thereprapkitstore.com Prusa Mendel construction Notes/Discussions

Sorry yes you're right, the smooth rods fit snug once the ends are filed smooth.

Hi everyone,

I also received my ABSPrusa (batch of 31 may) here in Belgium last friday. I started the building on sunday.

Hopefully for all of us, this forum and specifically all of the people that makes it alive are there to help as I must admit that the support from the store is almost inexistent.

The only problem I had was with the belt of the X axis which is too short (85 cm).

I arrived at the end of the wiki and I still have to do the heatbed and the head.

I guess that I will make it for the electronics with the answers I already found here and the pictures of the machine completely finished (I hope I'll be able to ;o))

But to finish it, do we have to switch on the wiki of the original Prusa for the software and calibration (I guess there's one, I must admit I didn't even get a look on it) ?

To really finish, I have some questions:

- As asked by remondo but not answered, should we use a thermal compound (used on CPU) on the resistance of the head (this compound is ok for temperature up to 250°c) and also for the thermistor ?

- On the Z axis, we have a nut under the coupling of the stepper motors on the 210mm M8 threaded rods as we can see on the pictures of the wiki. What should we do with it ?

- About he kapton, any idea about how many layers we should use on the head ?

By the way, how is your building remondo ? Did you finish it ?

Thanks to every one and especially nophead and also sorry for my english that can be sometimes bad.

Wrt to the x axis belt, I found the same thing. Then I realised that the x and y axis belts are a different lengths. I remember cutting off loads of spare when I did the x axis and later found the y axis belt was only just long enough! So, a good tip is to not cut any belts until both are fitted.

I had some thermal paste which I used on the heater resistor but I did not use it on the thermistor.

Pratick Wrote:
-------------------------------------------------------
> Hi everyone,
>
> I also received my ABSPrusa (batch of 31 may) here
> in Belgium last friday. I started the building on
> sunday.
>
> Hopefully for all of us, this forum and
> specifically all of the people that makes it alive
> are there to help as I must admit that the support
> from the store is almost inexistent.
>
> The only problem I had was with the belt of the X
> axis which is too short (85 cm).
>
> I arrived at the end of the wiki and I still have
> to do the heatbed and the head.
>
> I guess that I will make it for the electronics
> with the answers I already found here and the
> pictures of the machine completely finished (I
> hope I'll be able to ;o))
>
> But to finish it, do we have to switch on the wiki
> of the original Prusa for the software and
> calibration (I guess there's one, I must admit I
> didn't even get a look on it) ?
>
> To really finish, I have some questions:
>
> - As asked by remondo but not answered, should we
> use a thermal compound (used on CPU) on the
> resistance of the head (this compound is ok for
> temperature up to 250°c) and also for the
> thermistor ?
>
> - On the Z axis, we have a nut under the coupling
> of the stepper motors on the 210mm M8 threaded
> rods as we can see on the pictures of the wiki.
> What should we do with it ?
>
> - About he kapton, any idea about how many layers
> we should use on the head ?
>
> By the way, how is your building remondo ? Did you
> finish it ?
>
> Thanks to every one and especially nophead and
> also sorry for my english that can be sometimes
> bad.

Hi pratick, I finally received my bearings yesterday so I got all the axes fitted with belts. I too found that one belt was too short but it actually does fit if you stretch it which is not a problem because it need to be tight. The shortest one goes on the y-axis. James needs to be made aware that one belt is too short and the other is too long! His communication was great at first but progressively got worse.

I'm struggling with the lack of instructions for the head and electronics but I'm referencing the Prusa wiki page and making it up as I go along! I used aluminium foil around the extruder resistor as the J-Head wiki page suggests, it took about 10 attempts but I finally got it to fit snug. I'll probably wire everything up later tonight, it's fairly self-explanatory but drivers etc is another matter.

I would also like to know what the nut closest to the z-axis coupling is for. Presumably it is to protect the coupling if the z-axis gets too high because you wouldn't be able to screw it all the way down to support the x-axis brackets.

Edited 1 time(s). Last edit at 06/13/2012 11:59AM by remondo.

The normal reason for the second nut is for anti-backlash. It fits into the top of the hexagonal pillar in the X end, trapping a strong spring. I don't know if that is what James intends. Are there two big springs in the kit?

Most people don't fit them as gravity is usually sufficient, although it wasn't for me for some reason.

I can't think what else it would be for.

---

[www.hydraraptor.blogspot.com]

finished with the electronics :-)
found the trick to put a 1k ohm resistor between the green cable and a black one to get the power supply working :-))
the red led on the singubumboard lights up:-)))
did plug in the usb connector and with internet-driver-lookup (windows xp), a usb-serial adapter was installed.
downloaded the reprap-host-software [github.com]
starting the reprap.bat , the extruder-stepper makes some noise.

Now i would like to enter some coordinates in the xyz-tab of that software and see my beloved reprap moving.

but nothing happens :'(

and i have no idea what to download or install or to read or google or whatsoever :-)

ideas welcome !


@remondo
> Tie a small nut to a piece of cotton thread to create a makeshift plumb line

no, the working table does not have to be perfectly flat :-(
my approach is based on the symmety of the machine itself...


@nophead
> You shouldn't need to remove any material in the X-ends if they are the ones I printed. The rods should just push in. I just tested one here.

One coupling was perfect and with the other coupling none of the four nuts did fit, no way.

Edited 1 time(s). Last edit at 06/13/2012 03:58PM by roland.

Hi again,

Hope you will get some help Roland.

Your building is more advanced than mine and it looks like you will be more helpful to me rather than me to you, sorry.

I have another question: in the kit there are endstop holders (one for each axis) but no endstops at all (microswitches, I suppose).

How could the printer go to a zero position at startup without switches to detect the zero of each axis ?

So why aren't they in the kit ? Or was it only missing in my kit and you did indeed receive yours ?

@nopheah: there are no strong springs so these nuts on the Z axis will continue to keep their secrets...

Any entry point on another wiki for the firmwares, softwares etc ... ?

Thanks again.

I had micro switches with wires as part of the second delivery of my kit. I'm guessing that you should have them as well.

I never had springs in my kit for the z axis. I used some that I had lying around.

Pratick Wrote:
-------------------------------------------------------
> Hi again,
>
> Hope you will get some help Roland.
>
> Your building is more advanced than mine and it
> looks like you will be more helpful to me rather
> than me to you, sorry.
>
> I have another question: in the kit there are
> endstop holders (one for each axis) but no
> endstops at all (microswitches, I suppose).
>
> How could the printer go to a zero position at
> startup without switches to detect the zero of
> each axis ?
>
> So why aren't they in the kit ? Or was it only
> missing in my kit and you did indeed receive yours
> ?
>
> @nopheah: there are no strong springs so these
> nuts on the Z axis will continue to keep their
> secrets...
>
> Any entry point on another wiki for the firmwares,
> softwares etc ... ?
>
> Thanks again.

there should be a plastic bag with three micro switches and green/black cables already soldered to it.
(in my kit, the micro sd card was missing.)

If the switches are missing indeed, get yourself opto-endstops. they are said to be more precise.

--Any entry point on another wiki for the firmwares, softwares etc ... ?

the heatbed: [reprap.org]
and definitely [reprap.org] to get the wireing right:


here is how to plug in the SD-adapter: [reprap.org]

and here how to install the reprap-host-software: [reprap.org]

but all these links did not really help me ;-)

Quote

One coupling was perfect and with the other coupling none of the four nuts did fit, no way.

It looks like it was printed too low leading to a ridge around the bottom. Sorry about that.

I recommend for firmware use Marlin and for a simple host use Pronterface, which has Slic3r built in. Those are the most popular choices at the moment I think.

---

[www.hydraraptor.blogspot.com]

-- I recommend for firmware use Marlin and for a simple host use Pronterface, which has Slic3r built in. Those are the most popular choices at the moment I think.

this looks like trial-and-error to me :-/
and i don't like python. python is no better than perl :-( whoever starts a software with python can not be very clever.

i guess, there is allready a firmware on that solulo-board. i can read "Sprinter" on the chip label.
the repraphostsoftware automaticly is using some /usb/serial00 as serial port and there is only my reprap plugged in.
if that software can not move the x-axis, propably no other can do.

i rather think, i am overlooking something obvious in the software-chain.

roland
www.wikimedici.com

that red LED on the singolalulo board should better be blinking some status codes :-/

MotoBarsteward Wrote:
> on the heater to line up. I will be re- wiring my
> heated bed as the wire supplied for the job is not
> man enough and I'm loosing nearly 2 volts in the
> wires. Consequently drafts and cooling make it
> very difficult to keep the bed at 85 degrees. I
> used 2mm2 speaker wire (soft and flexible) for the
> print head. I shall do the same for the bed.

That seems to be a 5 ohm resistor for the print head !
The supplied wires have less then 0.1 ohm.
So i did not see a need to use thicker cables for the print head.


U = R x I
If you are losing 2 Volt at 10 Amps, the cables have R = 0.5 ohm.
That sounds strange to me.
10 Amps at 12 Volt will be 120 Watt for the bed. that should be enough.

The supplied wires have only 0.4 mm² copper.
specific resistance for copper: 1,678 ⋅ 10−2 [ohm x mm² / m]
so 1 meter of that supplied wire has R = 0,04195 Ohm

With 10 Amps that will make 0,4195 Volt and a 5 Watt loss
which is 4.2% of the energy spent to heat the bed.

Still okay :-(
1.5mm² would be a 3 Watt loss.


roland
the little physicist.

Edited 2 time(s). Last edit at 06/14/2012 01:55AM by roland.

I measured 10.5 V at the heated bed and 12.1 V at the PSU connector to the controller PCB. The wires from the controller down to the heated bed do get noticeably warm. I guess some of my voltage drop will also be in the MOSFET, PCB connector and tracks on the board.

Using thicker wire for the bed may not provide much improvement but it is the easiest thing to do (my speaker wire is 4mm squ). It takes quite some time for my bed to get to the 85 Degrees that I use for my PLA first layer temperature. However, this could also be due to a badly tuned PID loop. I'll stick a scope on it tonight to see if full power is being supplied. At what frequency does the MOSTFET PWM the current?

I know that people run their beds at 100 Degrees for their first ABS layer and I would be very surprised if my bed could get that hot.

Further investigation needed, I think!

roland Wrote:
-------------------------------------------------------
> MotoBarsteward Wrote:
> > on the heater to line up. I will be re- wiring
> my
> > heated bed as the wire supplied for the job is
> not
> > man enough and I'm loosing nearly 2 volts in
> the
> > wires. Consequently drafts and cooling make it
> > very difficult to keep the bed at 85 degrees. I
> > used 2mm2 speaker wire (soft and flexible) for
> the
> > print head. I shall do the same for the bed.
>
> That seems to be a 5 ohm resistor for the print
> head !
> The supplied wires have less then 0.1 ohm.
> So i did not see a need to use thicker cables for
> the print head.
>
>
> U = R x I
> If you are losing 2 Volt at 10 Amps, the cables
> have R = 0.5 ohm.
> That sounds strange to me.
> 10 Amps at 12 Volt will be 120 Watt for the bed.
> that should be enough.
>
> The supplied wires have only 0.4 mm² copper.
> specific resistance for copper: 1,678 ⋅ 10−2
> so 1 meter of that supplied wire has R = 0,04195
> Ohm
>
> With 10 Amps that will make 0,4195 Volt and a 5
> Watt loss
> which is 4.2% of the energy spent to heat the
> bed.
>
> Still okay :-(
> 1.5mm² would be a 3 Watt loss.
>
>
> roland
> the little physicist.

Well yes, 5 Watt in 1 meter cable will become noticeably warm.
i will use 1.5mm² cables for the bed. but i think that for the head, the 0.4mm² cables are okay.

from [reprap.org] :

Quote

PCB Bed has two sides, one with the traces (bottom side) and one with silkscreen (top).
Printing on top side is safer, works great.
Printing on bottom is more effective and heats up quicker. (Though the LED, resistor and wire connections are liable to collision with the print head. Also the tracks could be damaged if the print head collides. Make sure your z-bed-springs are not too strong to prevent damage during a collision).

you could insulate the underneath. stuff some foam plastic beneath the heatbed that can take 100°C. Or minearl wool.

Or pull a plastic bag over the entire reprap :-)

I guess i will do that ! good idea to mount the electronics at the bottom ;-)
But the neodym-magnets inside the stepper motors must be kept below 60°C !

roland
www.liebwink.de

as you are allready printing.. would be nice if you quickly post your software chain and whether you flashed a new firmware.
Something better than

Quote
roland
found the trick to put a 1k ohm resistor between the green cable and a black one to get the power supply working :-))
the red led on the singubumboard lights up:-)))
did plug in the usb connector and with internet-driver-lookup (windows xp), a usb-serial adapter was installed.
downloaded the reprap-host-software.
starting the reprap.bat , the extruder-stepper makes some noise.
Now i would like to enter some coordinates in the xyz-tab of that software and see my beloved reprap moving.
but nothing happens :'(

Quote

But the neodym-magnets inside the stepper motors must be kept below 60°C

The motors are rated for 80C temperature rise, which would put them at about 100C. I run mine over well over 60C.

[www.slidesandballscrews.com]

The PSU is turned on by connecting green to one of the blacks, there is no need for a 1K resistor.

The LED on the Sanguinololu is wired to the 5V rail.

---

[www.hydraraptor.blogspot.com]

you missed the important steps..

(you are right, 80°C is the temperature when neodym magnets start losing their power)

Edited 1 time(s). Last edit at 06/14/2012 08:32AM by roland.

Just found something on Amazon...

Getting Started with RepRap: 3D Printing on Your Desktop

Nice to have had before we started building!

this online "book" might also be helpful: [reprapbook.appspot.com]
roland
www.engelGesucht.de

still don't know if i have to (re)install a firmware in order to make the reprap go...

- About he kapton, any idea about how many layers we should use on the head ?

i don't think this is needed at all with this ABSPrusa.
And you would need 1cm capton for that: [reprapbook.appspot.com]
I guess, the 5cm kapton should be put over the glass.
And of course only one layer. Otherwise it would insulate the Bed :-(

Putting kapton around the head can have the same effect :-(
If you would wrap that nice black extruder thing that allready has some ripples for cooling, the heat could not meet the air and the extruder would heat up even more. And i guess, that aluminium plate will dissipate the heat so nothing is going to melt.

roland: there is no need for a resistor between the green and black wires, I actually bought an ATX plug from ebay as I didn't want a rudimentary setup considering the printer will be powered on for long periods. If anything you may need a load on the 5V rail to ensure you get the full 12V which can be provided by a 4.7ohm 10W resistor between the red and black wires on a molex or sata connector. More info can be found here. You will have to buy your own SD card, I understand that because you were supplied with the sanguinololu board James also provided the SD card READER but not a card. I bought a 16GB one for about £6 from Amazon. As a side note the later batches all shipped with the melzi electronics.

Pratick you should have received micro switches, I did not receive any LM8UU bearings but James eventually got them to me after I chased him up. Make sure you let him know you did not receive any microswitches, make a point to say you need all of them because James misunderstood me and only sent 2 bearings.

I don't think you need Kapton tape on the glass do you? I bought some Isopropanol to clean and degrease the glass from here which I understand will allow for the first layer to stick.

I should have everything wired up tonight and will hopefully plug it in for the first time later. Somebody on the reprap IRC told me there is an updated something or other for the Melzi board on the github repo but all I remember is he said it was number 63. I also need a guide to explain what to do once everything is wired up. I don't know where to start with the Melzi board but I suspect I'll get there after some trial and error.

Edited 1 time(s). Last edit at 06/15/2012 12:04PM by remondo.

well.. as i can not print, all i know is that expoy resin sticks to glass but not to PE, PET, PP, ... I allways use that cheap kitchen PE foil to wrap my new stuff into.

Thank you all for your advices.

I sent a mail to James about the belt which is definetely too short for X or Y axis (1cm more was enough) and the missing micro switches. Of course I didn't receive any answers and I ordered a belt on ebay and bought micro switches to a local store (1€ each so I won't be ruined ;o))

Though I must admit that this machine is surely a good Mendel, I'm not sure I would recommend thereprapkitstore. Too few explanations (incomplete wiki), support must be found here (thank you all) and no answers (till now) to requests. Maybe it's the price of success, I have no doubt about James goodwill.

Patrick

I feel the same way, the kit is great once you have everything but the support is terrible.

I have everything wired up and turned my PSU on but nothing happened. I have shorted the green and black wires and the psu fan spins. The voltage on the power terminals is a constant 12V on the multimeter. Do I need to switch jumper 15 to draw power from the PSU and not USB? When I plugged the USB into my computer it installed on com port 3. I really don't know what to do from here, there is very little information on the Melzi electronics. I tried to run the RepRap host software but the Command Prompt gives the error ''java' is not recognized as an internal or external command'. I have Java installed but my OS is 64bit Windows which I've read can cause problems.

Anybody got an up to date guide on what I do from here?

Edited 1 time(s). Last edit at 06/15/2012 07:22PM by remondo.

Hi Remondo,

You suppose that James already programmed the Melzi card and so it should work when powered on. I still don't know if it's the case, so you could check the following wiki:

RepRap.org

You should consider what nophead did recommend to use as software on the PC side and also the firmware on the Melzi board.

You will see that the firmware needs to be modified to fit the hardware like the termistor look up table and some constants for the 3 axis calibration.

As mentioned in theMelzi wiki

Quote
This is a Sanguinololu firmware compatible board, though does have a couple of extra features the Sanguinololu does not

So we can use any settings defined for the Sanguinololu.

May be I'm wrong and all this has perhaps already been done by James...I'll see when I'll reach that point in the building. But if you say that nothing worked when you powered on....

Anyway, let me know about your progress.

Patrick

Edited 2 time(s). Last edit at 06/16/2012 12:27PM by Pratick.

It probably has Sprinter on it. Connect to it with Pronterface and type M115 and it will report the firmware version.

Proterface comes with Slic3r built in so that should be sufficient to print. It may need calibrating first, depending on how close the settings James has put on it match your machine.

However, my favorite language is Python, so evidently I am stupid, so you might want to ignore my advice like roland.

---

[www.hydraraptor.blogspot.com]

Once again nophead is right, it is actually sprinter.

I don't know Python myself as I program in C/C++ but I surely wouldn't pretend that all Python programmers are stupid (I let Roland assume this).

However, I still have a little problem:

- my stepper motors only work in one direction: I guess I must reverse connections of one of the 2 coils
- Proterface works a few minutes and then can't talk to the printer anymore with the message "Can't write to printer (disconnected?)". I'm running the software on a laptop under Windows 7 32 bits, using a speed of 115200 bps. To be able to reconnect, I must remove and reconnect the USB connection to the Melzi card. Any idea ?

Thanks again nophead for your valuable help.

Patrick

Edited 1 time(s). Last edit at 06/16/2012 06:23PM by Pratick.

After some extremely helpful nudges in the right direction from the guys in the RepRap IRC (including kliment who developed the software) I have signs of life from the Prusa! I used Pronterface (or Windows Installer) and simply plugged the USB cable into my computer and turned the power supply on and it connected right away! I did press the button on the Melzi board which seemed to bring it to life but it may have just been coincidence. Anybody know what the function of this button? I then updated slic3r to the latest version (replacing the files in the slic3r folder within pronterface folder). I tested the heated elements and they heated to optimum heat within a minute or so. I then levelled the x-axis by eye and I'm now consulting this blog post to calibrate the printer and extruder. M115 tells me the firmware is indeed Sprinter v1.

Edited 1 time(s). Last edit at 06/16/2012 06:45PM by remondo.

Pratick Wrote:
-------------------------------------------------------
> Once again nophead is right, it is actually
> sprinter.
>
> I don't know Python myself as I program in C/C++
> but I surely wouldn't pretend that all Python
> programmers are stupid (I let Roland assume
> this).

I have also written plenty of C++ when people pay me to do it but when I am programming in my own time I use Python because it is much more productive being a much higher level language and also much simpler. Who needs template meta programming ,,,

>
> However, I still have a little problem:
>
> - my stepper motors only work in one direction: I
> guess I must reverse connections of one of the 2
> coils

No there must be something wrong with your end stops. The firmware must think they are triggered. Have a look at the Mendel90 testing instructions [reprap.org], they will apply equally.

> - Proterface works a few minutes and then can't
> talk to the printer anymore with the message
> "Can't write to printer (disconnected?)". I'm
> running the software on a laptop under Windows 7
> 32 bits, using a speed of 115200 bps. To be able
> to reconnect, I must remove and reconnect the USB
> connection to the Melzi card. Any idea ?

Seems odd. Check that the voltage regulator is not getting too hot and shutting down. It could also be a dodgy USB cable, or a bad ground letting ground current flow through the USB cable,

Edited 1 time(s). Last edit at 06/16/2012 07:10PM by nophead.

---

[www.hydraraptor.blogspot.com]

Quote

Anybody know what the function of this button?

It simply resets the processor. You can use it as an emergency stop. If you fit the auto reset jumper the USB connection will automatically reset the processor. That is good for loading firmware but you need to remove it when running stand alone from an SDCARD. If you don't have it fitted then you need to press the reset button at just the right time to load firmware.

Edited 1 time(s). Last edit at 06/16/2012 07:11PM by nophead.

---

[www.hydraraptor.blogspot.com]

And finally the first print ... not really convincing.



If anyone can make me benefit of his knowledge, don't hesitate ;o))

I haven't checked the temperatures measured by the board, I'm still waiting for my temperature probe and I also didn't read anything else in the wiki to improve the quality but of course, I will.

Thanks again to all that have helped and more specifically nophead, I think that James should pay you for doing the support in its place.


Patrick