Report on the PS1 mass prusa build, and an ask for help

[Sacha De'Angeli]

hi everyone,

I thought I'd report back to y'all about the mass prusa build that happened at Pumping Station One on June 18.

I was an attendee, but I also helped out a little with organizing and administrative stuff.

Basically, setting up a build of about 20 repraps at once is a huge pain in the ass, but ScribblJ (the organizer) took the challenge head on, and I really respect that moxie.

The primary lesson learned is don't bother using cast parts: they simply don't work as advertised.  

I'm almost done with my printer (ironing out electronics bugs), and as it stands, it's a mishmash of 30% cast, 20% printed, 50% custom MDF parts, and two plastic spoons (don't ask).  That's because I had to replace a lot of the cast parts in order to get a functioning gantry.  If you value your time at more than say $1/hour and you want to built a reprap, just buy the printed parts.

Unfortunately, this was the biggest obstacle to success with the class, and I certainly can't fault ScribbleJ for the decision - if you look, you won't find much negative written about using cast parts, but let me tell you: there are many instances where the design is moderately to disastrously faulty. 

Beyond the expected drilling of holes, all the parts need to be shaved, filed, expanded or modified to work, so a step that would take a few moments on a printed part takes up to an hour, all parts that have critical dimensions do not meet that critical dimension, and some simply can't be modified properly at all (without some specialty tools).

All of this led to slow class progress and some less than happy folks.

There are some bright spots, a few of us came up with an idea for making MDF replacements for the verticies (I created a decent template to post on thingiverse - I'll link to it here later).  Basically, you replace 6 hours of printing with 1 hour of really basic wood work, and you get perfectly aligned, heavy-duty verticies. 

Onto my ask for help:

I just wired up my motors and hooked them into my sanguinololu board, installed reprap software (on ubuntu 11.04), shorted the green wire on my ATX power supply to ground, plugged everything in, and flipped the switch, and nothing much happened.

The motors were "pulsing" gently and the pololu boards got pretty darn hot.

I entered 10.0 mm on each axis and clicked go, and nothing else happened.

where should I start with troubleshooting?

[Yepher]

I would say first step (which you probably have already done) is
determine if the failure is mechanical or electrical. If you disable
the motors can you manually move each of the axis? If not maybe the
problem is the axis is binding.

If you suspect electrical is it possible to connectors are wired
backwards?

Is there a way to adjust the driver voltages on your driver boards? If
so are those set properly?

I am just shooting in the dark.

-- Yepher

[Christopher Keller]

do you have an extra motor? can you disconnect a motor from the
hardware and allow it to freely spin? if so then you can do the
electronics dry run...

if not then yes, try turning the axes by hand to see if they are EASY
to turn. if the belts are AT ALL hard to move... you need to solve
that.

if, on the other hand, everything is butter smooth, then perhaps your
current trim-pots on the drivers are too low.... to set them, turn
them all the way (counter clockwise) GENTLY, you can break them... and
turn them up until the axes move freely without skipping, but no
higher.

you shouldn't have to set it above the middle, but that's just what I
THINK, I don't KNOW this.

Also - be sure you have heat sinks on the driver chips.

.... post a video on youtube, maybe we can help more :)

Regards,
  Christopher

> --
> You received this message because you are subscribed to the Google Groups "Chicago 3D Printer Enthusiasts" group.
> To post to this group, send email to chicago...@googlegroups.com.
> To unsubscribe from this group, send email to chicago-repra...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/chicago-reprap?hl=en.
>
>

[Sacha De'Angeli]

oh yeah, yes, the axes are all quite easy to move by hand, I've had friends with prusas verify that the resistance is fine.

I'm really thinking that my first step should be to verify that I'm actually communicating with the Sanguinololu.  Is there an easy way to do that.

In the meantime, I'll look at adjusting the pots.

[Yepher]

Sacha,

I am still not exactly sure how far along you are.

Were you able to flash the Sanguinololu OK? OR is this the first time
you are trying to communicate with the board?

-- Yepher

On Jun 29, 2:07 pm, "Sacha De'Angeli" <sachadeang...@gmail.com> wrote:
> oh yeah, yes, the axes are all quite easy to move by hand, I've had friends
> with prusas verify that the resistance is fine.
>
> I'm really thinking that my first step should be to verify that I'm actually
> communicating with the Sanguinololu.  Is there an easy way to do that.
>
> In the meantime, I'll look at adjusting the pots.
>

> On Wed, Jun 29, 2011 at 1:32 PM, Christopher Keller <cakelle...@gmail.com>wrote:
>
>
>
>
>
>
>
> > do you have an extra motor? can you disconnect a motor from the
> > hardware and allow it to freely spin?  if so then you can do the
> > electronics dry run...
>
> > if not then yes, try turning the axes by hand to see if they are EASY
> > to turn.  if the belts are AT ALL hard to move... you need to solve
> > that.
>
> > if, on the other hand, everything is butter smooth, then perhaps your
> > current trim-pots on the drivers are too low....  to set them, turn
> > them all the way (counter clockwise) GENTLY, you can break them... and
> > turn them up until the axes move freely without skipping, but no
> > higher.
>
> > you shouldn't have to set it above the middle, but that's just what I
> > THINK, I don't KNOW this.
>
> > Also - be sure you have heat sinks on the driver chips.
>
> > .... post a video on youtube, maybe we can help more :)
>
> > Regards,
> >   Christopher
>

[andres]

yeah. the nice thing about the system is you can run the components parts in isolation when you trouble shoot. (by components i mean software, electrical, and mechanical.)

mechanical is the easiest to see when there is a problem and you already looked into that.

for software/electronics, remember that you can feed gcode in manually with any ol serial terminal. Make sure your stepper drivers aren't pulling too much current by turning down the pots on the pololu drivers, and if you are really stuck you can use a normal arduino with a simple sketch to test each axis manually.

thats the quick run down, if you have any other questions let me know, i've definitely been right where you are, and we can get you past this easily enough.

andres