SSR failsafe
-
At a bare minimum you need to ground your bed chassis to the protective earth. If this is a stick on heater then please don't rely on a coincidental earth connection between the spreader and the chassis.
With regards to my earlier comments on two RCDs in series I could find no evidence of this negatively effecting safety. The only thing I could find was it was against regulation for the hard wiring of your house. This is because there is little ti know fault discrimination available and therefore a fault on one circuit could take out the lighting, leaving you wondering around at night in the dark, thus making it more dangerous. I always shut down my machines using the test button on a plug in RCD, which is therefore much better tested than the RCD that protects all the mains sockets in my house.
That said I will accept any warning against chaining RCDs in series if links to a good source (IEE etc) can demonstrate the safety hazard of doing so.
-
@singhm29 You need to consider both values when choosing the thermal fuse. I'd advise to go well beyond your normal current rating, something like a 20 A one. The max temp really depends on how high your bed can go and how high you're usually pushing it. I'd recommend bonding it to the silicone with RTV cement instead of a bracket when it comes to mounting it.
I'd second @DocTrucker on grounding - please don't rely on your bed being grounded via the chassis. Unless it's a delta and there are no moving parts in the bed assembly and you've tested that there's a low resistance electrical connection between the bed and the ground of the chassis.
-
Hmm not sure if I might be mixing things up but as it stands my heated bed came with two power wires but nothing for ground (I think that is normal for all heated beds) so I used a 3 prong power plug any have my SSR connected up correctly. However, the ground wire at this point is not connected and it was my belief based on this video that I should connect that ground wire using a screw in spade connector to my bed chassis. Does that sound incorrect/unsafe? My heated bed is secured from below by 2 aluminum extrusions, adhesive and 4 screws so I think I don't need to worry about it free falling from the printer if the adhesive wears out.
Currently, I'm only printing PLA at around 60C, what is the max temperature people want for their heated beds to cover most materials? I would assume 120C is a safe bet? So the thermal cutoff parameters I should be looking for would be above 120C and 15A or above? Essentially if either of those are exceeded it will trip the fuse?
Sorry for all the questions here I would just rather have things spelled out instead of making assumptions when it comes to safety!
-
I would suggest bringing the mains into your system through and IEC socket module combined with a double pole switch and fuse. Take the live and neutral from there to your relay. I would mount am earth stud on the chassis close to the socket and take the earth spade of the IEC to a ring connector around the earth stud. Ground the heat spreader to this point and ground any other bits of metalwork close to the heater to that too.
Grounding the negative of the psu to the main earth point has resulted in a more stable ethernet connection for me. I mean to do thus a little better but haven't had the chance yet.
-
You should consider a double mechanical pole contactor/relay between the mains input and the SSR. This doubles up the protection that the thermal cut off gives. If the duet detects a fault, it can be set to drop PS_ON which could be connected to this relay.
I'd always play on the safest side possible when using mains in anger.
-
@doctrucker I recommend a relay/contactor too, but my idea which I am currently executing on my printer is to wire it in a self-latching start/stop circuit. Wire the thermal fuse (be it as one-shot or a bimetall one) in series with the relay coil power. Maybe you could use the PS_ON to control it too. This way you should not have to use a fuse which able to disconnect 10-20A at mains voltage. It would be a pain to wire with thick wires and insulate properly WHILE attach thermally and mechanically at the same time anyway.
You could use 12V/24V in the control circuit. Electronically it will give you several possibilities:
-
It will trip on thermal runaway triggered by thermal fuse if SSR failed short or the firmware frozen.
-
You could trip it by software via releasing PS_ON before it actually ran away, but something wrong detected according to diagnostics.
-
You have to press a mechanical start button to activate. It will release on power-loss and never come back accidentally on power return.
-
It will not oscillate if you use bimetal thermal switch
-
-
@moczikgabor yup, all good stuff. Contactors can get pricey so you have to shop about a bit. If you get a three pole (look for three phase) you can use your reset button to initially power the coil then the third pole of the contactor to hold it on. Ground the coil (or coils) through the PS_ON with suitably sized flyback/flywheel diodes.
Edit: add you normally closed fault detecting fuses and snap switches between the coil and the PS_ON.
-
There are two issues here:
-
Use of a mains voltage bed heater. For me this demands using a 3-pin polarised mains plug so that you can be sure which connection is live and which is neutral (easy in the UK and Ireland, not so easy In many other countries), using a RCB/GFCI in the power feed to the printer, grounding the bed plate and metal frame of the printer, and using strain relief on the bed heater cable.
-
For any high powered bed heater (whether mains powered or low voltage DC powered), ensuring that under full power, the bed temperature will not exceed a safe limit. My preference is for passive safety; the bed temperature won't exceed a safe value when run at full power for an extended time. Failing that, use a reliable thermal cut out.
Using a SSR to cut all power to the printer is an extra level of safety. Using a dual pole contactor is IMO serious overkill, unless we are talking about printers costing many 1000s of pounds/euros/dollars and the cost of a failure is enormous. But it's often said that there is no kill like overkill!
-
-
@dc42 I think you should not just count the value of the printer itself. If it starts burning, the disaster could be much more than that. If you take this in to consideration, a contactor for 15-20 EUR isn't that much of a cost. An SSR would not be cheaper either nor the kW-powered bed heater was, and the Duet board itself isn't that cheap, may be an overkill for a cheap printer. But we want serious things.
If I build something, I always build as industrial grade, idiot proof, safety first. I don't like underengineered things which I am not fully confident in.
Limiting the power of the bed heater so that it could not reach unsafe temperatures would defeat the purpose of the high power bed: heating up fast.
The UK plug is polarised, but for example where I live, the Schuko isn't. You still can use a GFCI, but just assume both wire as live fuse them both and break them both at mains switch or GFCI or safety relays, everywhere where galvanic isolation is expected to happen. Controlling a mains powered device with SSR or relay in one line is okay, but safely disconnecting it you should use double pole switching. 3mm distance between open contacts... Etc...
-
@doctrucker Expensive is always depend on budget, but an SSR is somewhere in the same price range as a contactor.
I don't know where you live, but here in the EU you can buy quality ones for 20 EUR.For example:
ABB AF09-30-01-11
24V (20-60V DC), 3xNO + 1xNC, Imax=25A -
Quite alot of new terminology to take in here! Looks like adding all of this in will be a project itself lol here I was thinking adding in a thermal cutoff switch would do the trick.
-
@moczikgabor I said they can be expensive, shop around.
-
SSR relays are not safety devices! Omron make some of the best, and their documentation (which the duet fire safety documentation references - https://duet3d.dozuki.com/Wiki/FireSafety) clearly states do not use them in situations where they are the only thing switching a device off. Selecting a high current capacity isn't good enough. Spikes and such faults can kill them such as a meaty lihtening strike near your building or a faughtly appliance with high internal voltage on the same circuit, such as a microwave. I was advocating using the SSR to do the modulation of the heater and the mechanical relay to enable, or disable the heater.
@moczikgabor said in SSR failsafe:
The UK plug is polarised, but for example where I live, the Schuko isn't. You still can use a GFCI, but just assume both wire as live fuse them both and break them both at mains switch or GFCI or safety relays, everywhere where galvanic isolation is expected to happen. Controlling a mains powered device with SSR or relay in one line is okay, but safely disconnecting it you should use double pole switching. 3mm distance between open contacts... Etc...
I've made very similar comments on this forum about reversable plugs, perhaps even earlier on in this thread. Only thing to be aware of is two standard fuses isn't fool proof. If you have an over current fault that isn't a short to earth it's random which fuse will blow first. Both fuses may not blow, and the device may end up off but full of mains potential AC voltage. Yes procedure should always be turn off the mains when you think you have a fault on the mains side, but that's exactly that - a procedure - which can go wrong, be ignored, or accidentally missed of the training for a new helper that is standing in for you at a trade show while you take a break etc.
Four options. Only you ever touch the machine, habitually check your sockets (good idea, it checks earth!), clear warning notice on machine to disconnect mains after a fault, or use a dual pole (mechanically linked) resetable over current fuse.
-
@doctrucker The fues's purpose is to protect against fire, not protecting against electrical shock. If somebody is that dumb that carelessly touch parts connected to mains voltage without safely disconnecting it first - that will be their responsibility. In many printers even the mains switch is SPST only, so you can't rely on that either.
If you work on a circuit, it is yours responsibility to disconnect and check it first. No insulation, no fences, no other systems will save you with 100% certainity. Many times maybe, but once in a life maybe something will fail. Not following safe (I call it common sense) procedures but relying on someone or something to protect you is looking for a disaster.
You need two fuses, not because they will blow up at the same time and disconnect both lines, but because it cannot be guaranteed that which line is neutral and which is live. There is a possibility of earth (chassis) short anywhere in the system. Especially common in heaters, the moving bed cable's insulation can break, the cable end may break off, the SSR may fail short to it's heatsink, etc... And if you only have fuse in one line, which happened to be the neurtal today, then it will not blow. This is the ONLY reason you need fuse in both wire, not because it will safely disconnect your printer to work on.
Maybe you misunderstood my text before, I am not advocating to use SSR for safety disconnect. Exactly the opposite, exactly you worte last. I recommend using it for control, because it is fast and reliable for that purpose, but using an independent (manually enabled and thermal cutoff + PS_ON disableable) double pole contactor to disconnect in case of failure.
-
@singhm29 said in SSR failsafe:
Quite alot of new terminology to take in here! Looks like adding all of this in will be a project itself lol here I was thinking adding in a thermal cutoff switch would do the trick.
It all depends on you use case. If you have a low powered bed heater, which can not reach unsafe temperatures, and you always supervise your printer (frequently enough) you may live without it. But in a 12V system, for a 12V bed, using just an automotive relay and a thermal fuse is dirt cheap and easy to wire - yet greatly increase the safety.
Once you go for high power mains voltage bed heater... I should say that this safety system is mandatory. It could heat to the melting point of aluminium in no time. Maybe while you come back to check your print, it is already on fire. And use quality parts. You don't want the safety to fail.
-
@moczikgabor said in SSR failsafe:
Limiting the power of the bed heater so that it could not reach unsafe temperatures would defeat the purpose of the high power bed: heating up fast.
The bed heater on my delta heats to 100C in just a few minutes, and tops out at 200C at full power - within the rating of the silicone heater. There are no plastic parts in contact with the bed. I don't consider that reducing heating time by perhaps 2 minutes is worth sacrificing passive safety for.
If you are trying to make your printer fireproof and suitable for unattended operation, then IMO it needs to be inside a fireproof enclosure, with a smoke/heat detector to determine if there is fire or excessive temperature inside the enclosure. If you have that, then I agree with you that it should use a double pole contactor to disconnect the mains when fire is detected.
-
@moczikgabor I was generally agreeing with you. My comments about SSRs were in responce to another poster advocating them as a safety disconect in fault.
Reading through my other posts would also show I advocate double pole switches. I scrap single pole switches as soon as I can, and only run machines after checking the wall socket.
Regards the fusing there are two failure modes, overcurrent and fault. I agree two fuses protect against earth faults but in my opinion offer inadequate protection against over current. In that case something has gine badly wrong, and may have bared conductors which would not normally be bare, and may not have earthed. People may warm up machines first, and without large lights it may mot be immediately obvious that a fuse has blown, and therefore it's hardly about a user being stupid.
All that aside I think we are generally on the same page here. Mains heater beds are a luxury on most machines and need to be treated very carefully and nit just put on by hibbiests who have little experience or training in mains work.
-
For my clarity, in some other countries neither of the 2 current carrying conductors are intentionally grounded in general purpose outlets? Here in the US 1 is always intentionally grounded and is commonly referred to as the neutral. (Although it actually isn't a neutral by definition) The other is 120V RMS to ground. Our "240V" which is used for larger appliances (dryers, ranges, A/C, etc) is 2 ungrounded conductors but both 120V RMS to ground and 180 degrees out of phase.
-
@alexander-mundy said in SSR failsafe:
For my clarity, in some other countries neither of the 2 current carrying conductors are intentionally grounded in general purpose outlets?
I live in Hungary (but many places in the EU are similar too). We use three-phase Y connected distribution system where the central point of the Y is connected to earth AND carried by wire as a neutral. The phases are 120 degrees apart. If you use three phase power, those connectors are polarised, the order of L1, L2, L3 and location of N, PE are guaranteed.
If your house have only one phase, you will get one of the three, plus N. It is clear (should be...) through the whole house which is which. If you wire something directly in the distribution cabinet, you could rely on it. However, the one-phase consumer socket, Schuko isn't polarised. There is a standard that wire the L on the left side, N on the right on the socket, but the plug is intentionally symmetric, mechanically reversible, so you can't guarantee which wire is which in an equipment - the distinction ended at the wall.
Plus, because it is symmetric, the electricians rarely even care which is the live, even in the socket, so even there is a standard, better to expect nothing...
Polarised one-phase industrial sockets do exists, but it is rarely used, you'll never see it in a house.
-
Think the car charging sockets are industrial style 16 and 32A polorised connectors.
In a single phase UK system and much of the EU you have one hot 230VAC, one neutral, and protective earth. Ultimately the protective earth and neutral are the same thing, but by design the protective earth is not used as a current conductor. If there is a current in the protective earth, there must be a fault in the appliance. This results in a mismatch between live and neutral current flow, which trips an RCD if in the circuit.
The benefit of seperate earth and neutral is there can be a potential in a long length if neutral due to the resistance of the conductor and the number of connections in the loop or spar. Avoiding passing current in the protective earth under normal circumstances reduces this increase in potential.
The water gets slightly muddied by noise filters which can dump noise to the protective earth.
Edit on that last topic a sparky retold a story of a troublesome RCD tripping that ended up neing too many TVs on one circuit which dumped noise to earth.