Hot end overtemp shutdown

Hi All

It has occurred that when the processor stalls for some reason the hot end is left on and not regulated. It has also happened when the thermal sensor came out of the block.

This is not acceptable in my view.

Even the humble toaster will have some kind of over-temperature cutout.

What do we think would be a simple way of providing this facility independent of the processor obviously. I do appreciate that there has been talk of implementing a processor watchdog to help with this kind of issue but I think a simple even mechanical device is also needed.

Rory

I did some research on thermal fuses, however the highest holding temperature I found was 200C. This is a little too low even for PLA printing.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

The first thing that springs to mind is a more precise way to secure the thermistor in the block. I secured mine by bending the leads around the block, but its not very satisfactory.

One option is to get some larger diameter heatshrink ( or something ) to act as a stop and put that on one lead so you can slide the thermistor in from one end to a known point. Then add clamps ( e.g screw terminals) somewhere to secure the other end. (EDIT: Both ends!) Maybe the bit of paxolin above the block could be extended to mount clamps on.

regards
Andy

Edited 1 time(s). Last edit at 01/08/2014 03:54AM by kwikius.

---

Ormerod #318
www.zoomworks.org - Free and Open Source Stuff

Hi,

on some 3D-printer people take exhaust pipe paste (something like "GUNGUM") to fix the heater cartridge and/ or the thermistor/ thermocoupler. Some pastes are of course heat resistive and also remain flexible.

Markus

---

XBee & electronics blog: [lookmanowire.blogspot.com]

Quote
dc42
I did some research on thermal fuses, however the highest holding temperature I found was 200C. This is a little too low even for PLA printing.

Excellent idea, the following are non-resetable but go a little higher to 240C trip temp (not max operating temp which some searches provide)
[www.sinolec.co.uk]
These are AC current rated and I haven't investigated there DC performance (often a lower rating due to arcing)

---

Ormerod #007 (shaken but not stirred!)

Quote
kwikius
The first thing that springs to mind is a more precise way to secure the thermistor in the block. I secured mine by bending the leads around the block, but its not very satisfactory.

One option is to get some larger diameter heatshrink ( or something ) to act as a stop and put that on one lead so you can slide the thermistor in from one end to a known point. Then add clamps ( e.g screw terminals) somewhere to secure the other end. (EDIT: Both ends!) Maybe the bit of paxolin above the block could be extended to mount clamps on.

regards
Andy

I found by thermistor a very tight fit in the block, but can see the heat shrink idea is good and I could also benefit from securing the connectors.

---

Ormerod #007 (shaken but not stirred!)

At some future time before I need to leave my printer unattended for 'the big print', I plan to design some sort of independent 'watch dog' unit.

It will monitor the extruder and base temperatures using separate sensors and perhaps supply voltage (useful health indicator) and I was thinking this will trip an RCD to ensure a total shutdown in event of a fault.
The other thoughts on this is to power the extruder fan independently to ensure rapid cooling.
Monitor the movements to detect a stall (when the big print comes un-stuck).

I haven't searched on this so there maybe hundreds of solutions out there, but having worked in the chemical industry this is a typical safety feature.

But that's for the future once I am happy with my build and software versions.

---

Ormerod #007 (shaken but not stirred!)

Hot End needs its own MCU up there . Then only 4 wires to main Duet. 12V ,GND , TX, RX

regards
Andy

---

Ormerod #318
www.zoomworks.org - Free and Open Source Stuff

Quote
Treth

Quote
dc42
I did some research on thermal fuses, however the highest holding temperature I found was 200C. This is a little too low even for PLA printing.

Excellent idea, the following are non-resetable but go a little higher to 240C trip temp (not max operating temp which some searches provide)
[www.sinolec.co.uk]
These are AC current rated and I haven't investigated there DC performance (often a lower rating due to arcing)

I already found several with trip temp of 240C, but holding temp for those was 200C. You need the holding temp to be at least as high as the temperature used for printing, and the trip temp to be low enough to cut out before everything melts. I wouldn't worry too much about AC vs DC: these are single-use devices so I can't see that it matters if there is arcing when the thing trips, given that the voltage is low enough that the arc will soon stop when the device opens.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks for all the contributions.

I think it would be fair to say no solution has emerged as a clear front runner yet. One technique which would help with one aspect would be to drive the heater through monostable so that it would on for a fixed period when driven by an edge from the micro. This would not help with the missing sensor scenario, however that could be dealt with in code, something along the lines of if I apply power for 2 minutes and no temperature response occurs then flag an error and cease heating.

Some kind of last ditch overtemp device would be very desirable, shame that they are too low temp, but what if were mounted a little away from the actual block, such as on the mounting to the heatsink?

Rory

Nice of you guys to run over this ground again! It is a perennial problem, and often discussed in the main forum; to make a good, foolproof, safety system. Some printers have two thermistors, and cut out if there is a big variation in temperature between them. Probably the proper way to do it is to graph the temperature, and have a 'normal' response curve, and the heater turns off if it shifts out of these bounds. So, say your heater normally runs at 30% duty cycle, to maintain 200C, but then the firmware sees the temperature dropping (or drops sharply), despite running at 100%, then it kind of knows the heater or thermistor has dropped out, and can stop quickly.

At this point, the firmware starts to become a bit more like a series of tables, like the programming of car ECUs... but that's another story.

Ian
RepRapPro tech support

Rory: I think that mounting an over-temperature sensor on the heatsink is an excellent idea, because that would protect against fan failure as well as extruder overheating. Just need to find out what the normal max heatsink temp is.

Ian: I think this would be much easier if the ADC were run on a schedule, either using a time-triggered architecture with a scheduler (of which there are many freely available, including one of my own), or on a timer interrupt. Most of my designs work that way. It means that the code doesn't have to wait while the ADC conversions take place (very useful on the atmega processors, although less relevant on the SAM3X which has a faster ADC), and that the ADC readings are taken at regular intervals, which makes trend monitoring much easier. This approach would also make the implementation of a modulated z-probe easier and give more reproducible results.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Ian

The majority of Ormerod customers on the forum seem to be new to 3d printing but, due to the RS source, engineers of some sort. We are not therefore aware of the discussions that have gone before. If this is such a well known problem then why has the Ormerod been launched with no such protection. It can be a sophisticated as you suggest, or it could be a deal cruder, but there needs to be something. A processor based system is only good so long as the processor is running so something else is needed for this eventuallity.

Rory

Hows about this. Apparently it can generate electricity, which even if not much, might be sufficient to power a low power MCU when it all goes pear shaped. Can also be used for cooling

See Wikipedia

regards
Andy

Edited 1 time(s). Last edit at 01/09/2014 03:55AM by kwikius.

---

Ormerod #318
www.zoomworks.org - Free and Open Source Stuff

Quote
kwikius
Hows about this. Apparently it can generate electricity, which even if not much, might be sufficient to power a low power MCU when it all goes pear shaped. Can also be used for cooling

See Wikipedia

regards
Andy

I have a few of those devices here some place, 2 I think on an old PC to cool it's CPU's and the other on a portable fridge, I was aware that you can use it to generate small amounts of power but I think you would require heat on one side and a very cold temperature on the other for it to create power.

Paul

---

RS Ormerod No 436

Ian,

Do you print parts round the clock at your facility?
If you do, do you have any safety feature to protect your printers against over-temperature?
Do you consider overnight printing safe?

Hi All

Just seen another thread about overheated Ormerod. My fear is that once the firmware is more sorted and the processor is no longer freezing quite regularly this will all get forgotten about.

My approach at work some years back was to use a second temperature controller module set to a higher temperature to knock off the power if the primary circuit becomes disabled for any reason. There are some quite cheap temperature controllers on ebay these days possibly as little as 10 pounds which I am sure could be used. As discussed before this could be mounted on the heatsink to protect against two scenarios of excessive head temp and loss of fan rotation. As the unit has a temperature display it would be easy to read the normal temperature and set a slightly higher value for shutdown.

I am not sure if it is possible to pause printing but if so we could disable the ATX 12V power but still have the Duet running on 5V standby ready to resume printing when the situation is resolved.

Rory

Ray I believe storage radiators contain a non resettable over temperature device but I personally do not favor such an approach.

Edited 2 time(s). Last edit at 01/11/2014 01:05PM by Rory166.

Hi

I have now ordered some temperature controllers from the far east. At less than 5 pounds each including the sensor and a temperature display they should prove useful. I am thinking to have one as an over temperature shutoff for the hotend with the sensor on the heatsink and perhaps use another to control the bed temperature independently.

Rory

@arnaud31: Yes, we do print overnight, unattended. However, this is a case of "do as I say, not do as I do"! Generally, we wouldn't recommend that printing is left completely unattended, particularly if you go off to bed in the same house, if you don't have a smoke alarm. While I haven't heard of anyone burning their house down with a reprap, there certainly have been some incidents of overheating, electrical shorts, and other problems. In the past, this has been due to limitations on the controllers and the hobbyist nature that reprap grew out of, but with the Duet board, we feel that we have a platform that can better perform the control over the machine. This is still some way off; we were talking about it the other day, discussing ways that we could manage 'events'. Unfortunately, there is often a crossover between anticipated behaviour and unexpected failure!

@Rory166: We will not forget about this. Safety is of paramount importance. Programming this processor is quite new to us, so we are running to catch up with firmware requests. So there is only basic temperature protection at the moment - basically, it's okay if the temperature is between -30 and +300C (not quite sure of the exact numbers). What would be useful would be to have a list of conditions, the expected response, and the action if the expected response is false. Or, a list of possible failures, and to work out how to discern those from normal printing. The watchdog timer (which, I believe, has not been enabled yet) could be used as a final off switch, when things go wrong.

For example:
1. User or gcode asks for a particular temperature. Is this sensibly within range? (ie 250, not a miss-keyed 2500!)
2. Can I turn the heater on? Check what the current temperature is. If it's in range, okay.
3. Turn on heater. Take temperature reading. Wait 10 seconds. Take temperature reading, compare to earlier reading. If it's higher, ok, else stop. (Because the temperature reading is unresponsive, or the thermistor has fallen out. You could do this every time the heater switches on, and just stop and restart it if the heater switches off, as it maintains the target temperature)

That's a very basic example, but I'm sure do-able. I suppose it is surprising that this sort of functionality hasn't been built in before, when I think about it. We won't forget about adding this sort of thing, we just have to make sure it doesn't conflict with what people expect the printer to do.

Ian
RepRapPro tech support

Wearing my safety-critical software hat, what is needed here is a risk analysis. In essence:

- In what ways might internal component failure or external influence cause the system to fail with potentially dangerous or damaging consequences?
- Considering the possible consequences of each of these failure modes, which ones must be avoided?
- How can those consequences be avoided or mitigated?

It is possible to design software-based protection systems in conjunction with fail-safe electronics to make systems safe, but often simple electromechanical systems are easier and cheaper to implement.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

The is a very simply way to make the heater safe: chose a power that will not exceed the temperature rating of the hot end when permanently on. I.e. let it max out at about 300C. That way the worst you get is a some smoking filament. It doesn't catch fire.

The downside is a longer heating cycle, but the warmup time of the extruder is not normally significant compared to print time, and is often shorter than the bed warm up. Also a slow warm up gives chance for everything to reach equilibrium and expand fully.

---

[www.hydraraptor.blogspot.com]

A totally different suggestion!

If the machine is working correctly it will be extruding (except when travelling which should only be for a short period of time).

Something like a roller lever microswitch touching the teeth on the extruder cog could be connected to a simple timer circuit.
The timer is reset each time a tooth triggers the microswitch.
No reset and after an (insert appropriate adjustable time here) few seconds the timer would cut the supply to the heater (could also be used to cut main power supply)

There would need to be an override to enable heating befor the print starts.

Daryl

(still on the flat part of the steep learning curve)

---

RS Ormerod #299