater controller board. has extruder heater control of 12v via mosfet and 100K thermistor input nd fan connection (don't use it) this plugs right into my breadboard then out to my Arduino. same think as the polulo drivers. red led to tell me when voltage is applied to the heater and a green led to tell me its powered on. yea! thats success for me as I wasn't even sure which way to hold a soldering iron when i started.

WOW!
at first the freeze was a minor annoyance. now with the current git pull it is completely broke.
I only get about 5~10 temp readings before it freezes. if I reboot the MC I can get another 5 readings before it freezes again.
gonna need to put this out on the replicatorg dev list. meanwhile I'm reverting back to 23 for now.

Architect 230 certainly is high for PLA.

Usual for ABS (higher temp plastic than PLA is around 235.

I have no experience with PLA so can't give you real sensible figures.

Camile at MendelParts seems to work exclusively in PLA so would be a good guy to ask (He is frequently on IRC #reprap). As well as having a quick look at the Reprap wiki for info and maybe the BitsforBytes site I seem to remember they had a useful table with values in for different plastics that was an easy way to start off in the right ballpark.

What really works best for you will depend on what is in the formulation for your filament. SO you will have to experiment a little.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

I extrude PLA successfully at 210°C.

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse

Hmmm swapped the extruder stepper and the pololu driver board with one of the other axes.

The axis I swapped with works fine and whichever motor or driver board I use the stepper motor still just judders so there is perhaps something funny going off either with the firmware or the configured pin-outs for the extruder driver.

My Thermistor table is a touch out too. fine at low temperatures but an increasing margin of error at higher temperatures. I ran it up to 260 to get about 160. So not right yet, back to the python script. At 160 I got some extrusion if i pushed a filament through the bowden cable by hand, so a partial result there.

Edited 1 time(s). Last edit at 03/30/2011 10:48AM by aka47.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

sounds like your DIR pin isn't getting signal. that happens to me all the time cuz my drivers are still breadboarded and i fumble around the wires entirely too much.
:-)

Yeah, I think so too, time to move the commissioning into the electronics workshop. And use some reasonable test gear.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Hehehe, I sacrificed 2 reasabily good 3mm printheads this weekend for an as of tonight untested 1.75mm printhead. All because I ran outta 3mm plastic and every supplier online seems to be sold out. I guess that changed my parameters for me. :-) so far this weekend I got 1/2'dozen good test prints outta teacup. Started mOving on to host software and slicer. Thinking I might adjust my temps lut due to the fac I seem to have to print 3mm pla at 225 so I'm asumming my machine will need to hover around 260 for abs. Find out soon enough. Anyone out there have troubles getting plastic to flow? Once I get about 100mm flow then I know I can start a print. But if my heater cools bellow 205 I have to reflow another 100mm before printing again. Wondering if I should just setup up a starter routine in Gcode or if this might be a hardware issue. With the replicatorG instabilities it makes for a pain in the butt to get a print started. Once it does though no problems.

Your temperatures are sounding a little high there.

That is if they are that temp. Do you have a thermometer or something to measure what your nozzle is at really.

The figure you get read out from the firmware is heavily influenced by the lut.

I have to calibrate mine as it reads a lot hotter than it is really. At the moment I am checking using a temperature probe that goes up to 300 degrees C I got from a catering equipment suppliers.

Been fancying one of those IR gun things that measure temperature though primarily because I think it will be a touch more accurate and quicker than the amount of time I have to hold the probe against the nozzle.

If your lut says that ADC reading x is y degrees the firmware will report that irrespective of the fact that reading x is actually really y -/+ error Degrees.

Thermistors being non linear can produce interesting results. Particulalry if there are not enough values in the LUT or they are calculated for non optimal positions on the Thermistors characteristic.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

I undrestand that there's more than one type of PLA. I've found that the PLA (4043D) that I buy from Ultimachine needs to be extruded above 200. Below that it won't flow. I normally extrude at 230C for the first layer on glass, then I lower it down to 210 for the rest. I use a thermocouple on my heating block (old Adrian style) which I calibrated with another thermocouple inside the tip.

I have an IR gun. It's no good, as Nophead pointed out, for measuring tip temperature. It's too small for the gun. Even if you can, it won't be the right temperature since it only measures surface temperature and you need the temperature inside the tip. I've used it to measure the bed temperature and to see how much it varies across the surface.

The old way of using nichrome wire for heat and taping the thermistor over it gives low temperature reading. The new way using an aluminum or brass heating block with resistor gives a higher reading than the tip core. On mine, tip is 10C cooler than the reading at 200C and above.

yep thats near what i have experienced as well. I have a DMM with thermistor on it. I calibrated my thermistor to within 4 degrees Celsius to the inside of the barrel. in practice though given the way we handle temps and the readings it is often more like 10 Degrees off when over 220 Celsius.

the PLA I have is from Ultimachine.com as well. thanks for the info that means I'm not crazy and getting the same results as someone else.

Big News All,

ReplicatorG just had a new Driver Beta test release today by a Ben Jackson. I've been testing it a bit this evening with a new print head that I'm working on and it seems much more stable with Teacup.
here is the Source code link on git ReplicatorG245

I'd all for testing cuz this saved me a lot of work.
please give it a shot and let him know if you have issues.

><,'>

Architect

I have 8 different colors of PLA (4043D) and they all print well at 210°C.

---

Bob Morrison
Wörth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse

OK

Temperatures are now calibrated the manual way so are close enough for me for now. (Might revisit them later as I can be a bit picky, but will wait till I have a functional setup first before refining)

Extruder is still playing up in RepG24 even using bens version. ie it does'nt. But will under Repsnapper

As you observed (I can confirm) there appears to be little or no flow control going off between teacup and RapG24.

I sliced and attempted to build the mini-mug (Nothing clever) and it was something of a disaster. Although nothing broken.

Using Bens version of RepG245, it seems to ignore the driver info in machines.xml and insists on trying to home the Z axis to the max end even though the xml clearly states ensdstops min for all 3 of my axes. I aborted this manually when I saw what th G-COde was doing before anything got trashed.

Running the print job for the minimug after slicing it did'nt seem to pause for the head to heat up and also insisted on setting a bed temperature etc in the G-Code for a seting in the XML description that clearly said there was no bed.

Don't know what is going on with RepG but it seems to have problems understanding its own config.

Any chance someone could post some "Known Good" Teacup Test G-Code that works and I will try to print it from RepG to seperate out what Skeinforge is being told to do from what RepG is having trouble interacting with the machine with ??

I have tweaked the steps per mm etc but need to print something of know dimensions to check the scaling.

---

Necessity hopefully becomes the absentee parent of successfully invented children.

When i get home tonight I will post up a 20MM cube stl I've been using to test with. it slices in about 5 min. and prints in about 30 min. I've gotten it to be within .5mm in dimensions. i.e 19.5x19.5x20.5
I've noticed the heater bed thing to as well as my last few skeinforge changes have made the temp M104 commands disappear. they were there yesterday? it happened around the same time the heated print bed command showed up. and I just got my 6mm aluminum sheet in yesterday so maybe Its just a sign to make my heated platform ;-) of course i gotta get my extruder more reliable. Been doing a lot over the last few weeks mechanically hence the reason i haven't posted much here.
looking to rebuild my hot end and make my wades somehow easily switchable between 3mm and 1.75mm hot ends.
then onto fixing my end stops as i have suggested. found out last night i have broken one of the platform mounts by crashing my extruder into the bed. Not good. the reason for having end-stops in the first place :-)

Thanks I would appreciate that. If you could post up the G-Code that would be great. I can compare what RepG is turning out for me for the same object. Which might shed some light on what is going off.

I am rapidly getting to the point where I ready to bin RepG off and try using Skeinforge direct to file then streaming this to the machine using script.

I need to prove the machinery working and then regroup to mess with other stuff. Got some stuff I need to print.

My home position is Bottom, Left, Back and is set up to Min Endstops so G161 X0 Y0 Z0 give me a start from here point that matches.

No idea why RepG is trying to home to Z Max fatal for me as there is no endstop on Z Max and I need to print the bracket to add one.

Cheers

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

that is odd, I delete out the start.gcode that gets appended to the beginning of the gcode. until i get homing taken care of and decide what I want my Mendel to do at the beginning of each print, i just home it manually. then hit print. once replicatorg starts printing i have no issues with it at all. what i tell it to print on the other hand... :-)