Stratasys Genesis XS Retrofit/Upgrade

Hello everyone, my name's Tom and I've just been very lucky in getting a Stratasys 'Genesis XS' from the electrical waste area of where I work
I'm a bit of a tinkerer with Arduino, and am quite comfortable with programming them etc now, so this should help me with this new area of 3D printing!

Here's the printer.. its big and quite heavy!

I've seen a similar topic on this very forum discussing a Genesis XS, so to avoid flooding someone else's thread I thought I'd start my own.

The reason it was thrown out from work is because the printers' on board computer has a problem (error code 85?), the wafers of plastic that feed the hot end are very expensive now, AND someone sold all of the cassettes on eBay because apparently they're valuable.

it is for this reason I have decided to retrofit the 3D printer. my main goals are:

• Change the hot end to a filament type extruder
• fit in a new controller 'brain' for it, the 3Drag board
• try to integrate the existing PSU and stepper driver board if possible

With regard to the last bullet point, here's a bit of detail from what I've been looking at today:

Here is the stepper driver board. Ideally i'd like to reuse it, firstly to save a bit of money and learn something new, and because there is a nice 30V power supply built into the machine.



The board has four separate stepper drivers on them, a closeup of 'one driver' is here


There are a pair of LMD18200T H-bridges, the datasheet is Here

I've been tracing the board tracks as best as i can (it's an 8 layer board i think!), and I know that:
The PWM input (pin 5 of the LMD18200) goes to the NAND gate chip (IC9 as printed on the board),

The two PWM pins of each LMD18200 are not linked; they go to a separate gate on the NAND gate chip.

there are two outputs from each LMD18200T's, does anyone know how they are 'paired' to run the stepper? each pair runs a 6 wire stepper motor.


The direction pins lead back to the PIC, which i think is programmed to match each axis. they have labels on indicating which axis they belong to, and what i think is a software version (e.g. 4.03)



I would like to interface these with the 3Drag controller board instead of using pololu drivers etc.. mainly because these stepper drivers may be more suited to the 3 Amp beefy steppers built into the machine.

Ultimately I'd like to know if the step and direction pins from the 3Drag would be able to drive the pair of LMD18200's.. could anyone give me any pointers please?

Do you think this is a waste of time and I would get similar machine response from pololu drivers or another 'off the shelf' driver that would work with the 28-30V supply? I jumped the gun a bit and have already ordered the 3Drag board so either way i'm sticking with that!

Link to my Google+ page with more photos

All the best,
Tom

OK, after a bit of research i don't think it's possible to run the two LMD18200's from just one signal/direction signal from the 3Drag board... so I'll go down the Pololu route which will just plug into my board... laziness but it will get me up and running more quickly

my decision comes after reading this tutorial on stepperworld [www.stepperworld.com]

Limit switches / extruder next!

hello everyone! it's been a while since I updated this, but here's the latest progress: the axes, hotend and extruder are now working:

Short Youtube video.

I can't find conclusive proof that my 3Drag controller will be ok with 36 volts going into it, so i'm using the old 'AT' power supply that powered the old computer inside it. it seems to have more than enough power to run the head, axes and lights at the same time... the heated bed is rated at 36V so I'll use an SSR with the gold 36V supply to heat it up instead of burdening the 12V power supply which probably won't get it hot enough to begin with.



the stepper drivers used are the Blue Pololu drivers with a small heatsink attached. they didn't get warm during a long run which i was surprised with given how small the chip is, and how big my stepper motors are! The 3Drag board is fitted around the back of the unit (i hope these pictures are working for everyone else lol):





for the extruder I've got an ABS printed Gregs Wade extruder mounted to a J-head, and reused a NEMA 17 stepper motor from the original biscuit feeder.this was easy to fit on - there's a nice vertical plate where the old extruder used to sit; I basically extended this plate 29mm downwards because the new extruder nozzle assembly was 29mm shorter than the old one. simple!

Here's the old assembly and biscuit/wafer store.. this thing could hold a lot of plastic when fully loaded :O









There's a lovely long 12V fan that i've wired in to the 3Drag board, it controls nicely under PWM despite having its little own controller in it. it blows across the printed component near to the nozzle. I might make an air scoop to grab some air to cool the fins of the J-head.. not sure if this will be enough of a breeze but we'll see!





For the Endstops there are N.O (normally open) reed switches at the maximum of each axis. I'd have preferred to have NC switches but these are free, and i have no money

I'm using Repetier-host to move the axes for now, i couldn't find anything on getting ReplicatorG to work with the 3Drag board, but Repetier-host seems fine for me anyway

bits to do now are to calibrate the axes, get the endstops working properly. I need to have a look at what i've put in the firmware because the axis just glides right past without acting on it, haha!

I also need to find the thermistors that I bought to control the heated bed.. they've completely disappeared!

Cheers,
Tom

Awesome project. I'm not familiar with the 3drag board but it looks similar to a ramps style setup. What firmware are you using? If you are using marlin then you may need to see if your end stops are inverting and wired properly. Traditionally you want your endstops to be at your min. so that you home to 0,0,0.

Thanks for that, yeah I'm using Marlin, i think it's really good! I thought i was clever with what I do with my Arduino but wow i'm very impressed.

I'll double check that they're inverted, i'm quite sure they are but i will have a look tonight. I had another go at making them work last night to no avail They are definitely wired correctly, as in they are functioning at the plug end where it goes in to my 3Drag board.
I've got the switch wired to "S" (signal) and GND (ground!), with pullups enabled in the firmware. no joy though... it doesn't even try and home itself when you turn it on; i've been doing it manually for now and making sure my commands don't make it crash.

Another problem is that my Z axis is lifted up by the stepper motor to meet the extruder... the firmware is set to keep this axis active and it should hold the motor in position. However after about two minutes the motor switches off and it starts drifting back down due to gravity... this is very annoying because i have to re-home the axis all the time
I thought it might be the stepper driver tripping out due to overheating, but the heatsink very rarely gets past room temperature.

I had it running for about 4 hours last night until it was bed time, I might turn the acceleration down a bit because I don't like the way that the axes snap back and forth. other than that it's almost ready for its ABS due to arrive on wednesday. woo!



jzatopa Wrote:
-------------------------------------------------------
> Awesome project. I'm not familiar with the 3drag
> board but it looks similar to a ramps style setup.
> What firmware are you using? If you are using
> marlin then you may need to see if your end stops
> are inverting and wired properly. Traditionally
> you want your endstops to be at your min. so that
> you home to 0,0,0.

There is a setting in the config.h to keep your steppers on even when they are not in use. It sounds like that needs to be adjusted on your Z. When you home your axis to they move towards the side with the endstops? If so are you sure you don't have your endstops connected to the max instead of min plugs (rambo has this option, I'm not sure about your board)? I have helped someone convert a cube-X. It was quite a bit of work but it paid off in the end.

jzatopa Wrote:
-------------------------------------------------------
> There is a setting in the config.h to keep your
> steppers on even when they are not in use. It
> sounds like that needs to be adjusted on your Z.
> When you home your axis to they move towards the
> side with the endstops? If so are you sure you
> don't have your endstops connected to the max
> instead of min plugs (rambo has this option, I'm
> not sure about your board)? I have helped someone
> convert a cube-X. It was quite a bit of work but
> it paid off in the end.

Thanks for the suggestions:
Yes, the axes definitely move in the direction of the limit switches.
Yes, the setting for keeping the motors enabled is set to do so. However, I had a look in the 'configuration_adv' tab and on line 172 there is:
#DEFAULT_STEPPER_DEACTIVATE_TIME 60

I timed how long it takes for the z axis to start drooping and it was indeed 60 seconds, so I've now put a load of zeroes on the end of the 60, and it holds forever now (well, more like 6 million seconds or something).


With regard to the endstops not working, I measured the voltage on the pin that is supposed to be pulled up by #define ENDSTOPPULLUPS and it isn't being pulled up to any voltage whatsoever.

I've had a look at the board and traced the endstop pins back to the arduino. just for reference:
X endstop signal is PE5 (digital pin 3 as seen in the firmware etc)
Y endstop signal is PJ1 (digital pin 14 ....)
Z endstop signal is PD3 digital pin 18...)

with the 3Drag board there is only one endstop switch for each axis, i.e. a minimum only, or a maximum only.

I then had a look in the Pins.h and found that for the 3Drag board, the default is MIN endstops (understandably), and the MAX ones were disabled (-1 is disable).
I've now changed the pins.h to act as a MAX endstop with MIN disabled... and the configuration *should* now work... I'm not going back to the printer now until friday afternoon so I'll have a go then. It's a bit worrying how they didn't pull up despite temporarily having MIN endstops enabled as well
just to double check: to enable a pullup, is the syntax JUST

#define ENDSTOPPULLUP_XMAX 
  #define ENDSTOPPULLUP_YMAX 
  #define ENDSTOPPULLUP_ZMAX

and not

#define ENDSTOPPULLUP_XMAX true
#define ENDSTOPPULLUP_YMAX true
#define ENDSTOPPULLUP_ZMAX true

or something? from what i understand they are activated unless you comment out the line...?

I know it would have been easier to fit my own NC endstops at the minimum of each axis, but where's the fun in that?! Plus the factory limit switches have lovely mounts and are good quality reed switches so I'd like to keep them, and the tidy wiring loom.

some useful references for people to look at, if needed:

[www.atmel.com] atmega 2560 datasheet
[www.open-electronics.org] Link to the 3Drag board schematics (they only show a low resolution picture :@)
[spreadsheets.google.com] A link for Arduino pins - firmware numbers, which i found really helpful

Well, good news is that my updated arduino pin trick works perfectly! The unit now homes and parks properly I had a go at printing, but initially had problems getting the bed past 26C (its a 30volt bed which I had running on 12v) but I've sorted that by using the old 30v supply and a relay, and the temperature now goes right up. I'll put an SSR in there soon but this will do for now

I blocked the nozzle so I've got to clean that out before trying again
Tom

well i've finally got it printing!

initially had a problem with repetier-host / slicer making the coordinates all messed up after it had finished slicing and started to print the job. it would try and center the print around 0,0 and as a result, print a quarter of the object.. no matter where i placed it in the preview screen!

I've overcome this problem by not letting slic3r home the axes before it starts. i just comment out the G28 command and it works fine then.
if i manually send G28 it works correctly too... very strange but i'm happy!

here's my first ever print... a 20x20x10mm shape to chack that my axes are correct.

blurry zoomed in shot:



The plate needs levelling, which I thought is odd for a 'proper' machine, BUT i don't know its history





ta-daaaaaa





I've also tuned my heated bed and the nozzle, PID settings for my machine are:
hotend (a J-head, cant remember what version but i think mk V)
kp: 20.95
ki: 0.95
kd: 115.35


the heated bed, which takes far too long to heat up due to the 25% duty cycling being on no matter what settings i put in the firmware....! (checked this with a o'scope)

kp: 205.61
ki: 7.2
kd:1468.76

the bed won't go past 78 Celsius or so.. which is rubbish.

I've put some blue tape on the bed, which made the filament stick nicely


I also added a cooling fan to blow across the fins of the J-head after my first blockage it's just a PCI-slot cooling fan from a PC.

It now stays nice and cool and hasn't jammed yet, the keener eyed folk among us will notice the almost failure about 30% of the way through the layers.... I tightened the filament gripping bolts up and it seemed to rectify the problem. They're now up all the way; the springs don't seem to have much resistance!



Cheers,
Tom

Hi all,

Seeing as though I wasn't impressed with my heated bed warmup time, I decided to make a MOSFET switch, using a P30N06LE N-channel mosfet that I scavenged from the scrap circuit boards of the printer.

seeing as though my heatbed is 36V, and my 3Drag board is running on 12V I thought that a separate MOSFET was the way to go I've used a 7805bt voltage regulator to convert the 12V signal from the 3Drag's output MOSFET into 5v for the mosfet's gate.

(mosfet driving a mosfet you say?! isnt that stupid?? not in my mind when I want to make sure I don't blow my 3Drag board up!)

Here's a picture: Red pair and green pair carry the cathode current of the bed; single red + green twisted pair go to the 12V mosfet output of the 3Drag.




It now warms up very well indeed... so much that it blew the thermal fuse inside the heated bed at 100 degrees C... I think the printer was only ever designed for lower temperature plastics. I've just had a look on the heatbed manufacturers website and it should be ok for up to 260 Celsius so that'll do me! it has a silicon shielded nichrome foil heat pad.

I've also moved the thermistor from its old location of touching the side of the aluminium plate (not very good) to touching the center of the plate on the underside where the original one was.

The stratasys control system used a LM34 linear farenheight sensor. I simply desoldered this and put a good old 100K thermistor in its place:




I've now got to level the bed and then start printing some very urgent parts. Yes steve I haven't forgot about you

well the heated bed is now nice and level (used a piece of paper to feel the distance between the nozzle and the heated bed). adjustment on this is done by adjusting the tension on the two vertical belts.. there is no other adjustment as the machine's frame is so solidly built. each belt has a cap screw with locking nut to adjust the tension.

I'm having so much trouble getting the ABS to stick to the bed :@ :@ argh! I've cleaned it meticulously, used blue tape (posh masking tape in the UK from B&Q), tried lemon juice, ABS juice made with acetone... different bed temperatures between 90 and 130 celsius, reduced the feedrate down very slow almost everything!

I've just ordered some 200mm Kapton tape which ought to do it... hopefully.

The best results I had was when the print didn't stick and I just turned the feedrate up to the max just to watch it go really fast.. and it stuck!

..but then repetier host crashed with a 'user breakpoint' error :@

at this point i closed the garage and went home. lol

Tom

well I got it to actually print something last night! I printed A Snowflake on Thingiverse





My heated bed is now covered with 200mm wide kapton tape. This stuff is fiddly to fit when there's only one person to peel it off the roll, cut it AND get it into the printer/ aligned to the bed! I didn't get it quite right first time but this will do for now:



I noticed that the extruder was retracting too much and the filament wasn't coming out of the nozzle when there were lots of retractions close to each other... it would extrude about a quarter of a turn of the big wheel, then retract about half a turn... and not return to where it was.
There's a setting in slic3r to get it to 'un-retract' so I'll definitely have to use that!

for this print I just used a small screwdriver as a ratchet to stop the extruder wheel turning backwards. the stepper motor current is turned down low so it only buzzed a little bit. Holding the screwdriver was really boring after 50 minutes!

The Kapton tape worked very well for sticking my ABS filament to the bed. To remove the print I cooled it down with the inbuilt fan, and used a roll of solder as a sort of 'hockey puck' to strike the base of the part.


The snowflake isn't perfect but it's pretty good for a first attempt


My only gripe to sort out now is that Repetier-Host seems to crash during a print.. straight after the first bit of infill tries to print. I printed the snowflakes with 0% infill to avoid this problem.
I can't remember the exact error but a popup appears on the screen saying 'repetier host has reached a user-defined breakpoint'. any ideas??

All I can say is that I'm impressed. Keep up the good work. I'm looking forward to more updates.

I have just picked up a Genisys Xs 3D printer all most like new. It only had a few hours on it when the person got sick and the shop parked it covered up in a room. no one in the shop want to learn how to use it so it set. I got it home dusted it off and plugged it in. After it warmed up and I got into the menu I found some parts to print out. It printed some parts beautifully. I was think their should be no reason that I can not buy flat stock abs and make the abs plastic biscuits.
I was going to change every thing over to the lighter cheaper stuff but if someone out their that has used this printer at work and would use it as is would be great and I would sell it to that person.

Edited 3 time(s). Last edit at 03/11/2017 09:08PM by donelliott.