Going crazy configurating extruder settings for TOM, sailfish and acceleration

[Daniel Springwald]

Hi,

I am working many hours to find out what are usefull settings for my TOM - without success :-/

Seems to me to move in a circular course:

When I use a "Filament Packing Desity (ratio)" of 0.85 the print has hills at the edges (see picture) and at the same time the infill-lines have disruptions and holes.

Changing the value higher than 0.85 the edges get a litte better, but the hole are bigger. Lower than 0.85 are very bad edges but less holes.

So it seems to be too much material and too less at the same time ?!?

My machine:

TOM ABP MK6 

Thing 32084, Sailfish 4.0 Rev. 720_

ReplicatorG 0040

0.4 Nozzle

ABS 3mm

Thanks a lot for any hint!

Daniel

[David Slater]

Hi Daniel,

  I have not run sailfish yet.  So I dont know what I am talking about.  But I will eventually and what I will do is put everyting back to default and do a test print.  If the corners are still rising like in the above pix I would lie to bot about the filament diameter.  I would tell it the filament was fatter than it really is in 0.03 mm increments.

later, slater

--
You received this message because you are subscribed to the Google Groups "Vintage Makerbot" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vintage-makerb...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Daniel Springwald]

Hi David,

thanks for your reply.

As I wrote I allread tried to increment the fillament diameter.
When doing this the corners get better but the holes and disruptions inside the infill get worser. Lower diameter makes better infill but worser corners.

This drives me crazy :-/

Best
Daniel

[David Slater]

Hi Daniel,   Well I guess this is the time to bite the bullet.  I wanted to delay but I will "upgrade" to sailfish 4.0 rev720 and rep 40.   and see what I see.... please send me the file you are printing.
later, slater

[David Slater]

Have you done much googling on the issue.  Seems to me I saw something about the type of infill you use with one having a bug.

[Daniel Springwald]

Hi David,

it´s the 20mm calibration cube. I just interrupted the print to show the infill problem, too.

Best
Daniel

[Daniel Springwald]

Hi,

yes I googeled a lot about the problem ;-)

I only found descriptions of a bug in hexagonal infill, but the effect of this bug would be 100% infill.
But my problem is too less infill.

Best
Daniel

[David Slater]

That hexagonal may be what I remember.  Are you running under XP?  I think I understand that sailfish uses the infill number .085 to define filamant diameter?

[Dan Newman]

On 05/07/2014, 2:32 PM, David Slater wrote:
> That hexagonal may be what I remember.

In SF-50 (and maybe 47), the hex infill in Skeinforge has a bug
when you do 100% infill. Lower infill percentages should work
okay.

Dan

[David Slater]

Thanks Dan !

Daniel is using:

 

My machine:

TOM ABP MK6 

Thing 32084, Sailfish 4.0 Rev. 720_

ReplicatorG 0040

0.4 Nozzle

ABS 3mm

I assume it is ReplicatorG 0040 - Sailfish not just straight rep40.  Is that correct Daniel?

I am starting up another xp machine and of course it has a bunch of updates to do before I can start installing Rep40-Sailfish & Sailfish 4.0 rev 720 on my thing 5094.

--
You received this message because you are subscribed to the Google Groups "Vintage Makerbot" group.

To unsubscribe from this group and stop receiving emails from it, send an email to vintage-makerbot+unsubscribe@googlegroups.com.

[Daniel Springwald]

Hi,

this is correct.

I´m using Windows 8.

BTW: PLA printing with acceleration works fine on my TOM. I just cant get the settings for ABS...

Best
Daniel

[David Slater]

Looks like I will continue to not know what I am talking about.  My computer would rather crash than switch to sailfish.

Hope someone else can help you quicker than I can sort this out.

Sorry,

later, slater

--
You received this message because you are subscribed to the Google Groups "Vintage Makerbot" group.

To unsubscribe from this group and stop receiving emails from it, send an email to vintage-makerb...@googlegroups.com.

[daves...@gmail.com]

Hey Daniel,
What version of repg are you using?  I  finally upgraded the firmware on my tom, hbp, stepstruder mk6 to sailfish 4.7 and am unable to connect to it with repg40r28.
I just downgraded the firmware to 4.0r730 as you are using in your July emails so am trying to find a repg that will connect to it now.
Thanks,
later, slater

[David Slater]

Hi Daniel,

I am using a vintage tom with heated build platform, mk6 stepstruder, makerbot ver 2.5 motherboard with atmega 2560, extruder v3.6 firmware 3.1, homebrew lcd interface, 3mm filament running, repg40r28sailfish software and sailfish 4.7 firmware on an XP laptop.

My first print was the calibration cube with 99% fill using the gcode they supplied.  Second print was the calibration cube with 10% fill and I generated the gcode.  pix attached

Had a lot of trouble uploading the firmware to the extruder and motherboard with rep40r28sailfish.  Was much easier with rep40r24sailfish for some reason.

Will try to replicate your issue with abs calibration cube after I get my setup a bit better tuned and figure out how to modify the start and end codes.

later, slater

[Dan Newman]

On 11/08/2014, 12:23 PM, David Slater wrote:
> Hi Daniel,
> I am using a vintage tom with heated build platform, mk6 stepstruder,
> makerbot ver 2.5 motherboard

Really a v2.5? Black? That makes it a very late model ToM. Most
of them except the last hundred or so came with v2.4 motherboards.
I actually have a v2.5 as a spare and when I asked about it, was
told they had to run off a few more as they wound down the ToM.

> My first print was the calibration cube with 99% fill using the gcode they
> supplied.

Should be 100% using non-hex fill. 99% shouldn't matter much, but as
long as we're discussing it, it should be 100%. There is, however, a
bug in SF-47 and -50 as regards 100% hex fill which may be why they
gave you one with 99% fill.

> Had a lot of trouble uploading the firmware to the extruder and motherboard
> with rep40r28sailfish. Was much easier with rep40r24sailfish for some
> reason.

We updated the version of avrdude in RepG at some point in the r2X line. Don't
recall when. HOWEVER, uploading EC firmware has *always* been a PITA.
I have the best luck, disconnecting the EC board from everything including
power when updating them via USB. One of the reasons I won't distribute
the couple of bug fixes I have for EC firmware: I don't want to deal
with all the difficulties people have updating EC firmware.

Dan

[David Slater]

On Mon, Aug 11, 2014 at 4:15 PM, Dan Newman <dan.n...@mtbaldy.us> wrote:

On 11/08/2014, 12:23 PM, David Slater wrote:

Hi Daniel,
I am using a vintage tom with heated build platform, mk6 stepstruder,
makerbot ver 2.5 motherboard

Really a v2.5?  Black?  That makes it a very late model ToM.  Most
of them except the last hundred or so came with v2.4 motherboards.
I actually have a v2.5 as a spare and when I asked about it, was
told they had to run off a few more as they wound down the ToM.

​

​Yep but I would have sworn the card was green..I guess it is really black though.  My son bought this thing 5094 for my birthday  July 20, 2011 for 1318.57  it has been a joy building modifying and understanding it.​

 

​

My first print was the calibration cube with 99% fill using the gcode they
supplied.

Should be 100% using non-hex fill.  99% shouldn't matter much, but as
long as we're discussing it, it should be 100%.  There is, however, a
bug in SF-47 and -50 as regards 100% hex fill which may be why they
gave you one with 99% fill.

​I am thankful for that. I will use non hex fill hearafter.  Hex does look nice though.​

 

Had a lot of trouble uploading the firmware to the extruder and motherboard
with rep40r28sailfish.  Was much easier with rep40r24sailfish for some
reason.

We updated the version of avrdude in RepG at some point in the r2X line.  Don't
recall when.  HOWEVER, uploading EC firmware has *always* been a PITA.
I have the best luck, disconnecting the EC board from everything including
power when updating them via USB.  

 

​I did disconnect the motherboard from the  EC board but not the power supply.. Will try power supply disconnect next time. Had less trouble with EC than motherboard.  I beleive I did update firmware in EC using 40r28 but I just could not find the right timing for 'reset-upload' until I switched to 40r24 then it was easy ​

 

One of the reasons I won't distribute
the couple of bug fixes I have for EC firmware: I don't want to deal
with all the difficulties people have updating EC firmware.

Dan

​Yes I understand.  I am retired field service engineer worked at power plants across the USA.​

 

​Nukes in particular were difficult.  There was a servo valve that controlled the main steam valves and would work for years if refurbished by the OEM,,,if you told them to what particular rebuild level to use,,, of course that often fell thru the crack..  I gave up and began telling them to pitch them and buy brand new instead.  Even the fossils had problems.  One plant in Texas had them refurbished by an airplane maintenance facility and claimed they never had any troubles.  I found several defective and overnight one blew the electrical connector right off with an internal pressure leak.  At first I thought somebody tramped on it.  Oh well. Now that I am retired all I need to fight is TOM!!

Thanks again for you interest and help.

later, slater

--
You received this message because you are subscribed to the Google Groups "Vintage Makerbot" group.

To unsubscribe from this group and stop receiving emails from it, send an email to vintage-makerbot+unsubscribe@googlegroups.com.

[Dan Newman]

Sure enough, a v2.5.

By the way, if you leave the USB plugged in to your computer but the ToM turned
off, you'll see some of the LEDs on the Gen 4 LCD interface occassionally blink
(if you wired those LEDs). That's because there's just enough power for the
ATmega 2560 to run a little before it gives up the ghost, stops, then reboots
itself and tries to run again. The USB 5V is only at 500 mA; all the peripherals
such as the stepper drivers and EC are trying to run themselves and the voltage
sags to around 1.2 - 2.5V. The ATmega will try to run. I've taken the ATX
PSU connector's "power good" pin and wired it to the Arduino card's RESET
line via a 1K resistor. That is, a 1K resistor connects the two. Additionally,
but this is less necessary, I've programmed the ATmega to have 4.3V brown out
detection (BOD).

If you don't do the above, then do not leave USB connected to the ToM. Otherwise,
you may get EEPROM doo-doo: the AVR starts to run at low power, finds itself in
random code sections, and ends up occassionally trashing the EEPROM. Known
problem with AVRs and has nothing to do with Sailfish. There's even an
App Note from Atmel on the issue.

Dan

[David Slater]

Thank you.  I think I saw that happening while messing around with the various configurations and am sure I killed bot power with usb connected.  Actually I was wondering what was running that was pulsing the debug led.  Never imagined it was showing me low power supply voltage.  I also thought the eeprom got trashed by something once but thought it was me.

Ha Ha I just put the bot together and moved it down to it's normal cellar location, wouldnt you know the bot is off and usb connected... I will implement the PSU power good pin to the Arduino reset via 1K.  Will have to research where this stuff is.

I have an Atmel AVR Starter Kit but have not done much with it.  I have been focusing on getting the bot running right before I delve into the inner workings very far.  Early in my career I repaired Westinghouse P50 computers which had core memory ,2K I think, 13 bit words, all transistor, teletype I/O.  Nand logic was used throughout and the nand gates were implemented with ?two? transistors.  The transistors along with bias resistors and diodes were sandwiched in between two pc boards so looked like logs sandwiched in between walls.  Cordwood Nands is what they were called.  Pins comming off the bottom edge much like a dip but bigger were a real bear to desolder.

So Thank You again for the heads up on by LCD interface issue.  I am on my way to fire up the soldering iron... ha ha ha

later, slater

Dan

[Dan Newman]

> I have an Atmel AVR Starter Kit but have not done much with it.

Don't play with fuse bits until you're comfortable doing AVR stuff.
Get a fuse bit wrong and things are bricked until you get your
hands on a high voltage AVR programmer.

> I have
> been focusing on getting the bot running right before I delve into the
> inner workings very far. Early in my career I repaired Westinghouse P50
> computers which had core memory ,2K I think, 13 bit words, all transistor,
> teletype I/O. Nand logic was used throughout and the nand gates were
> implemented with ?two? transistors. The transistors along with bias
> resistors and diodes were sandwiched in between two pc boards so looked
> like logs sandwiched in between walls. Cordwood Nands is what they were
> called. Pins comming off the bottom edge much like a dip but bigger were a
> real bear to desolder.

I still have some core memory boards about, but not from a Westinghouse.
I loved the fact that on some machines, you could IPL them, switch off
the power, crate them up, ship them across the country, uncrate, power
on, and everything was good to go with no need to reload.

Dan

[David Slater]

Ha Ha that is neat on the core memory..

It took about 3 hours to load the P50.  Started out hand loading a short program to load a binary loader. You toggled the address bit and the data bit with a ground probe on each of the 13 bit cards to do the hand load.  Then the data was read in off of paper tape.

Dan

[JustMe]

Hello;

I noticed your mention of the Westinghouse P50 computer.   I was a field service guy back in the mid 70's working on the P50.
I would like to get one for restoration.   It is an important part of computing history.   The full service manual is available online,
just search for "Westinghouse P50 computer".   I worked out of the Pittsburgh office then went international.

Sorry to hijack the thread.

Pete.

To unsubscribe from this group and stop receiving emails from it, send an email to vintage-makerb...@googlegroups.com.