Ramps 1.4 - dual extruder - four thermistors for runaway redundancy?

[TopJimmyCooks]

I will be installing an itty bitty double extruder soon. When I got the hexagons, I drilled a second recess in each heater block for a second redundant thermistor. I am trying to prepare for the future and make the machine safe to be used with low supervision levels. I knew there were only three connections for thermistors on Ramps. I decided I would figure it out later or perhaps I wouldn't have the ramps board permanently.

The question is, is there any way to monitor a fourth and fifth thermistor with the auxiliary pins? I am already using the LCD panel adapter pins and I also have the ramps dual fan extender available to use although I don't think I will need it since the hot end fan will always be running 100%.

[printbus]

You would have to connect to one of the ATMEGA pins that can be selected by the internal analog switch. A glance at the RAMPS schematic indicated both AUX-1 and AUX-2 have some pins that do that. You'd have to add a 4.7K pullup to the Arduino VCC, and the other thermistor inputs have a 10uF capacitor on them to provide some low frequency filtering. VCC and Ground are also available on both AUX-1 and AUX-2

The schematic for RAMPS 1.4 is at http://reprap.org/wiki/File:RAMPS1.4schematic.png

On the schematic, the ATMEGA pins with ADC support are A0 through A15. The schematic also shows the pullup and capacitor on the other three inputs in the Thermistors area.

EDIT: I answered the question from the RAMPS hardware perspective. I'm sure you'd have to figure out how to add the additional monitoring into the MARLIN firmware.

EDIT #2: In case someone is looking at the schematic as a way to learn more about RAMPS, it can be confusing to see pins for the stepper motor drivers connected to what I said can be analog input pins. The key is can be. All ATMEGA pins have multiple functions, and firmware flashed into the ATMEGA determines what available function a pin will be used for.

[TopJimmyCooks]

Thank you. I knew the arduino had the capability to assign pins to analog input so I thought it was doable. Now that I know it can be done, wiring it up is pretty easy. the dual thermistor for at least one extruder is already defined in marlin, not sure about for 2 extruders.

Dumb question - a 4.7k pullup is just a 4.7k ohm resistor?

[Roxy]

Dumb question - a 4.7k pullup is just a 4.7k ohm resistor?

It is just a 4.7k ohm resistor tied to the V+ (because it is pulling something up).

[printbus]

Do you know what kind of failure you're worried about? If the thermistor or wiring shorts out or opens, Marlin already detects an error state. Don't those already cause an abort? I'm not aware of thermistors to all of a sudden change how they read, so I don't know that is a likely failure.

One possibility is that the thermistor becomes partially detached and starts reading low. That's definitely a possibility if the thermistor is just taped onto the hot end. I had good luck with the muffler sealer being used as glue. That's an improvement over the tape. Tapping the aluminum block for use with a stud-mount thermistor is even better (gmay3 has done this; I'm using a different aluminum block on my hex hot end that was already tapped).

Then there's the possibility of a RAMPS failure where a heater circuit sticks on. The heater MOSFET locking up is probably the most likely failure that could lead to heater runaway. Firmware can't do anything unless some form of redundant power control is also added in. If you were really concerned, I'd have an independent circuit monitoring those new thermistors (and the bed as well?), with an independent power cutoff to the printer as your ultimate failsafe.

[Mjolinor]

I think I would go with a separate circuit with a PIC that just shut down everything if they read over or under some specific value with two separate thermistors.

I too have never heard of thermistors reading incorrectly outside tolerance range, only either open or short.

Pretty easy to do with a PIC and a relay in the PSU feed line and no need to bother with what Marlin can or can't do.

Marlin itself is very complex for such a small micro but with the safety off board in a small PIC you would have no bugs in the firmware for something so simple.

[Roxy]

I think I would go with a separate circuit with a PIC that just shut down everything if they read over or under some specific value with two separate thermistors.

I too have never heard of thermistors reading incorrectly outside tolerance range, only either open or short.

Pretty easy to do with a PIC and a relay in the PSU feed line and no need to bother with what Marlin can or can't do.

Marlin itself is very complex for such a small micro but with the safety off board in a small PIC you would have no bugs in the firmware for something so simple.

If you do this with a PIC... Maybe you should consider going a step further. It would be good to watch the stepper motors to make sure they are moving and not frozen in place. And of course, if you go that far, you could watch the limit switches and make sure they are not staying activated too.

That really would be a nice 'Fail Safe' circuit to have controlling the power to everything. And while you are at it, maybe provide a sensor to make sure the filament has not run out. I bet a lot of people would be interested in having something like that.

[AbuMaia]

One possibility is that the thermistor becomes partially detached and starts reading low. That's definitely a possibility if the thermistor is just taped onto the hot end. I had good luck with the muffler sealer being used as glue. That's an improvement over the tape. Tapping the aluminum block for use with a stud-mount thermistor is even better (gmay3 has done this; I'm using a different aluminum block on my hex hot end that was already tapped).

The E3D hotend I use has a screw to hold the thermistor in its hole. I like it much better than wrapping the head in Kapton tape to keep it in.