Pwdr Building Questions about pwdr-model-0.1 HOT 50 OPEN

from pwdr-model-0.1.

from pwdr-model-0.1.

Hi Eddie!

Great! You are the third one building a Pwdr machine, somewhere in Germany, printer #2 is being built. Feedback is very useful for the design of Model 1.0! Let's answer your questions:

1. The default speeds worked for me, using gypsum powder. Just fiddle a bit untill it's working. It could also be that the new layer height is too small and the powder can't be deposited in such a thin layer.
2. True, the current firmware is limited to 2000 pixel bitmaps (but also jpg, png). When printing 3D files, you'll definitely need a SD shield as the 256kb on the Arduino Mega isn't sufficient. I'll post a wiring scheme and firmware for the SD shield soon.
3. Yeah, the C6602A are quite unstable if another curent is applied to them then the control current. I had major problems with this as the Arduino sends out random signals during booting. If the current is connected at that stage, nozzles can burn... A switch (preferably controlled by the Arduino) is the solution, if you've made one, please send me the code and a picture so I can include it in the codebase and documentation

Future Recommendations:

1. This would increase the complexity of the machine, but maybe reset buttons on one end would be useful so you don't have to move the gantry around by hand
2. OK, great! Could you give me the details on the powder supply you're using?

I really appreciate your feedback and I'm very proud other people are building my machine. Could you send me (via private e-mail) some pictures of your machine?

Alex

from pwdr-model-0.1.

I suggest you forget about Arduino and start using a real IDE, such as Atmel's excellent Atmel Studio. It's free, it's super powerful, and it isn't that hard to learn.
Arduino doesn't send out random signals during booting. It's completely wrong to assume so. Microcontroller, when powered on, has the pins floating. See Atmels PDFs. The state on these pins is not known and not guaranteed. Therefore you should either pull these up to VCC or down to the ground.
I suggest: use one pin to control a NPN transistor, which will then enable LM317. Put a 470R resistor between the pin and base of the transistor, and 4k7 between pin and ground. Collector and emitter go onto your 24V or whatever you had there... and that's it. Oh, that pin is called "digital" in Arduino world I think :-) When it's set (1), your print head gets power, when it's not (0), it doesn't.

from pwdr-model-0.1.

Thanks for you comments. The Atmel Studio is currently unavailable for Mac OSX and the Arduino fits just for this job and allows basic users to program microcontrollers.

On the 'random signal' problem, thanks for your explanation and more important, thanks for the solution. But as I'm a mechanical engineer, could you help me to create a electric scheme? This circuit will than be included in the design for Model 1.0.

from pwdr-model-0.1.

Thanks Alex. Does your C6602 cartridge spray in a dispersed manner after printing more than 1 minute? It seems that it is not because of the voltage applied to the cartridge but the high nozzle spray speed (15 times with an interval of 800us). I am making improvements slowly with depositing new layers of powder.

I am using this power supply: http://www.ebay.com/itm/24V-DC14-6A-350W-Regulated-Switching-Power-Supply-BRAND-NEW-Ship-From-US-/150899550132?pt=LH_DefaultDomain_0&hash=item23225067b4.

But I think like nothingman said, a 100 Watts power supply should suffice.

from pwdr-model-0.1.

@Pwdr : I wanted to modify schematic, but I didn't see any. Do you have a schematic of a board in something easy to digest, say Eagle (preferably, as it's free for boards up to 10cm x 10cm), Protel, etc?

from pwdr-model-0.1.

@SungwonJ: it most probably is a problem with voltage source. Firstly, it may be too high/low, secondly it really has to be filtered and low noise... cheap switched mode power sources are not good :/

from pwdr-model-0.1.

I didn't notice any problems with dispersed spraying. Sometimes nozzles go
defect though. The speeds are based on information and experiments of
Nicolas C Lewis and his InkShield.

The smaller supply is indeed sufficient, but I'm told that a switching
power supply isn't ideal for the stepper boards, maybe Nothinman knows
something about that?

The relay to disconnect the inkjet isn't needed if Nothinmans suggestion
is incorporated.

from pwdr-model-0.1.

I am not a hardware guy and I just take the power supply from my RepRap Mendel machine. @nothinman Could you please instruct? Something like a filtering electric scheme will be highly appreciated!

from pwdr-model-0.1.

I do not know the power consumption of the head and the recommended voltage. Don't know whether you get high ripples or not. This would all have to be measured. I got a head from Epson printer and will be probably playing with that instead... did anyone ever find any PDF documents for that HP print head? That would help a lot, as they all tell you what's required to drive the head. One could also measure voltage/current directly on the printer...

from pwdr-model-0.1.

I am using 18V to power the ULN2003 as well as the cartridge. No official document released by HP. But thingiverse.com/thing:8542 had done some experiment with C6602. It could be helpful.

from pwdr-model-0.1.

As Eddie mentioned, information about the HP head can be found in a Parallax manual [http://www.parallax.com/dl/docs/prod/robo/InkjetKitDocs-v1.0.pdf]. The HP head is intended as OEM cartridge, for other manufacturers to build a printer, so there must be some more extensive documentation than this [http://h10088.www1.hp.com/gap/download/HP_C6602A_US_Print_Rev_E.pdf]

What kind of Epson head will you be using?

BTW, there are electronic schemes in [https://github.com/Pwdr/Pwdr-Model-0.1/tree/master/Electronics], could you see whether you could incorporate you solution for the floating pins? There is no PCB yet, it's high on the feature request

from pwdr-model-0.1.

http://www.parallax.com/dl/docs/prod/robo/InkjetKitDocs-v1.0.pdf page 33 is showing 24V on the voltage regulator side and 22V on the print head side, not sure which one is correct, but it looks like the voltage should be increased...
I'm using a print head from Epson R800 photo printer. It's fast as hell, and nozzle count is very high. I think the printer does ~5700dpi in one direction...

I saw that "schematic", but it's something between a schematic and a wiring diagram ;-) So I guess there isn't a real schematic? ;)
Power source from that page 33 could be used (left-top corner), since the kit is released by Parallax I guess it is rather stable...

from pwdr-model-0.1.

I tested the voltage range it seems the print head is fully functional between 18V-24V.

from pwdr-model-0.1.

@nothinman Could you provide a sketched schematic for the layout you're proposing? Don't really get it... Thanks in advance!

from pwdr-model-0.1.

An Eagle schematic file is present now. Could you help with the scheme for the pin pull up, @nothinman ?

from pwdr-model-0.1.

Sorry man, enjoying my holidays now hence being non responsive. Will be back on the 6th!

from pwdr-model-0.1.

Did you have a good holiday @nothinman ? Maybe you've had some time to look at the scheme?

from pwdr-model-0.1.

Surprisingly it was pretty shit haha, but thanks for asking. I did have a look and can modify it this evening. Will get back to you shortly.
(sorry, I was still working on my extrusion based printer, couldn't figure out what the problem was and it took me ages...)

from pwdr-model-0.1.

Has anyone built one here in North America? I know there was parts availability issues and I wanted to see if that was resolved.

nothinman [email protected] wrote:

Surprisingly it was pretty shit haha, but thanks for asking. I did have
a look and can modify it this evening. Will get back to you shortly.
(sorry, I was still working on my extrusion based printer, couldn't
figure out what the problem was and it took me ages...)

---

Reply to this email directly or view it on GitHub:
#6 (comment)

from pwdr-model-0.1.

Hey @digitalgods, sorry for the late reply. I thought that @SungwonJ lives in North America and he didn't seemed to have problems with parts.

from pwdr-model-0.1.

I'm building one in Dallas Texas. I don't anticipate having too much of an issue finding the parts, but am just in the initial phases. Let me know here if you are interested in collaborating on this in the near future.

from pwdr-model-0.1.

@digitalw00t : I'll be seeking your assistance on getting this going after first quarter 2013. I put pwder printer on hold till I can get space setup for Fort Worth Makerspace with a spot for my kelms. My thoughts were to print onto dry powder ceramic slip that can be fired to harden piece.

from pwdr-model-0.1.

Well I am just startinv to gather the stuff to build it now.
On Dec 13, 2012 5:04 PM, "Robert Bradbury" [email protected] wrote:

@digitalw00t https://github.com/digitalw00t : I'll be seeking your
assistance on getting this going after first quarter 2013. I put pwder
printer on hold till I can get space setup for Fort Worth Makerspace with a
spot for my kelms. My thoughts were to print onto dry powder ceramic slip
that can be fired to harden piece.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/6#issuecomment-11357777.

from pwdr-model-0.1.

Hi Alex, can you specify how to connect Q2347A cable? Please!

from pwdr-model-0.1.

Ok, I found some schemes, tell me please, it will be the right one? And where to connect pins 10 and 11 from U6?

from pwdr-model-0.1.

The ULN drives 7 lines each, but the print head has only 12 nozzles, so I assume that 10 and 11 are simply not connected. There is no need to connect the input either, and those two pins could be used for something else.

Seeing pin 15 and 16 on the connector being connected on the ribbon, shorting 12 and COM, I assume that the connector or ribbon cable is flipped?

There is a tiny drawing at the bottom of this page:
http://pwdr.github.com/documentation.html

from pwdr-model-0.1.

Sorry for the late reply. I had the schemes updated already, but didn't commit them to Github.

ULN 10 and 11 are indeed not connected. The rest of the connections look similar, on first sight.

On Sunday, January 27, 2013 at 19:47 , MatthiasWM wrote:

The ULN drives 7 lines each, but the print head has only 12 nozzles, so I assume that 10 and 11 are simply not connected. There is no need to connect the input either, and those two pins could be used for something else.

—
Reply to this email directly or view it on GitHub (#6 (comment)).

from pwdr-model-0.1.

Alex, Matthias, thank you! I asking it because there is at list three different schemes in the archive, and it confused me, so I just wanted to make sure. Thanks!

from pwdr-model-0.1.

ULNs are Darlington arrays. Think of them as voltage level converters. They make sure the voltages between the printhead and microcontroller are at correct levels. Each of them has inputs and outputs, and these are physically the same, so it does not matter which port you connect it to, as long as your software knows it! This is probably why you're seeing many different schematics.

from pwdr-model-0.1.

I found a bug in Pwdr_GUI_V0_3.pde - in the tab "Load Model" filename "file://"+FileName; is wrong, it must be - "file://localhost/"+FileName;

from pwdr-model-0.1.

@Boreaz: 6 mails up. The drawing seems to have the flat wire coming from the cartridge flipped vertically. What seems to be a loop between the bottom two pins should probably be at the top, right?

from pwdr-model-0.1.

Yes, the looped contacts must be at the top, I paint this picture from two different schemes, when I'm tried to understand how to connect printhead and it completely wrong. This scheme https://f.cloud.github.com/assets/3283492/163144/5b784fe8-7801-11e2-8b3b-6e356af6352c.png should be right but I'm not shure with how to connect upper nozzles - from 1st to 4th or from 4th to 1st.

from pwdr-model-0.1.

I will try a few more changes to your plans when I solder my ink shield this week:

1: the ULNs fir when the output goes high, To avoid false highs while the port is an input, I will try pulldowns (R2.2k to GND) on all ULN pins. That should keep the cartridge safe.

2: I will not cross the lines for the nozzles. All this does is create chaotic layouts and chaotic wiring. The straight wiring is easier to solder and to verify. The nozzle assignment is easily fixed in software, since we can only fire a single nozzle at a time. Using a switch/case statement is extremely fast because it is translated into a jump table (3 instructions only).

Matthias

PS: I know that we have readers from all over the world here. Being German, I tend to write with no sugar coating. This is not meant as a criticism, I am not trying to be rude. I am very very thankful that Pwdr is OpenSource and I appreciate the great efforts. Fantastic work! And I am looking forward to a working machine soon, trying to contribute all my findings.

from pwdr-model-0.1.

1. there's no need for that. there's a resistor on a base of every transistor in an ULN2803.
2. schematic is chaotic indeed, it just hurts my eyes to looks at it :-)

from pwdr-model-0.1.

The Eagle file has been updated (with pdf). Indeed very chaotic, so please suggest something for the. It's either on this side of the ULN's chaotic, or on the other side, due to the usage of the (quick) PORTA and PORTC

from pwdr-model-0.1.

@Boreaz I ran into a similar problem when trying to neatly connect the FFC back to the ULN arrays. After a lot of searching, I found this board: http://www.newhavendisplay.com/nhdffc161-p-2479.html
I've added two holes and two more screws to the outside of the Y-carriage to carry this board. The FFC now terminates neatly on the side of the carriage assembly and is easily removable from the rest of the device should it ever short out (thankfully, though that hasn't happened yet).

from pwdr-model-0.1.

Hello will you answer my question is anyone here

from pwdr-model-0.1.

How to use sd card and do i need add something in the pwdr 0.2. I uploaded pwdr model0.2 and added sd file to sd card

from pwdr-model-0.1.

It didnt work can u guys help me

from pwdr-model-0.1.

from pwdr-model-0.1.

from pwdr-model-0.1.

from pwdr-model-0.1.

from pwdr-model-0.1.

Hi how i use processing what should i put (string,string). How to uplaod stl file to processing

from pwdr-model-0.1.

can you show me some example of how to use codes

from pwdr-model-0.1.

and processing 4.2 didnt work
some errors

from pwdr-model-0.1.

from pwdr-model-0.1.

from pwdr-model-0.1.