Author Topic: EEVblog #326 - Makerbot Replicator Teardown  (Read 42023 times)

0 Members and 1 Guest are viewing this topic.

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
EEVblog #326 - Makerbot Replicator Teardown
« on: July 31, 2012, 02:13:10 pm »


Dave.
 

vlf3

  • Guest
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #1 on: July 31, 2012, 02:29:43 pm »
Hi Dave, would be interesting to compare the overall manufacturing speeds, between Thing-O-Matic and this Replicator version up-grade, if any change has been incorporated ! is it worth same object production, side-by-side  ???
 

Offline ProBang

  • Contributor
  • Posts: 30
  • Country: de
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #2 on: July 31, 2012, 03:27:47 pm »

Hello.


At 12:20 it looks like the red wire (R1) is not fixed in the connector.
(I don't like this connectors at all...)
I hope, you have recognized it while the teardown.

Does the Replicator works well now?
Needed more adjustment of the heating-plate, as mentioned? Or still produces crap?

Greets,


Hartmut
If you think my english is bad...
- then should you read my french!
 

Offline scottwolf369

  • Newbie
  • Posts: 6
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #3 on: July 31, 2012, 06:17:29 pm »
I'm confused by the bodge on U7, the AQV252G high-current PhotoMOS.

From the schematic, it looks like they're using U7 as a safety cutoff to kill power to Extruder 2.
But with the way they've soldered together pins 4 and 5, power will always be applied to EX2-PWR regardless of the state of U7.

I wonder if the schematic is correct, or if they had an issue with the safety cutoff circuit.
 

Offline charlespax

  • Regular Contributor
  • *
  • Posts: 191
  • Country: us
    • Pax Instruments
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #4 on: July 31, 2012, 09:15:26 pm »
A few responses to items in the video.

During development we tried several fan configurations. Having the fan directly against the side panel with a laser cut grill was pretty noisy. Just air movement in the chamber is enough to accomplish what we wanted, so we didn't need ducting; stand-offs worked, so we checked that off the list and moved on.

We wanted to put the SD card in an easily accessible place rather than on the motherboard in the back. We did some testing, but we were not 100% certain we could operate with the SD card on the interface board without signal issues. Given the short development cycle the footprint was left on as a fall back in case it didn't work. In the end it did work and it was decided that the micro SD card foot print should stay to provide hackability if someone wanted SD card support without using the interface.

Also, there's a few *hidden* messages on the Thing-O-Matic and Replicator silk screens.
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #5 on: August 01, 2012, 12:05:11 am »
During development we tried several fan configurations. Having the fan directly against the side panel with a laser cut grill was pretty noisy. Just air movement in the chamber is enough to accomplish what we wanted, so we didn't need ducting; stand-offs worked, so we checked that off the list and moved on.

Yes, because you have a small, cheap noisy fan  :P
Hiding inside the box will always attenuate the noise!

Quote
We wanted to put the SD card in an easily accessible place rather than on the motherboard in the back. We did some testing, but we were not 100% certain we could operate with the SD card on the interface board without signal issues.

I get SD card read error messages a bit, perhaps you still have issues here?

Dave.
 

Offline mariush

  • Super Contributor
  • ***
  • Posts: 4982
  • Country: ro
  • .
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #6 on: August 01, 2012, 01:01:41 am »
charlespax : 

If just the air moving around was enough for cooling, why didn't you guys go with a 120-140 mm low speed fan?  I realize it would have been probably a 12v one but it's probably less than 1$ in parts to add a 24v to 12v dc converter if that was even needed (a proper quality fan would run as low as 5v and probably still move air around)

Why the poor wire management inside, was it so hard to find some double sided tape or something like that and fix the cables to wood board or route them in some way?

Did you guys have a say in component selection?  Why the so-so capacitors from Lelon? I see them fail often in power supplies and monitors and I guess your board isn't subjected to the same stresses and temperatures as the mentioned devices so they would be all right but still... why not go for a brand name with reputation like Nichicon or Rubycon or United Chemi-Con?  Costs ? The pcb assembly firm/manufacturer defaulting to that brand?

--

I agree with Dave on the so-so decision to mount the capacitors under the boards with the stepper driver and I also remarked the issue with the boards potentially being mounted the other way around. A simple mark on the silkscreen would have been nice...  it's not a flaw, as the makerbot isn't supposed to be built by end-users.

Not very fond of the speaker choice and the fact that's surface mount/ not removable.

Considering all those small pcb with drivers and the cables all over the board and so on, I'm wondering how well would the board pass EMI tests, radiations etc..
 

Offline FenderBender

  • Super Contributor
  • ***
  • Posts: 1115
  • Country: us
    • The Solid State Workshop
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #7 on: August 01, 2012, 02:44:09 am »
I agree. The Replicator has the potential to be an excellent product. It's just a bit unfinished and I don't agree with all of the design decisions. Makerbot must buy in fairly large bulk. I'm sure it's nothing like Dell buying motherboards from Foxconn but still enough where prices differences can't be too much. While Makerbot, I think, has the edge in the at-home 3D printing market, there are other companies who'd like to take that prize. Simple things. I'm not so sure how much the electrolytic capacitor issue really matters in the long run. Lelon isn't terrible, but a Nichicon cap must cost like $0.001 more per than the Lelon part in bulk. Hell, they're practically the same when I buy a mere 5 of them. At a volume of what 10,000 the difference in price can't possibly be significant. Again, atleast they didn't put complete crap in there, but it makes me think...if you put Lelon caps in there, what else is being cheaped out on? Well there's one thing: The fan. 40mm...The only place these should exist are in 1U server racks where they NEED low profile because of the density of the machines. The Replicator doesn't NEED low profile so this situation could have been easily rectified with a standard 120mm fan, which are inherently more reliable, move more air and are quieter because they operate at lower RPMs and are bigger. You can get a brushless computer fan for not much more than a 40mm junk fan. Those usually consume at most 6W. 0.5A maaaybe. 24->12V would be like $0.50 in parts.

When you're in a new market like this, 3D printing, there's no room for error. If you want to leave the competition in the dust, do it right.

 

Online mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13694
  • Country: gb
    • Mike's Electric Stuff
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #8 on: August 01, 2012, 07:15:37 am »
Quote
Why the poor wire management inside, was it so hard to find some double sided tape or something like that and fix the cables to wood board or route them in some way?
DS tape will not stick to wood long-term if there is any static pull on it.

Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #9 on: August 01, 2012, 07:18:34 am »
I agree. The Replicator has the potential to be an excellent product. It's just a bit unfinished and I don't agree with all of the design decisions. Makerbot must buy in fairly large bulk. I'm sure it's nothing like Dell buying motherboards from Foxconn but still enough where prices differences can't be too much.

Not much bulk buying power here for say a one-off part per board, they have only shipped 10,000 units over about 3 years.
Sounds like a lot, but it's not in the scheme of electronics bulk buying.

Dave.
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37661
  • Country: au
    • EEVblog
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #10 on: August 01, 2012, 07:24:12 am »
Lelon isn't terrible

No, they are one of the biggest cap makers in Taiwan (and top 5 in the world IIRC)
The BadCaps forum has a lot to answer for in some respects, as the infamous bad capacitor incident happened over 10 years ago now and many of the major players were inadvertently caught up in it, including Lelon. So the dodgy cap label sticks to them and others, and is likely all but invalid in today's market.

Dave.
« Last Edit: August 01, 2012, 07:25:59 am by EEVblog »
 

Offline sciguy14

  • Newbie
  • Posts: 2
  • Country: us
    • JeremyBlum.com
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #11 on: August 01, 2012, 05:43:45 pm »
Our BotStep17 motor drivers were designed to maintain compatibility with the popular Pololu motor driver (http://www.pololu.com/catalog/product/1182).  That's why the capacitor is off the board -we didn't want to increase the size or change the form factor to ensure backwards compatibility.  Our main change to the board was adding an external VREF pin and removing the on-board trimpot, so that we could use a digipot controlled by the MCU to dynamically change the motor current.

With regards to motor diodes that Dave mentioned, we decided to use just the parasitic diodes after finding that putting the additional diodes in (while it does prevent the LCD from illuminating on motor movement) could, in some circumstances, cause damage to the BotSteps if the gantry was jerked forcefully.

As for the buzzer - It's open source firmware.  You can easily disable the noises in the firmware if you want.  But, I think they're cute.
« Last Edit: August 01, 2012, 05:48:54 pm by sciguy14 »
 

Offline Neilm

  • Super Contributor
  • ***
  • Posts: 1545
  • Country: gb
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #12 on: August 03, 2012, 04:38:30 pm »
Dave,

The trouble with using datasheets to learn how to layout circuits is that sometimes the thinking is old in regards EMC compliance. I have seen some manufacturers who assume that star point grounding is the way to go. This is not always the case (and in my experience I wouldn't use it except for a very simple board).  Indeed, some times what they are recommending is now specifically recommended against by EMC experts.

Neil
Two things are infinite: the universe and human stupidity; and I'm not sure about the the universe. - Albert Einstein
Tesla referral code https://ts.la/neil53539
 

Offline Rigby

  • Super Contributor
  • ***
  • Posts: 1476
  • Country: us
  • Learning, very new at this. Righteous Asshole, too
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #13 on: November 12, 2013, 07:31:16 pm »
Resurrecting an aged thread here, but I just watched this and opened my employer's MakerBot Replicator 2 - the electronics has changed a great deal.  It's still based on the Arduino Mega, but it looks like it's been redesigned.



I removed the Y axis BotStep board so you can see what's underneath.
 

Offline techydude

  • Supporter
  • ****
  • Posts: 45
  • Country: us
  • E.E. from .au in SF.
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #14 on: November 22, 2013, 12:01:28 am »
When I recently had a quick look over the schematic for the MightyBoard I noticed something odd - the RGB LED driver circuit has a PCA9533 generating PWM, whose outputs are going through a 10:1 voltage divider, then through a *schmit-trigger* inverter to drive the FETs.   :wtf:

Why bother with dividers & a schmigger option?

I tweeted Jeremy Blum about it, but his PM reply suggested it was something he shouldn't discuss publicly.  Mike Senese favourited my tweet immediately (I guess he has a full-time search on Makerbot), but also wouldn't be drawn on any further comment.

Have the Illuminati infiltrated the OSWH scene?!?  :-DD
 

Offline nitro2k01

  • Frequent Contributor
  • **
  • Posts: 843
  • Country: 00
Re: EEVblog #326 - Makerbot Replicator Teardown
« Reply #15 on: November 23, 2013, 02:21:51 am »
When I recently had a quick look over the schematic for the MightyBoard I noticed something odd - the RGB LED driver circuit has a PCA9533 generating PWM, whose outputs are going through a 10:1 voltage divider, then through a *schmit-trigger* inverter to drive the FETs.   :wtf:
This looks like several people or ideas went past each other. Those are pullups, which will in turn make the inverter show a low output and thus open circuit by default. In a different circuit I would expect those on the FET gate, though pulldowns in the case of a regular enhancement mode FET. Looking at this schematic I'm getting the feeling that someone read a guide on noise immunity and didn't actually understand squat of what they were reading.
« Last Edit: November 23, 2013, 04:26:31 am by nitro2k01 »
Whoa! How the hell did Dave know that Bob is my uncle? Amazing!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf