Filament Drives

[Amedee]

Yes, indeed -- the one proposed here is more something we can use to say: "I pushed 1m for this print, and it actually used 95cm, so I need to increase the feed by 5% or so". For a more in-depth analysis you need a much more accurate encoder (plus something to handle it -- not sure we have enough pins and processing power on the board to handle a fast encoder)

[johan]

... not sure we have enough pins and processing power on the board to handle a fast encoder)

And that's another reason why sticking with an 8-bit MCU is a bad idea

[Amedee]

Agree. We are at the limit of this ATmega2560, we must be very careful when we add something that could delay the planner loop.

[Meduza]

I could see a use for just taking a high resolution encoder and tappning into the step/dir signals and log data to a computer as a first step to analyze the problem and doing some testing before actually implementering compensation in firmware.

I also agree with Johan, we should really move to more powerful processors than the atmega 2650 and away from 8-bit towards 32-bit.

[Neotko]

The problem I see moving to 32bits it's all the pt100 (bed specially) umo/um2 uses.

I always keep an eye for a 3d board with pt100 but so far:

http://reprap.org/wiki/Alligator_Board
The creators website never had an update after the indiegogo campaing.

http://smoothieware.org/forum/t-1086570 ... 100-pt1000
Smoothieboard pt100 sounds doable but never actually explains how to do it or if they did it.

Replicape seems to be a really nice board, I almost bought one but cancelled in the end. The kickstarer worked just fine and Elias the creator said a few times that the expansion board called Reach might have 2x pt100 connections.

[reibuehl]

The other thing to watch out for when selecting a new board is in my opinion the 5th motor as a number of people are experimenting with dual extrusion or alternative feeder concepts. The Smoothieboard seems to be available as a 5 stepper driver version to start with while the Alligator board needs an additional board for it.

[Amedee]

e3D is selling PT100 amplifiers -- I have not looked in detail, but I guess you could then use PT100 with almost any board...

On a side note, there is a new version of the Smoothieboard in the make, with 5 motor drivers and with either a co-processor or a FPCGA for the stepgen. You can get a smother motion plan with smaller steps, but with software stepgen you hit easily the limit (I don't know how fast Marlin/Arduino can pulse, but with GRBL the limit is around 50KHz if my memory serves well)

[Neotko]

About pt100 amplifiers, I asked the Replicape guy and that was his answer

Creator Elias Bakken Octubre 22
@Sebasitan, that souds like a sweet setup! There are many people that have asked for pt100 support. I've had a look at the schematic for the amplifier, and from what I gather, the amp is set up with 10 times gain, so at high temperatures, the output will approach the input voltage. There is 1.8V , 3.3V, 5V and 12V available on the board, but the maximum input voltage on the analogue pins is 1.8V. The PT100 requires a minimum of 2.7V, so 3.3V is the lowest possible option which is also available. That means that a voltage divider will need to be used to limit the voltage. It's not a big deal, but it does mean an extra two component circuit between the PT100 and the input pin. So the answer is yes, it will work if you use a voltage divider.
PT100 support along with more steppers/extruders is the number one request people are making, so I've started dabbling with Reach for Rev B. It's not done yet, but it would solve both of these problems in an elegant way. There are 4 analog pins left on the BBB, so there would be room for both two more thermistors and two more thermocouples (PT100) inputs. However, having the main board perform optimally has to be the number one priority.

[johan]

Pt100 requires a minimum of 2.7V for what ?

[Neotko]

No idea XD! He was answering if the e3d pt100 module could be used on his board.

[Amedee]

The PT100 itself is just a variable resistor, so voltage does not matter (to some extends of course).
Limitation might come from the amplifier board.
Sanjay's design assumes input voltage is 5V and provides a voltage table (link)