LaserCut Mendel by Tech Zone Remix

It looks like I have to eat my words (at least to some extent). I modified the firmware to echo the current temperature directly back to the host computer (connecting USB directly to the extruder board), and set the temperature to 100c. Temperatures went well above 100, and were continually jumping around. I then unplugged the 12v connection (because things were getting too hot), and the temperature it read immediately settled.

At a guess, noise on the heater circuit is causing communication troubles. Unfortunately, looking at the OneWire protocols, I don't see a good way to compensate, short of re-reading the data multiple times, and bit-averaging the values. Since it takes 40ms to read, doing that would probably mess up the servos.

I guess I can try cutting the cable apart, and separating the wires, to see if that helps, but honestly I'm starting to think that using serial communication to read the temperature is just a bad idea.

Quote
raldrich
I guess I can try cutting the cable apart, and separating the wires, to see if that helps, but honestly I'm starting to think that using serial communication to read the temperature is just a bad idea.

I came to that conclusion after using it for more than a month and returned to using a thermistor. I was hoping that you would prove me wrong.

Well, I found out what was keeping me from seeing valid temperatures in repsnapper - the RS485 cable between the two boards was bad. Now that I've fixed that, I can see how badly the PID code is behaving.

The problem seems to be that temperature values going all over the map (such as when the serial comms to the thermocouple are messy) really messes up the PID code - in many cases convincing it to set the voltage much higher than it should.

The odd thing is that when I was running testbench code (just thermocouple and PID, with no motor or intercom), I wasn't seeing nearly as much comms interference. It makes me wonder if the intercom code messes up timing for the OneWire library.

The other thing I'm wondering about is why the RS485 baud rate is set so high - 1,000,000 baud. Seems like total overkill now that step and direction are on their own pins.

Edited 1 time(s). Last edit at 06/09/2011 04:49AM by raldrich.

The one wire protocol needs some short, but accurate timing delays, which makes it a pain to do in software on a micro that is doing other things at the same time. You generally need to disable interrupts during each bit transfer to avoid the delays being too long, but then you tie up the CPU for about 75uS, not good for stepping, etc.

I certainly would not use it in this type of application without hardware support.

---

[www.hydraraptor.blogspot.com]

I have to agree, nophead - between all of the calls to delayMicroseconds, and all of the cli()/sei() calls peppering the OneWire.cpp file, it can't be healthy for the Servo library.

I'm still unsure whether it's noise on the "OneWire" or interrupts that are causing all of the flakey sensor reads. I may try modifying OneWire::read_bit() to multi-sample, and see if it helps at all.

Still, as soon as I get all the parts together to rebuild the hot end, I'm going to make one with a thermistor, and jettison OneWire. It's sad, because otherwise it's a fairly elegant hot end interface.

[Edit]

Apparently it was noise on the OneWire that was causing much of the trouble - If you replace the TechZone cable with actual ethernet cable, make sure that the OneWire signal wire gets paired with the nichrome wire's ground, rather than it's +12.

Now I just need to get matching mainboard and extruderboard firmware so I can get rid of the 1million baud hack that's in there, and run the RS-485 comms at an achievable baud rate.

Edited 1 time(s). Last edit at 06/11/2011 01:36AM by raldrich.

OK I was able to create an account finally. Not sure why no confirmation emails make it to my mac.com account but that's another problem...

Anyway my machine was working and I was able to print using repsnapper on Mac OSX. What I want is to create the GCODE with Skeinforge on the OSX machine and print on an Ubuntu or non intel Mac (Repsnapper doesn't run on it). So I have been struggling to get Skeinforge to create gcode that makes a good print and last night I managed to knock the extruder tip off the oak block. Not sure exactly what caused this but my bed was high and the tip was gauging into the aluminum due to poorly adjsutable endstops and adjustments to the crappy popsicle stick bearing supports. There may have been problems with the extruder trying to push more plastic through the tip then could flow through also assisting in the damage. I have ordered a couple more oak blocks and was tempted to preorder the metal extruder but then had second thoughts based on the substandard quality of the laser cut printer...

Anyway will somebody please share with me their skeinforge settings. I'm trying to use SF41 without reflashing the roms... I think I am having trouble with the speed settings which is probably compounded by not reflashing the roms... Or convince me to reflash the rom and give me some guidance on the e_mm_steps settings...

Cheers,
NATE

Hi N8Morriss,

Sorry to hear about your extruder troubles, I had similar issues with the oak block extruder. I went through several oak block type extruders until i finally broke down and bought the metal type extruder. It has been working great ever since i got it.

I found this post very helpful on configuring skeinforge settings:
[www.mendel-parts.com]

My current Settings are based off ohmeyes detailed post. I have yet to perfect the settingsfor my printer but I am now printing and using skeinforge.

I had issues with skeinforge generated gcode not moving the z-axis probably. The z-axis steppers were being moved to fast to turn. Hopefully you don't have that problem but if you do:
[forums.reprap.org]

Good luck!

Trevor

Does oak handle the temperature or does it char? I see various references to using wood but in my experience softwoods char at extruder temperatures. Do hardwoods handle higher temps? They are all basically the same stuff, just more dense.

---

[www.hydraraptor.blogspot.com]

[i205.photobucket.com]

Yeah they char. I think it is possible that a good design with oak could work well for an extruder tip, but i don't think I'll try it again... The extremely charred one on the right could have possibly been caused by an overtemp situation. The other two are blackened and never made it through 10mins of printing @ roughly 230c. The problem I experienced with this extruder tip design was that the copper tip gets pushed out of the oak.

NOPHEAD:
What temp are you running your bed at for PLA, I looked at your blog and I thought 55C should be OK.
[hydraraptor.blogspot.com]

I'm running my bed temp at 57C, printing techzone purchased PLA (extruded at 230C) and I am still getting a great amount of warping on large parts. I am printing on the blue tape though, is the print surface the problem or should I increase the temperature?

Quote
tcool1008
I'm running my bed temp at 57C, printing techzone purchased PLA (extruded at 230C) and I am still getting a great amount of warping on large parts. I am printing on the blue tape though, is the print surface the problem or should I increase the temperature?

I think the best bed surface for PLA is heated glass. I got window glass from a stained glass shop and had them cut and grind the edges and corners a bit. You can use bulldog clips to attach them on top of your heated bed. Check out Adrian's video on the reprap wiki.

Quote

NOPHEAD:
What temp are you running your bed at for PLA, I looked at your blog and I thought 55C should be OK.

Yes I do PLA on a bed at 55C. When using glass I do the first layer at 90C. For PET and blue tape I use 55C for all the layers.

To make blue tape work you have to do the first layer a bit low and the plastic a bit hotter on the first layer so it is liquid enough to soak into the tape. Adrian also used to rough up the tape and change it for each build I think.

Glass is by far the best for PLA. The only reason I use PET or Kapton is they also work with ABS. The first layer temperature required for glass seems to vary wildly according to the glass. The temperature for the rest of the build should be just below the glass transition of the plastic.

---

[www.hydraraptor.blogspot.com]

The oak tips from techzone have a few typical routes toward failure, though they usually hold up for a period of time, and would probably be better with a different choice of wood. I suspect they went with Oak due to the low thermal conductivity, but Oak maintains this property through a network of rather large pores, which tend to fill with polymer over time.

With experience of some ~10 wooden hot tips, these are the failure modes we've identified:

1) charring of the wooden threads, leading to the brass hot end popping out. This is exaccerbated by having a tip overheat and start smoking. If you can avoid this, it is less likely to happen.
2) cracking of the wood near the brass, letting the brass slip out
3) cracking of the wood between the brass and the teflon tube insert. Typically this is preceeded by polymer oozing from the pores connected to this juncture.

This is all using PLA - I'm sure results using ABS would be worse. I recently recieved some of their metal tip designs, and they seem quite good in comparison.

I'd like to suggest to techzone (maybe we can all sing a chorus to them together) that their kits include a 4th stepper controller and metal tip from the get go - even if they have to up their price to compensate, there's no reason not to sell the most complete product you can and i worry that this isn't happening, but could easily be happening with some input.

EtherDais Wrote:
-------------------------------------------------------
> The oak tips from techzone have a few typical
> routes toward failure, though they usually hold up
> for a period of time, and would probably be better
> with a different choice of wood. I suspect they
> went with Oak due to the low thermal conductivity,
> but Oak maintains this property through a network
> of rather large pores, which tend to fill with
> polymer over time.
>
> With experience of some ~10 wooden hot tips, these
> are the failure modes we've identified:
>
> 1) charring of the wooden threads, leading to the
> brass hot end popping out. This is exaccerbated
> by having a tip overheat and start smoking. If
> you can avoid this, it is less likely to happen.
>
> 2) cracking of the wood near the brass, letting
> the brass slip out
> 3) cracking of the wood between the brass and the
> teflon tube insert. Typically this is preceeded
> by polymer oozing from the pores connected to this
> juncture.
>
> This is all using PLA - I'm sure results using ABS
> would be worse. I recently recieved some of
> their metal tip designs, and they seem quite good
> in comparison.
>
So far, the oak block on mine is holding up, I think because of two reasons.

1) I've been very careful to avoid overheating the tip - I made sure to solve hot tip communication problems before attempting a print.

2) I believe that the Nylon liner in my tip is pressed very tightly to the brass tip, keeping plastic from leaking into the wood.

> I'd like to suggest to techzone (maybe we can all
> sing a chorus to them together) that their kits
> include a 4th stepper controller and metal tip
> from the get go - even if they have to up their
> price to compensate, there's no reason not to sell
> the most complete product you can and i worry that
> this isn't happening, but could easily be
> happening with some input.

TechZone appears to be transitioning to a "monolithic" design for their electronics, which uses a real stepper controller for the extruder. I suspect that they also intend to discontinue the wooden hot end.

The question is, what do they plan to do with their current stock of gen 3 electronics, and wooden hot ends? Keep sending them out to unsuspecting new customers? They'd do better to stop shipping them in kits, and sell them all as repair parts.

> With experience of some ~10 wooden hot tips, these
> are the failure modes we've identified:
>
> 1) charring of the wooden threads, leading to the
> brass hot end popping out. This is exaccerbated
> by having a tip overheat and start smoking. If
> you can avoid this, it is less likely to happen.
>
> 2) cracking of the wood near the brass, letting
> the brass slip out
> 3) cracking of the wood between the brass and the
> teflon tube insert. Typically this is preceeded
> by polymer oozing from the pores connected to this
> juncture.
>

Unfortunately, you can chalk up one more that has failed this way. I haven't disassembled it yet to see exactly what's going on, but it has pushed out several mm, and has been oozing PLA other than through the nozzle.
Is there a reasonable repair that can be made with the existing parts? Are there some types of plastic that would stand up to this temperature? What about MDF instead of oak?

I'm sure when you dissect your failed oak tips, you will find that there has been some crack or leak which originates from the porous bands which permeate their blocks. This is after something like ~5 of my own tips, and ~5 tips we've gone through in the course of last semester's class.

We believe that even using an Oak branch, with the 'center bore' drilled through the center of the branch would provide a significant improvement, as it wouldn't have the pore structure of the oak blocks which techzone generally provides.

We've considered using other types of wood as a replacment for the oak block design.
Gary Hodgeson suggests bamboo: [garyhodgson.com]

There are a number of discussions about using different hard woods for such things, inclusing references to this table: [en.wikipedia.org]
But, it's important to balance hardness (higher being better) with thermal conductivity (lower being better).

EtherDais, thanks for the tip about bamboo. I will see if I can work up something using it. As fate would have it, I ran into this problem printing out some parts in preparation for a demonstration of the printer at a local Linux User Group. Perhaps I can get things working again in time for the demo.

Looking back over some of the earlier messages in this topic, I see that there was some discussion about communication problems possibly affecting the temperature readings. I am seeing quite a lot of errors. There is a constant parade of checksum error messages in the terminal window that I launched repsnapper from. The most common error is

** (repsnapper:2304): WARNING **: Error (-4) 'Serial Error: Received a LineNr code without a Checksum code
 or Checksum without LineNr' - user popup?

Occasionally I get a pop-up window that says "Error reading from device! Resource temporarily unavailable". Printing stops when this comes up, but usually printing resumes when I close this window. When it doesn't, I click the "Kick" button in Repsnapper, and usually that gets things going. Once in a while, it will not start back and I'll have to just start over.

But things should not be this way, and perhaps it is contributing to my hot-end problems by causing problems reading the temperature. So if anyone has suggestions about how to resolve these comm errors, I'd be grateful.

I got things apart and found that the hot end hadn't pushed out of the wood, the oak block seems to be a lamination of several layers and the interface between two layers split apart. There was a single screw through the layers to hold them together, but it spread apart on the end opposite that screw. [edit] Now that I've removed some of the PLA, I can see that I was mistaken, it's a solid block of wood, and it split along the grain. There is some charring around the hot end, but not enough that it had pushed out of the wood. The split began at the upper edge of the hot end.

On the plus side, considering the construction, I think I may try just getting a piece of wood (for now) and trying to re-build it as-built, just so I can get through a Linux User Group demonstration I have scheduled for this evening. Long-term, I will try to find some other material to replace the wood.

Apparently the wiki doesn't like picasa photos - they aren't showing up in the preview of this message. So here are the links to them:


Here's a photo with the hot end attached to the extruder:
[picasaweb.google.com]

and a close-up after taking it off:
[picasaweb.google.com]

and here's a photo after cleaning out the PLA and separating the broken parts:
[picasaweb.google.com]

I wonder if my feed rate is too high? I was definitely getting too-thin an extrusion the last couple of parts, and it didn't seem that it was too thick previously.

Edited 1 time(s). Last edit at 07/12/2011 09:28AM by ShayWalters.

Okay. So I'm back to working on this after another several months. I've switched over to a thermistor now, and I'm looking for firmware to put on it. I've got a patched together version of the 'plain old' reprap firmware mostly working, in that it moves the X, Y and Z axes and reads temperature, but the extruder stepper doesn't turn. I'm still poking at getting that working, but I'm wondering if anybody what you guys would recommend.

I've got the techzone remix electronics(Not the monolithic board).

Hey Shaywalters,

Ive had similar experience with techzones poor tip design. I purchased the following from CdnReprap
[reprap.org]

Just PM him and tell him you want to purchase one. He makes them and sends them from Canada.

This is more reliable and only 45$



ShayWalters Wrote:
-------------------------------------------------------
> I got things apart and found that the hot end
> hadn't pushed out of the wood, the oak block seems
> to be a lamination of several layers and the
> interface between two layers split apart. There
> was a single screw through the layers to hold them
> together, but it spread apart on the end opposite
> that screw. Now that I've removed some of the
> PLA, I can see that I was mistaken, it's a solid
> block of wood, and it split along the grain.
> There is some charring around the hot end, but not
> enough that it had pushed out of the wood. The
> split began at the upper edge of the hot end.
>
> On the plus side, considering the construction, I
> think I may try just getting a piece of wood (for
> now) and trying to re-build it as-built, just so I
> can get through a Linux User Group demonstration I
> have scheduled for this evening. Long-term, I
> will try to find some other material to replace
> the wood.
>
> Apparently the wiki doesn't like picasa photos -
> they aren't showing up in the preview of this
> message. So here are the links to them:
>
>
> Here's a photo with the hot end attached to the
> extruder:
> [picasaweb.google.com]
> EZvl5_oVKA?feat=directlink
>
> and a close-up after taking it off:
> [picasaweb.google.com]
> LD1UjzvkNQ?feat=directlink
>
> and here's a photo after cleaning out the PLA and
> separating the broken parts:
> [picasaweb.google.com]
> KzxdXKp55g?feat=directlink
>
> I wonder if my feed rate is too high? I was
> definitely getting too-thin an extrusion the last
> couple of parts, and it didn't seem that it was
> too thick previously.

Hi,

It seems like a few of you have got further than me with the TechZone kit, which is encouraging for me I started very slowly with mine but have been making more progress since the new set of instructions went up. The problem I'm having at the moment is with assembling the 2, 3 and 6 pin ribbon cables as described here:
[www.reprap.org]

I split the ends of the cable, tucked them into the space past the contacts, folded them over the tops of the contacts and into the slots on the other side of the cable that hold everything in place, and then poked the wires into the contacts with the smallest screwdriver I have, which seemed to be easily small enough to do the job. The resulting cable has no electrical connection between its ends. The description seems to imply that if you do it right, the contacts will somehow pierce through the wires (or at least through the insulation to make contact with the wires), but I don't see how that can happen. There are no sharp points I can see on the contacts, and the wire fits comfortably into the space in the contacts without anything getting bent out of shape or coming close to piercing the insulation.

I tried stripping the insulation off some of the ends, but then the wires are so thin that they simply fall out of the connector. I can't see any way to keep them in short of soldering them, which looks like a pretty fiddly and difficult thing to get right.

Where am I going wrong? Can anybody describe the process in more detail, and/or with better pictures? Failing that, where can I buy some suitable pre-assembled cables of the right length with the right connectors? I'm in the UK

electrodruid,

Hi sorry to hear you are having trouble with the ribbon cable connectors. This was also somewhat of a problem for me because the screwdriver kept slipping and hitting my fingers! If you look at the top of the connectors there should be sharp metallic slots. I would recommend putting the the connector in a vice then positioning the wire over the slot, then press the wire very firmly down into the slots with a small flat head screwdriver.

In between the slots are sharp, this should cut the wire insulation and make contact between the wire and the terminal. I would definitely check each connection with a multimeter after you are done.

Here is a top view of a 2 pin connector(look for the metallic slots):
[i205.photobucket.com]

If you have any further problems post a pic of what your doing and hopefully we can figure it out.

hope this helps,
Trevor

I find it makes an easier result with those connectors to first remove the insulation from half a centimeter of the wires and fold the stripped metal onto the insulated part.
You then proceed as recommended on the wiki, making sure the stripped part is under the plastic, so that even if you don't pierce the plastic the contact will be made anyway.

Make sure to check for shorts before using the cable then.

Hi guys,

thanks for the advice on the ribbon cable assembly. I've successfully built the 6 pin cable now. I have another question about the electronics - specifically, I'd like someone to confirm how many cables of each type I should end up with, and what each of the cables is used for.

This is what I got in the kit:

1 * 10 pin ribbon cable
1 * 6 pin ribbon cable
1 * 3 pin ribbon cable
1 * 2 pin ribbon cable
6 * 10 pin connectors (black)
2 * 6 pin connectors (blue)
8 * 3 pin connectors (7 green ones, 1 blue one)
3 * 2 pin connectors (blue)
4 * opto endstops

From the quantities of the parts and the instructions, it's easy to infer that I should cut up some of the cables and make 3 * 10 pin cables and 1 * 6 pin cable. The 3-pins and 2-pins are more confusing though. One of the 3-pin connectors (a green one, FWIW) is attached to the CAT5 cable in the tip assembly, so I'm left with 7 * 3-pin connectors and 3 * 2-pin connectors. The instructions seem to imply I need to build 3 * 3-pin cables and 2 * 2 pin cables - which I could do if I cut one of the 3-pin connectors down to a 2-pin, but it would leave me with one endstop I can't connect to anything. If I actually need 4 * 3-pin cables then I'm missing a 3-pin connector and a 2-pin connector. What am I supposed to be doing here?

electrodruid - We've been having HUGE issues with missing parts on our orders lately. Techzone is, thankfully, generally willing to correct their mistakes when informed kindly.

I believe you have an extra 3 pin connector that should be a 2 pin connector. This is not a big deal. Just use it for that and make sure you know what wire is going into what hole. 3 3 pin cables are used for the optos, and 2 2 pin cables are used for communication from the motherboard to the extruder controller/extruder stepper controller.

The documentation for the Gen 3 boards has some images which may help you.

I just got done assembling my tech zone laser cut mendel and reading through this thread. The consensus I have come to is that I am in way over my head. I get some blinking lights and can't communicate with the machine all i get is a constant electrical whirl from all my motors it changes pitch when I open my reprap software but won't jog or do anything. I have downloaded more software than I know what to do with and haven't heard back from tech zone.

Devos,

I wish there was a subforum to direct you to for techzone help, but that seems to be this thread. =/

You may want to start off with a description of what they sent you, then we can work toward solving the problems.

Devos, you might want to visit the reprap IRC. There's plenty of helpful people there, some of which have already been through the "TechZone Experience"

Here is the update, I have got all of the individual components working using repsnapper I can control the machine but I can't get any code out of my models using repsnapper on(win7 64). I have loaded Skeinforge but can't make heads or tails of it, I have tried to generate code in replicatorg and import that into repsnapper but it just moves to a corner and stays there making lots of small moves and not much else.

The learning curve has been brutal on this project. After going through the hassle that is know as bot maker we got a full mendel kit from techzone and was excited that it actually showed up although was very disappointed at the amount of modification that the parts took to make them work and the smoke when you get the extrude tip up to temp. I have been going to Metrics for a lot of help and that has greatly lowered my stress level although not my bosses who though we would be printing parts after putting the thing together.

Right now I am planning on hand coding for a bit just to make sure I have everything working correctly and then trying to figure out why my stl aren't getting sliced and coded in repsnapper.

After a bit of a hiatus (some of it spent waiting for a replacement tip assembly which I was told by the UK supplier that I'd be getting, which hasn't materialised), I'm back trying to finish the electronics on my Mendel. I got the Generation 3 Electronics, and I've got the cables made, connected all the boards up, and hooked in the motors. I'm totally stuck with how to actually connect up a power supply, though. I have a mains adaptor (with a US plug, but I can get a UK one), a socket for the adapter to plug into, some yellow and black wire, a metal plate and a couple of screws, but it's not at all obvious to me how to actually hook it all up to the Mendel. I've found several photos showing how to assemble all the rest of the electronics (Example: [reprap.org]) but nothing that shows the power supply connections. I don't see a space on any of the boards to connect the socket to, or have any idea how to get the power from the socket to all the different boards that need it.

I'm probably missing something incredibly obvious, but can someone describe or show pictures for how they've done it? It'd be much appreciated...

In short: You need to find and hook up the 12v and ground wires from the psu to the electronics. That means to the extruder board and the stepper controllers. The motherboard gets power from the usb, but will benefit from an extra grounding cable to the extruder.

Pic of the extruder 12v supply: [reprap.org]

---

--
-Nudel
Blog with RepRap Comic