Is It Really A Deaf Extruder?

Well many hours have been spent trying to understand what's going on with the extruder controller and the mother board. The issue still eludes, but it does look as though the extruder is doing what its supposed to.

Here's where I have been looking: -

Having looked at the design it appears that the RS485 interface buffer is enabled and disabled by the on board computers such that bursts of data can be passed between devices. The Mother board seems to transmit a data burst possibly having disabled its receiver and the extruder appears to respond having its transmitter disabled initially to allow it to receive data. Presumably the mother board then listens for a response to allow remote devices to transmit. Research seems to indicate that its a half duplex transmission system.
 
So it looks like the mother board just isn't 'seeing' the data sent by the extruder. Understanding why the motherboard is not 'seeing' the data is the next step as its definitely reaching the pin on the integrated circuit.

The top trace is the mother boards TX2 signal trace on Pin 11 of the 'shield' connections, Pin 12 of the IC. The lower trace is the RX2 signal on Pin 10 of the shield connections., pin 11 of the IC.

 

The RS485 circuit looks like this.

Electronics - Yes - Extruder Deaf

After the present of some money, I have acquired some electronics for my McWire.

Electronics
IPrint3D Motherboard, extruder board, 3 stepper driver boards and 3 limit boards.

The electronics have been undergoing testing as well as 4 stepper motors and an extruder board. The software came to life most easily with ReplicatorG (version 20). A TTL-232R-5V was used to connect the P.C. to the v1.2 RepRap motherboard. Power was furnished by the P.C. power supply which will be used for the dedicated P.C allocated to the project. I am planning to acquire an old power supply at some point in the near future. A network cable connects the extruder board to the mother board.

The current roadblock to progress is the extruder driver which was intended to be used with a stepper motor. Communication with the extruder board seems to be the issue at the moment with no output from the board seemingly available. The board responds to commands well enough with the heater and fan lights responding as anticipated from the ReplicatorG manual control panel.

When the extruder stepper motor is tested with the Extruder Controller V2.2, once configured for a stepper extruder (see below). The unit just seems to overheat the H bridge stepper driver chips; whilst the stepper motor sits there and screams. Both of the driver chips are activated when a command is sent to the extruder as can be seen from the indicator lights on the PCB. Remembering that there seems to be no output response from the extruder board, when a stepper runs there is noise on the thermistor channels which had a 100K Ohm resistor fitted for testing. The unit reads 0 for temperature and returns packet errors to the ReplicatorG software. There is continuity between sending and receiving RS485 chips with data appearing on the lines. The ReplicatorG software doesn't seem to be able to locate the extruder. I upgraded and at least once managed to trash the extruder board software. Fortunately I managed to re-program the little beastie with an AVRISP MkII and the HEX files located in 'C:\Documents and Settings\username\.replicatorg' folder. You don't see these folders from 'My documents' on Windows XP, you have to use the file explorer.


Stepper Motors
The stepper motors are SY42STH47-1684B-1 models manufactured by Soyo. So as these are not the Nema 23 motors I had planned for I need to update my 3D model and see how the whole thing will hang together. I have also acquired some M8 threaded rod but don't have any suitable nuts to drive with yet and would need some more imaginative coupling between the motors and rod if I intent to use the M8 threaded bar.

 

Setting up ReplicatorG
ReplicatorG requires some configuration to set the limits and scaling for the axis drives. Here is the segment I added to the ReplicatorG machines.xml configuration file, no doubt it needs a little further refinement. Note that the configuration shown below is between the 'machines' tag


From the information shown in the status window the lack of communication with he extruder board can be seen. With the board in this state the ReplicatorG software can control the Extruder.

 So this is where much of the effort will now be focused to attempt to resolve the communication issue between the two devices (I think I have tried all the firmware variants) - These are a bit of a work up if you are new to it all. The reset / upload sequence is a bit hit and miss from my experience.

   
    

Alternates and equivalent costs

So as an alternate to not having a Reprap device and having no money. I decided to try looking at alternatives. One particular one that came to mind as I have some exposure to the company was a rapid prototyping service called Protomold . Out of curiosity I decided to feed their system with one of the simpler parts required for RepRap Mendel device to look at the commercial cost of producing a small quantity of parts.

Well if I have all the settings correct you can see the price for one part is easily more than a RepRap device. I also found the moulding information very interesting.


So on with saving for RepRap bits I think.

Still no money...

It has been a long time since I posted, principally as I still have no money for hardware.

Earlier this year I filled in an on-line survey for Rep-Rap and seem to remember that he promised to email results but I don't recall seeing anything.

Also been working on the CAD model of my McWire and its quite comprehensive now. I haven't managed springs yet though. The ends are giving me some problems. All the parts are modeled and assembled and I can even do animations in my CAD program now, so I may try to get one of my virtual McWire.

Came across a German distributor of Rep Rap hardware http://shop.grrf.de/ which looks very interesting but the high entry price to buy, say all the electronic parts, in one go is a barrier. I have been considering if there is a way I could provide parts at a lower inital entry cost and use it as a lever to help me get a set of parts.

Have been programming PIC parts and AVR parts in the mean time to become familier with the hardware that's used. I had a go with a micro power AVR model which plugs directly into a USB port. The ATTiny85-20SU-ND model was set up as a simple data logger to make entrys into any application as it emulates a USB keyboard.

On the PCB hardware related side I have updated on Eagle CAD PCB http://www.cadsoftusa.com/ and got a POV ray viewer to render the hardware I have been experimenting with.

Slight tweaking required...

A further interference seems to have popped out of the woodwork. I think that Replibot saw this one as well. The Y axis sprung bearing assembly might be too deep. In this instance the easiest option may be to pop a few washers under the support rails between the X stage and Y stage. Of course washers are cheap and the punching operation used to make them can leave a raised edge as well.  I had a variation of 0,10 mm in my small batch. Not a great error for one washer but if I add them up for a spacers I could easily end up 0,5mm out which would be about the height of an extrusion. Well at least I am prepared for it now, and can think about an avoidance strategy.

Watching with interest...

I found the Pipe recently that summarises updates in one place for a related set of blogs. Pretty cool although I enjoyed surfing somewhat.

I have been watching Replibot being constructed (with some envy I might say). But at least my efforts in modelling seem worth while as I had seen the issue with the interfering bolt heads. At first I thought it was just my model that needed some tweaking but it seems I can have some confidence in it now. My proposed approach at the moment is to acquire some countersunk bolts to hold the PTFE bearings in position. I hadn't intended to use self taping screws in this location from the outset because I was concerned over the forces applied to the Perspex. I imagined I also might want to replace the bearings quite frequently due to creep and wear of the PTFE bearing, so presumed a self tapping screw might degrade rapidly. This will also keep the bed height down and thus keep the build height and maximise tool clearances.

Santa got my youngest an NXT Lego kit. He needs a help with it so I have been assisting with the build, programming and watching it fall over. We have already burnt our way through a set of batteries after a couple of hours, so we are looking to sort out some rechargeable. I have formed the impression that it has a similar processor to the version 3 electronics, it also apparently may support RS485.... interesting. It might be an alternate if it hasn't already been done.

Found....

I re-visited the Nanotech web site looking for data on Nema 17 stepper motors and found CAD models of Nema 23 and 17 devices. They are STEP models so I have imported them and checked them against my models.

Saw the Tweet note on the Blog created by Replibot regarding the Generation 3 electronics being available. By the time I got to it they seem to be out of stock again, so no need to panic about expanding the countrys burgening debt burden spending money I don't have. Imagine, I might have used my credit card and added to the 175 billion or so. Still as the Scots say "many a mickle makes a muckle", it should help a bit ;)