Replicator3

299 views
Skip to first unread message

Jose DeMille

unread,
Jan 18, 2013, 12:41:41 AM1/18/13
to make...@googlegroups.com

A Manual Feed Rate Over Ride giving the ability to change Feeds and Speeds of builds while the jobs are being constructed]\

Have the ability to restart a job from anywhere in the build process.

A Hand Dial for jogging the X-Y and Z axis, tied into a LCD Digital Read Out

Wingcommander whpthomas

unread,
Jan 18, 2013, 1:10:52 AM1/18/13
to make...@googlegroups.com
I recently received an email from Bre pointing to a survey - which I dutifully filled out - and on all the aspects of the Replicator 2 I was dissatisfied with I made sure to provide in detail a personal account of my experiences with the printer - anyhoo they has a section at the end to provide comments about future directions. I would encourage anyone who got the survey request to politely fill it out online - but in retrospect I wold have been good to have chatted a bit about future directions here first.

For me I feel they really need to deliver a reliable out of the box experience, first day, first week, first month... If it had done that I would have ben extremely satisfied with the product. That being said, I am satisfied with my Replicator 2 now that I have upgraded the extruder http://www.thingiverse.com/thing:42250 (couldn't help it - another shameless plug - trying to hit the top 10 most popular things on thingiverse :) .

For me the things I would like to see improved are:

Fix the extruder.

Improve the overall rigidity of the build platform, it seems as you go higher in your prints this may lead to vibration related print defects. Although it may also be that this is helpful for overhangs where the edges curl up as they cool. The flex allows the extruder nozzle to graze over them without breaking the part loose.

Being able to adjust temperature, extruder flow (100% +/- n) and slow the feed rate (100% - n) LIVE would also be great.

Not sure about restarting the build process anywhere on Z - I am assuming that if they fix the extruder problems this capability would become redundant.  

You can already jog but it is a bit too basic, it would be nice to have the menu option refined a bit so it accelerates if you hold the button down (top speed set in settings) and stops the minute you take your finger off and steps in smaller increments if you just momentarily toggle the button.

Jose DeMille

unread,
Jan 18, 2013, 1:35:46 AM1/18/13
to make...@googlegroups.com
Great points!  As far as the jogging goes... what I would like to see is a hand wheel with selective toggled graduations of .1", .01", .001", .0001" per increment of the dial, incorporate this with a DRO, then machine can be used as an inspection device, thus allowing the operator to be able to determine exactly where the build needs to be restarted.  Tie that into having the ability to start the NC code from anywhere in the program...  Now you have a fail prof production quality piece of machinery.

These features would give a shop floor manager the flexibility to stop lower priority jobs mid-build, switch build plates, then start or resume another build without having to scrap the time and materials.

These "features" have been standard controls on every NC/CNC machine tool that I have worked on since the early 1980"s. BTW We fed the "Gcode" to the machines with paper ticker tape back then, and they still had these "features" that the Rep2 is missing.

Grumpy Jose

Wingcommander whpthomas

unread,
Jan 18, 2013, 3:49:03 AM1/18/13
to make...@googlegroups.com
Great points!  As far as the jogging goes... what I would like to see is a hand wheel with selective toggled graduations of .1", .01", .001", .0001" per increment of the dial, incorporate this with a DRO, then machine can be used as an inspection device, thus allowing the operator to be able to determine exactly where the build needs to be restarted.  Tie that into having the ability to start the NC code from anywhere in the program...  Now you have a fail prof production quality piece of machinery.

Now I get you. Yes something like this would be useful. 

Andy

unread,
Jan 18, 2013, 4:25:21 AM1/18/13
to make...@googlegroups.com, make...@googlegroups.com
1.  Faster processor, more ram.  Lets get that s-curve acceleration going.  One of Steve Job's greatest insights was the power of software combined with good hardware.  There is so much amazing potential in these machines.  Giving them powerful brains allows for the firmware and features to blossom.  Pause at z-position is only the beginning!  The market advantage of having the printer vs sending prints off to shapeways is one MBI should expand and exploit.  Transition to using cheap pc or mobile hardware so there isn't any serious limit to what the firmware has at it's disposal in terms of resources.  This is the eventual end game no matter how you look at it.  

2. Filament diameter sensor wired up to the firmware to dynamically adjust extrusion rate.  As I said on a previous post, this what I see as the last "random" unaccounted for in the system.  Everything else at least has some engineering attempt to solve or minimize the problem.

3.  For dual extrusion, a servo mounted on the gantry with a wiper blade that spins, continually wiping the inactive extruder drool.
Or:
4.  For dual extrusion, raise the inactive extruder .2mm and block it with a brass strip to prevent drool.

5.  Solve printable/removable support.
It doesn't have to dissolve, it just has to detach effortlessly.  It doesn't even necessarily have to be filament.  What about a glue extruder that just creates an interface layer between two pieces of print material.  Maybe this glue has a lower transition temperature, so you can just use a hair dryer to detach the glue from the print?  People want support that doesn't affect the print.  It would also allow these printers to print much more complicated designs.

6.  Spend some serious quality time engineering the machine.  At $2,800, these machines need to be extremely reliable and should be expected to run w/ minimal problems.   The rep 2 and rep 2x are now for sale.  MBI has made the transition to big brand and big business.  The next generation of printers are going to bring even more first time operators. They will have even less of a willingness and patience to make these machines work.  Plus, the competition will have made an year's worth of progress, including the RepRap community.   Make the Makerbot brand known for quality, dependability.  They already have a well earned reputation of good customer service. 

-Andy

On Jan 18, 2013, at 12:49 AM, Wingcommander whpthomas <m...@henri.net> wrote:

Great points!  As far as the jogging goes... what I would like to see is a hand wheel with selective toggled graduations of .1", .01", .001", .0001" per increment of the dial, incorporate this with a DRO, then machine can be used as an inspection device, thus allowing the operator to be able to determine exactly where the build needs to be restarted.  Tie that into having the ability to start the NC code from anywhere in the program...  Now you have a fail prof production quality piece of machinery.

Now I get you. Yes something like this would be useful. 

--
 
 

Jetguy

unread,
Jan 18, 2013, 6:54:55 AM1/18/13
to MakerBot Operators
Well, here's the first tip on expectations that are highly unlikely to
happen for technical reasons.

Lets face it how the company started in the first place. They took the
Reprap project, put it into an easy to manufacture frame basd entirely
on the tools they had in front of them (lasercutter). They have always
pulled in open source and were open source until recently. The big
thing about all that is they heavily borrow others ideas and
specifically code. They truly haven't revolutionized anything other
than making an easy to produce machine that used open source software
(replicator-g), until their own released closed source Makerware.
Think about that, do you want the same guys to started Makerware to
start from scratch?

Dan has already said that in the current firmware, due to space and
other issues, the entire position reporting has been stripped out of
the code. Why that is siginificant is that in order to jump this to a
new processor, they likely are going to need to completely from the
ground up rewrite the entire firmware and a lot of the stuff they take
for granted needs overhauled. If you guys are hacked off about
Replicator2 problems, at least you aren't having firmware issues.
Imagine the same group of folks who slipped and threw in scripts for
the 2X into the Replicator2 firmware release are the same guys about
to rewrite the firmware from scratch-bad days are ahead.

For technical reasons alone, I don't expect anytime soon and probably
not in an R3 such processor intensive features.
And, let's be reaslistic about this, the very machines you are
describing with said features might cost more than the average house.
Sure a Haas mini mill has great features, but not exactly in the price
range most people ae looking for. And, the guys who build that machine
are an entirely different group of engineers and mindset. Also keep in
mind that I don't know the engineering staff at MBI today, but knowing
how the company works, they were hiring interns not fully qaulified
engineers (mightyboard anyone?). Makerbot simply doesn't have the same
investment in the engineering, production and programming/sofware side
of the house that the other CNC players do. Which is my gripe, the
price of the bot increases but the engineering investment is not going
up, evidence by re-use of the same parts over and over that failed
then and fail now (HBP connectors, warping Z arms, filament plunger
drives?)

Let's just make a list of what they did and what the community did:

Community developed:
Stepper extruder
Dual extrusion
1.75mm filament adoption (taken heavily from other bots there)
Acceleration firmware
Upgraded constant pressure filament drive

Makerbot developed:
The frame?
Makerware-well wait, a lot of that is derived from open source, then
they close the source right?

Exactly, this is the finest example of a company who gained most of
its IP from open source and then closed the source. They claim they
developed a lot in house but the truth is, they were inspired by what
they saw others create and in a closed source patented world, they
would be in court right now.

And, if Bre hadn't been such a @#$@#$2 to others in the beginning
about open source, then he wouldn't look like such an @#$@#$ now for
closing the source after taking and continuing to take open source and
put it into their products.

And sorry, I'm about to offend some people, but I feel that Sailfish
is an enabler in this continued trend of giving Makerbot IP and
Makerbot, in turn keeping closed source and not giving back to anyone
but themselves. If they claim to be that inovative, then they need to
hire the engineers and put forth the R&D money and effort and develop
this crap on thier own. This is MakerBots own undoing and at some
point, the very business model and how they inovate is going to fail
if the community stops giving them the latest trends or IP.

What I'm trying to say is that you have the bots and the technology in
your hands that most of the IP and design was created outside of
MakerBot. Sure, they make the hardware and provide support, but let's
not pretend that certain people in the community couldn't join forces
and design a closed source bot too, that would blow anything MakerBot
sells right out of the water. Right now, they back Makerbot because
that's the sharing mentality that was started but with the companies
new rules, not exactly enticing these same folks to stay Makerbot. A
shining example is Type A machines. It's one guy who started in a
Makerlab. But he designed what the T-O-M could have been 2 years ago.
IF he had backing and a production facility, he could easily follow
the exact same steps as MakerBot initially but remain open source and
get a following.
QU-BD has the same potential. So yes, there are companies building
better hardware, and if the current trend keeps up, the guys doing the
real grunt work outside of the company are going to back someone else.
> >http://www.thingiverse.com/thing:42250(couldn't help it - another

Eighty

unread,
Jan 18, 2013, 9:31:37 AM1/18/13
to make...@googlegroups.com

My wish list is pretty lengthy, but I have to be realistic at the same time.  While MBI markets this to the "professional designer", it's just a prototyping machine.  It's not a high-end production machine.  So with that said, I think the low hanging fruit is:

1.)  Better processor, allowing for more firmware features (such as Pause at ZPos).  Some gcode overrides would be spectacular - I like the idea of temp and/or speed overrides.
2.)  More solid gantry and build plate.  My Rep2 is capable of much faster speeds, but the ringing in the exterior finish forces me to go slower.
3.)  As Andrew suggested, we really need a fix for the filament variations.  Seems like a bot-mounted caliper at the pinch wheel would be fairly straightforward.  Then couple it with a firmware upgrade that automatically adjusts the feedrate.  I did mention a better processor, right?

Wingcommander whpthomas

unread,
Jan 18, 2013, 9:41:30 AM1/18/13
to make...@googlegroups.com
3.) As Andrew suggested, we really need a fix for the filament variations. Seems like a bot-mounted caliper at the pinch wheel would be fairly straightforward. Then couple it with a firmware upgrade that automatically adjusts the feedrate. I did mention a better processor, right?

The interesting thing about this is it would really fix a lot of the issues I have changing filament. I have to measure then rebuild files I want to print in different colours because the diameter is different - and there is a huge difference in surface finish between 1.69 and 1.75 if you get it wrong - its quite surprising. Would be awesome to just feed the filament in and have the pinch wheel calibrated with a digital index to adjust the flow based on the present filament diameter. All other things aside, this would be a big step up in usability from my perspective - but we are just dreaming at this stage.

Jetguy

unread,
Jan 18, 2013, 10:30:24 AM1/18/13
to MakerBot Operators
The issue here is not the wish list, you guys are coming up with great
stuff.
It that most if not all of them requires a new processor, which
requires a new board, whihc requires new firmware, which requires
matching software.

Most of the improvements have been made by MakerBot taking the code
and inserting it into their released versions. Yes, the firmware is
open source but the primary contribution is going the other way back
into MakerBot, not the other way.

So what I'm saying is, make an educated guess of performance based on
past experience of the firmwares from MakerBoot's internal programming
team. Functional, but not exactly ground breaking and ususally takes 1
or 2 releases to get anything working correctly.

The mightyboard is the most shining hardware example besides the
extruder plunger fiasco. Both had to go to a major makover or even a
complete re-design to reduce major failures.

So your saying you want the next bot to be improved with all new
hardware, new electronics, and new firmware/software.

I'll tell you this, don't buy the first batch.

Further, let's cut to the real chase. MakerBot is NOT about building a
better bot for a cheaper price. In fact, so much so they are highly
reluctant to put any new tech into the bot. Arguably, Replicator 2X is
just the stuff that should have been on a Replicator 2 and then all
that is basically from the Replicator 1. Tech investment is a big fat
zero for a bot sold at a higher price. Are they improving the quality?
We hope yes but won't know for 6-8 weeks! And the fundamental
principal is again, they knew how and had the means to build a better
bot at the release of the Replicator2. The sale of that bot was a
market test exercise to see how much they could cut down a bot and
people would still be dumb enough to buy it rather than waiting on the
2X. What people should have done is boycotted, forcing them to sell a
better product. Be honest, those who bought one, it's not a life
critical thing to have had that bot? Those making money with those
bots could have used the older version or something else that was
likely cheaper and more reliable at printing.

Which then goes back to my argument. They closed the source making it
harder for us to even make improvements. Why should we continue to
contribute any advancements to the current or future designs? Say for
example a method to measure filament diameter. We could develop an
external module and I think optical measurement is best and then
easily via a microprocessor feed that into a free input pin and a
modified firmware could take that variable and use it as described.
But oh wait, we don't have schematics or know whats free to even use.
Because they use 1280s we are likely out of space. If you have the
illusion Makerbot would look at such tech to incorporate, think again.
It's about economics for them. The module cost could be $100, but by
nature they charge $200, but to put that into production and support,
they charge $400. So $400 premium, but wait, we said the current
processor can't do it. So new board $500. But wait they always screw
up the first board so then the next version is $700 more. So for
aproximately $1,200 they might be able to put that into a production
bot. And while I'm making this up, I think you'll find the pricing
structure and specifically upgrades in the past supports the numbers
presented above. So one feature is $1200 price increase but that
doesn't account for normal, every year price increases because I
assure you, they couldn't release that next year.

So look at the comptitors appraach to the issue of filament diameter.
They did the reasearch and know what the market will buy and can
afford. They handle this problem not with tech, but with quality
control. They did the math and said to put the tech into the bot, plus
the risk of failures and support calls, plus loss of reputation over
failures, plus the actual cost of the tech was so high, it would push
the price range of the machine out of being competitive. So, they
teamed up with the filament provider, or do it in house, and only
garantee if you use their more expensive filament. They traded machine
price for cost per print on consumables. Makerbot is more likely to
follow suite. And so it goes with almost every other item on the list.

Eighty

unread,
Jan 18, 2013, 11:10:19 AM1/18/13
to make...@googlegroups.com

This is probably a post for a completely new thread, but I think we're being too hard on MakerBot.  Here's why I say that:

Starting with the Rep2, this is no longer a DIY setup.  We're not buying a box of parts to customize as we see fit.  We're buying a complete system (well, it's supposed to be complete anyways).  It's a different mindset, since Bot operators have traditionally been tinkerers.  But MBI is trying to sell the equivalent of an HP LaserJet.  You buy it, you plug it in, you do some installation, and then you print with it.  If you suddenly decide that you'd like more options, then you need a different machine.  HP doesnt' sell upgrade kits to turn a LaserJet into an all-in-one.  So I think it's a bit unfair to demand upgrade paths for older machines when a newer model comes out.
 
Now, if MBI sells a backwards-compatible extruder upgrade, I'll probably be one of the first to buy it.  But I can't see where I have the right to demand one.
 
I'm not trying to open a can of worms here.  It's just that I see so many posts expressing outright anger at the 2X improvements.  Companies have the right (almost an obligation) to improve their products.  That will ALWAYS mean that the older versions are suddenly inferior.  So now I have an inferior Rep2.  It still works, so I'm fine with it. 
 

Mark Cohen

unread,
Jan 18, 2013, 11:29:33 AM1/18/13
to make...@googlegroups.com
You missed the point. I bought a $2000 machine in August 2012. That
never worked right out of the box. People have been buying $2200
machines later that also do not work right out of the box. The company
has an obligation to provide the upgrades for free to make these
machines work as advertised. They have no right to sell me a machine
with a bad motherboard and then replace it with another bad
motherboard. They have no right to sell a machine with a warped
acrylic bed and replace it with another warped acrylic bed. They will
not in any way get my repeat business. This is from a person who has
had 9 Makerbots and has given that company thousands of dollars so
far. They need to get with the program and take ownership of their
problems and then provide the fixes. People should not have to say how
thankful they are that customer support is so great because they sent
me replacement parts right away, rather they should say I never have
to call customer support because their products are so good.
> --
>
>

HelpingHands

unread,
Jan 18, 2013, 11:58:36 AM1/18/13
to make...@googlegroups.com
Christ, who pissing in your tea?
I've been building reprap machines for years, and I'm still feeling a little slighted at your suggestion that MakerBot has not done anything for the community.
When you say that makerbot did not develop things like the steppstruder, dual extrusion, or push adoption of 1.75mm filament, I think you're just blowing hot air. Makerbot did not do these first,  but they also did not simply copy someone else's work and slap their logo on it. They COULD have, and other people DO, but they at least had the integrity to spend R&D hours on it (maybe not enough, but they did, and that's what counts).
ReplicatorG was written mostly by makerbot. You can check the git commit history if you doubt it. Makerware was not based on previously existing community projects, with the exception of simply pulling RepG's version of Skeinforge and including it. I will admit that their move to MakerWare was not smooth in the least (it still lacks any dualstrusion support!), and I feel they have rather left ToM users in the lurch, but I have not had any issues running it with my Rep1 and Rep2.

My point is, your statement that "most of the IP and design was created outside of MakerBot" is blatantly false. I dare you to try and defend it with facts. The firmware overhauls by the community have been some of the best things every to happen to makerbot, but there's not much in the way of obvious hardware "borrowing."

Message has been deleted

Jose DeMille

unread,
Jan 18, 2013, 12:26:03 PM1/18/13
to make...@googlegroups.com
Mark,

Well said!

Jose

Darrell jan

unread,
Jan 18, 2013, 12:47:59 PM1/18/13
to make...@googlegroups.com
Think about multiple product lines. Look at 3D systems, which has the 3D Touch. That model goes head to head against the Replicator 2 or 2XL. It has some additional features, but overall it might actually not be as good, and is way more expensive. 3DS also has the Cubify, aimed at people who want to make things but don't want to get anywhere near the guts of the machine. And of course 3D systems has even higher end models as well.

Makerbot is a relatively tiny company that's trying to carve a niche. Business-wise, it looks like it's doing pretty well. Some people here want them to head into a different market (professional-industrial), others want them to continue to do better in their current market(consumer). They have decided, for now at least, to leave behind their old market, hobbyists. Well, they could do all of them. It would take more than one new product. You don't want to try to sell the professional machine to the consumer, or vice-versa.

And, of course, it takes money.

Darrell

Jose DeMille

unread,
Jan 18, 2013, 1:18:02 PM1/18/13
to make...@googlegroups.com
Darrell,

The problem that most people have with MakerBot is that they are marketing the Replicator2 as a Prosumer Product.  I have been bashed and trashed for what I'm going to say but I feel that is THE major source of frustration for most of the complaints on this board.  They are marketing the product as being ready for prime time, works right out of the box.  I have two bots and they have both taken a tremendous amount of time to tweak and more QC issues than I care to even recall.  Yes the customer service is wonderful, but the point is: we shouldn't need to call customer support at all.

This maybe fodder for a new thread but I'll post the question here: How many of us have set up a machine and have had it run more than 1000 hours without having to involve MakerBot support?

Cymon

unread,
Jan 18, 2013, 1:19:43 PM1/18/13
to make...@googlegroups.com
Keeping it simple. Of course I want better, harder, faster, stronger, but if Makerbot is going to stay ahead the next step has to be a leap and here's what I think that needs to have:
  • Filament diameter measured as the filament comes in and flow rate adjusted on the fly. I've discovered that calculation is actually pretty trivial.
  • Slicing handled on the board so all you have to do is send it the STL and all the details are handled on-board.
  • Smarter slicing that prints well 100% of the time. This may limit certain options like 0 fill layers for springs or multiply, but in the end making this easier and faster to use is necessary to blowing the lid off this.
  • Cheaper, not more expensive. I don't care if it needs to be smaller as a consequence, if makerbot makes the $300 3D printer and can keep up with demand then kiss the challengers good bye.

Darrell jan

unread,
Jan 18, 2013, 1:31:47 PM1/18/13
to make...@googlegroups.com
Jose,

I don't disagree, but just trying to stay on topic. You're basically saying that for Replicator 3, the emphasis should be on QC as opposed to new features? Or maybe new features which results in better reliability.

Darrell

Jose DeMille

unread,
Jan 18, 2013, 4:05:34 PM1/18/13
to make...@googlegroups.com
Darrel,

My bad, I got off my own topic.  Opps....

John Watson

unread,
Jan 18, 2013, 4:15:58 PM1/18/13
to make...@googlegroups.com
What I find interesting is that makernot is looming around this place..no one from there seems to dare make a comment about the issues ..they must have a very closed policy about  interaction on this site
but I suppose thats what the support tickets are all about  ..I personally found the answers to be almost useless
I was offered to send my machine back or get a new extruder and I said whats the point its a bloody defective delivery system ..to which I got a reply ..oh so happy you upgraded to the mk8 for the rep 2 ..we hear good things about it ..if you need further support ..yadadadadadadad
sigh...

I think that before any rep 3 versions are  even attempted ...they should get the rep 2 people sorted and Happy ..as we are the  potential next buyers ..I for one wont be buying anything else from MBI
till I am satisfied they  sort out  all the rep2 owners ..I also think the rep2x should be totally boycotted untill the rep2 is more reliable..
ah well back to work  after my micro moan
Message has been deleted

Billy Zelsnack

unread,
Jan 18, 2013, 9:49:14 PM1/18/13
to make...@googlegroups.com
On Friday, January 18, 2013 9:30:24 AM UTC-6, Jetguy wrote:
The sale of that bot was a 
market test exercise to see how much they could cut down a bot and 
people would still be dumb enough to buy it rather than waiting on the 
2X.

They made the correct move with the 2. The mistake was not fully embracing PLA earlier in the product line.

Reply all
Reply to author
Forward
0 new messages