Print 3D

Edit:21 janv. 2018, Cre:21 janv. 2018

Silicon Heat Pad Incident

Using heaters on the mains

Silicon heating pads for beds are more and more used in the domestic 3D printer world.
There are a few reasons for that.
When you use low voltage heat bed, you may have some problems:

  • Insufficiently rated MosFET on Cheap Chinese board have caused fire
  • Insufficiently rated Terminals and copper tracks on Cheap Chinese board have caused fire
  • Absence of ferrules or wire terminals caused false contacts then fires
  • Insufficient size on wires also cause overheating and fires
  • Rising size of printers needs more power and the inability of cheap boards to handle 24V drive to quite high current in 12V, aggravating all factors above

All these problems could be solved by proper engineering and good manufacturing, but the recent years have seen a ‘race to the bottom’ in term of quality from equipment coming from China and the search of the lowest price and insufficient engineering knowledge caused many troubles. The open sourcing of equipment design and cloning mentality of Chinese manufacturers drive people to manufacture stuff they did not properly understand and this is a major safety concern.

From an engineering perspective, there is not much sense to transform current for the simple purpose of heating. We all use appliances with direct heating with the mains voltage (Irons, ovens, cookers, toasters, etc.) So using the mains for heating is nothing new.
However, home 3D printers are a new equipment with no traditional design practice and if caution in design is not applied, it can end up in quite dangerous machines. The fact that this area was driven by do-it-yourselfers without the mindset required for safety handling don’t help either.

Incident with a heating silicon pad on the mains (230V)

I here report a very serious incident with a silicon heat pad, quite unexpected which occurred me two years ago. I already published photos, but never detailed context. A silicon pad glued on aluminium plate 3mm thick inflated and started to burn the printer wood box. It was a test, so I was in the same room but for some reasons, it takes time to me to notice and I stopped the machine and that ended the incident? There was no flame and an enclosed machine may not burn very quickly due to oxygen depletion.

What was this heater for ?

I designed and built an enclosed Delta printer? in plywood.
It can be interesting for an enclosed printer to have a controlled temperature inside the chamber, so I wanted to design a chamber heating system. It does exist industrial chamber heaters, but they are costly and they all rely on a fan for circulation. I thought it was wise to have a purely static heater not relying on any fan. My D-Box being quite large, I have sufficient room for such a static radiator, set on a side wall of the printer, so I decided to give it a go. I bought a PAD of 500W and installed it on an aluminium plate 3mm thick of 200×300. This makes a power density of 0.75W/cm2 if you consider one face, but the two faces were exposed and I expected good cooling on both face due to the vertical position and rising airflow due to the heat. Thats makes 0.375W/cm2 if you consider the two faces. This was a radiator, not a bed, so with good cooling. I expected a temperature which cannot exceed 130/140°C. The silicon pad was glued on the aluminium and set on the face which was on wall side at two centimeters of the wall, a gap sufficient for a good airflow.

Test

I decided to makes a test at full power to see what I could expect in term of chamber temperature rise.
Not quite confident in the marking of the bed, I did checked the bed resistance, which was in agreement with the indicated power.
The thermistors were not wired (I should have) and I was willing to check the plate temperature with a good quality infrared thermometer. I just forgot to coat the aluminium of set tape on it to have a proper emissivity coefficient and my Infrared measures on brilliant aluminium plate proofed immediately wrong. I was wrong while not properly checking the temperature, but a test at full power while checking the machine is the most basic safety test you shall run. And, sort of, this test was very successful, as it proofed this setup to be very dangerous. It is much better to have troubles in a supervised environment than when you are a few rooms away or inattentive.

What occurred.

The silicon pad inflated. I did not noticed as it was hidden behind the plate. As my surface measurement was wrong, I closed the printer and go checking the chamber temperature, which was not rising very quickly. As the pad inflated, there was no longer any thermal mass to heat and the silicon temperature rise to high values. As the pad inflated, it closes the gap behind and was no longer cooled. The wood being a relatively good insulator, temperature raised to values sufficient to burn the wood (without flame). My printer was relatively tight and is equipped with an odor filtration system, which delayed me to notice the smell of burned wood. I was at less than 3 meters from the printer, but closely looking the chamber temperature which somewhat distracted me, then the smell became intense and some smoke came from the printer. I stopped the printer with emergency button.

Why did it occur ?

I have absolutely no idea why the pad inflated, it might have been some chemical becoming gaseous, or internal heat wires shorting, or a local ‘hot spot’ which destroyed the silicon and creates gases, I do not know and haven’t expertised the stuff. What can I expect in term of responsibility for an equipment I bought directly in China on ebay. Nothing. So there was not point to pay big money for an expertise. And The result were positive, I was now knowing what NOT to do.

My mistakes

First, I should have installed insulation along the box wall were the plate was set. Wood burn slowly and is a lot less dangerous than plastic but it burns anyway. Plywood in itself was a mistake to build a box, you shall use melaminated chip boards as melamine produce nitrogen when heated so cause oxygen depletion. But the best is plaster or cement.
Installing the pad inside the box was a possibility, but the aluminium plate was protecting it from mechanical aggression in case of troubles with the arms (detached arms are a common problem with deltas).

I should have connected the thermistor, I had the possibility on my board as the bed heater was not installed. I was knowing that you cannot measure temperature with an infrared thermometer on a glossy surface, but I forgot it. I should have stopped the test as soon as I understood my measure were wrong, just adding tape on the surface is sufficient. However, as the pad inflated, the heating element was probably no longer in contact with the aluminium plate so not heating it and this measurement won’t have given me a valid information. But opening the box for measure will have helped to notice earlier the smell. Frankly, this behaviour (inflating) was completely unexpected.

Always try to think to ALL what can occur. My main concerns about silicon beds were (and still are) electric, about the contact between the thermistor and the heating wires. The photos shows that, indeed, in case of overheating, all wire insulation fuses, plastic self-destroy and that creates a risk.

(c) Pierre ROUZEAU
Privacy - Vie privée - Imprimable - Rechercher
Page mise à jour le 21/01/2018 02:15