Watson Electronics

From RepRap
Jump to: navigation, search

This page is a development stub. Please enhance this page by adding information, cad files, nice big images, and well structured data!

Page is under construction and is unstable
Crystal Clear action run.png
Watson Electronics

Release status: experimental


CAD Models
External Link

This page is intended as an initial review of the current state of the art in reprap electronics. The ultimate goal is to look for areas where the quality of the electronics can be improved, the features expanded and the cost maintained or reduced. This page will evolve into a new electronics design

Stepper Drivers

X axis, Y axis, Z axis, E1 axis, E2 axis will cover most situations. An upgrade of the stepper drivers to 1/16 stepping might be implemented. Also, it would be nice to programatically control max stepper current rather than messing with single turn potentiometers. It seems that a lot of folks don't have the tools or knowledge to set the stepper current using a DVM and a potentiometer which can lead to frustration with under/over driving the steppers.

Considering using Microchip MCP4802 or MCP4902 DACs for setting the stepper current programatically, the 4802 has an internal voltage reference while the 4902 has an external reference. The external reference may be a little cheaper since we need 3 of the DACs.

We are leaning toward the Allegro A4983 or A4988 (this is what the Pololu uses). They are a bit cheaper than the TI equivalents with 1/16th stepping at the loss of 500mA of maximum current output. Update Using A4988

Temperature Measurement


Sticking with thermistors is probably the lowest cost solution. Though the range of accurate measurement is small thermistors are still pretty effective. There will be need for improved A/D and signal conditioning. Reading accuracy and repeatability should be close to 0.5 degrees C or better in the operating temperature range. Gen6 electronics seem to vary more than this.



Silicon Based

A silicon based detector should be fine for the heated bed as it would be unusual to need the bed to reach over 100C (most silicon integrated temp detectors can handle up to 125C or 150C)

Fan Controller

A fan could be a useful addition, especially during high speed prints with a lot of overhangs or bridges. Most hot ends tend to work better when the cold end is kept under airflow. PWM control or simple on/off control may be implemented. PWM control will likely be added as it is a pretty insignificant cost at low power levels. Support of 12V fans at a few hundred mA would probably make sense. 2 fan controller outputs will be planned. One for the print head and one for general cooling.

End Stops


Optical limits are simple and easy to use. We will stick with this. 1 per axis (don't need them on both sides) for a total of 3 limits.

Limit Switch

Heater Control

Bang Bang

Bang Bang control is a method of on / off power control. The on and off time are controlled proportionally to maintain the average power delivered and needed to maintain a certain temperature. The reality is that this doesn't differ from PWM control except for the fact that bang bang control is typically implemented with on off cycles with a period of hundreds of milliseconds, or many seconds.


PWM control is an efficient method of proportional control which is much more power effecient than linear control. The disadvantages are that it can be audible at low frequencies (which is dwarfed by the audibility of the PWM stepper motors in most reprap designs) and at high frequencies switching losses can become significant. That said PWM control can easily be 90% efficient. One of the problems seen in several of the heater controller designs has been use of arbitrarily high frequencies. Frequencies above a few hundred Hz or a few KHz, although potentially audible, can not be exceeded without proper MOSFET drivers to pump the MOSFET gate capacitance. The higher the frequency, the more time the MOSFET spends in it's switching transition state which causes excess power loss. This is especially problematic when using microprocessor I/O pins to drive a MOSFET gate as they can not deliver enough current quickly enough to charge the gate and switch the MOSFET rapidly.

Future Direction

It would seem that the continued use of PWM control, with sufficiently sized MOSFET along with a low frequency (100Hz with capacity to utilize up to 1KHZ is a good target for heater circuits.

Heaters for 2 extrudes and 1 heated bed would seem to cover the most ambitious of reprap users.

Hot end targets will be for 5A delivery for extrusion heads allowing for as little as 2.4 ohms and a maximum power of 60W. This is likely overkill and will be rolled back as competing interests (such as cost, board real estate etc. come into play)

Heated bed targets will be for 15A 20A at 12V delivery for a maximum power of 180W 240W. This power level might be overkill but there is a desire for a larger print area with the Watson printer. Separate power input for the heated bed may be considered.

Computer Interface

It may be possible to create a simple modular communication interface that would allow for a plugin "option" for users to select which computer interface they preferred whether it be USB, RS232, or Bluetooth. This would likely increase overall system cost slightly (an additional daughter board would be required along with connectors to implement each of the 3 options.


This will likely be the primary interface. FTDI offers a nice usb to serial bridge which is used by other designs FT232R. We are considering the MCP2200 as a lower cost but fully functional alternative. The FT231XS-R might also be another lower cost option


given the ubiquity of USB and ease of creating virtual comm ports there isn't much value in offering RS232


Bluetooth might be fun to offer but isn't critical for most users. Serial integrated bluetooth modules are not cheap either. This would likely be something for expansion unless the community really demanded the features

RS485 Interface

Scrap. I don't think enough people use this to be worth having in the hardware.

External Storage


Adding SD is the lowest cost and easiest approach, the increased board cost is minimal and the cost of a connector is only a few $.


microSD can be handled with use of an adapter. There might not be any reason to consider a dedicated microSD slot unless board space were an issue.

USB Mass Storage (Flash drive or Hard Drive)

User Interface

Character LCD

A character display would likely be included as an addon. It would be valuable to have some feedback during a print, especially if external storage is used and standalone operation is desired. This may be implemented as an addon module to the base electronics as a plugin type character display is pretty easy to obtain and implement.


A 4 direction mini joystick might be an option. Also considering PCB with capacitive touch points. This would be cool but not necessarily "better"

Buzzer, Beeper, Speaker

Some type of audible notification would be nice. One consideration is alarms if something goes drastically wrong that the printer can detect. Also in conjunction with a user interface some feedback is nice (especially when using not haptic inputs like capacitive touch sensors)


Thermal switches, redundant temperature sensing or some type of overheat protection is advisable. Consideration of max torque and current limits will also be important to avoid physical damage to the system, or injury. In general you should only apply enough torque to move the axes with some headroom.

Loss of Filament

This isn't strictly safety, but it would be darn nice to know when you run out of filament instead of printing nothing for an extra hour.