AC Heated bed?

Hi,

I wanted to ask most of electronic experts there although I myself not one but might be a good question and idea, is it possible use an AC heater cartridge for heated bed instead of a large pcb board that most often either not up to spec or too high in current?

For example using a kettle's heating element, flattened and sandwiched with silicone cover at bottom and aluminum on top, a relay to control the circuit I believe similar to bang-bang method instead of pwm in marlin's config file.

e3d has one but it is a cartridge style in here. I'm thinking that this might be another step forward than keep relying on dc heated bed that might have problem reaching to target temp

What would you guys think, just an another idea I got

You're right that DC beds are generally underpowered, and it's sort of pointless to use regulated DC, and so a bigger, more costly power supply, to provide power to a big resistor. I've done that in SoM, my second printer (450W heater @24V) and won't do it again. The power supply fan sounds like a vacuum cleaner.

I think that using a small, AC powered heating element on an aluminum plate is going to end up creating a hot spot right above the heater and fall off in temperature as you move away from the heater on the bed plate. Something like that would be a good, cheap way to make an enclosure heater- just bolt a heatsink to that small, high powered heater and blow air over it with a fan.

You can get quality AC powered bed heaters- I used a 300x300x750W Keenovo heater on a 300x300x8mm MIC6 bed plate in my last printer and it has been great. I can print edge to edge with no worries about prints letting go before they finish.

Here are thermal images of it:





When you wire an AC heater you have to think about all the things that can go wrong and plan for them. I used an electrical fuse in case the wires get damaged/pulled (ground the printer's frame!), and a thermal cut-off on the bed plate in case the SSR driving the heater fails (they fail "on") or the controller board loses its mind.

Details on the bed design here: [drmrehorst.blogspot.com]

Thermal performance here: [drmrehorst.blogspot.com]

Edited 1 time(s). Last edit at 01/26/2018 12:46PM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

You may find [duet3d.com] and [duet3d.com] helpful, in particular the safety recommendations at the end of the second one.

---

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].

I am building a 400x400 printer using a 800W 380x380 Keeenovo silicone heater. It will be attached to a 6.6mm thick MIC6 plate. (Trying to find a 8mm source in the US that is reasonable priced)

I was intending on use the integrated thermistor, but I recently noticed the Duet firmware supports what I believe is is averaging off temperature inputs? I've been considering use of a PT100 or Type K sensor for robustness.

So... are there measurable gains to be had with averaging, or is there an option to use the the PT100 or K and then use the thermistor as a fail safe should there be too much of a temperature differential?

This is my first round using AC mains as a heating source and I'm looking for a bit of extra safety and trying to avoid thermal fuses, etc..

Bed temperature is not especially critical, also the ideal indicated bed temperature will depend on what surface you put on top of it (PEI, glass, etc.) as well as on the filament. So high precision in measuring bed temperature is unlikely to be worthwhile, and a thermistor is adequate. OTOH if you are using a Duet with a 2 x PT100 daughter board and you have only one hot end, then you have a spare PT100 channel anyway, so use a PT100 if you prefer.

The Duet firmware doesn't currently support averaging of temperature inputs (although it does support multiple bed heater zones - perhaps that is what you saw). You can get an averaging effect by connecting multiple thermistors or PT100 or PT1000 sensors in series or in parallel. However, with 6.6mm alu plate and a silicone heater, you shouldn't get much variation across the bed anyway, so I doubt that it would be worthwhile.

---

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].

Quote
gen2eng

This is my first round using AC mains as a heating source and I'm looking for a bit of extra safety and trying to avoid thermal fuses, etc..

Extra safety and avoiding thermal fuses are opposite concepts. If you use active measures to shut things down in the event of runaway bed temperature, it will probably be fine until the controller reading the bed temperature decides to ignore it. Using a TCO in thermal contact with the heater/bed to interrupt power to the heater is the ultimate in runaway protection. You can use both methods, but if I were choosing one or the other, I'd definitely go with the TCO. Printer controller firmware usually has pretty extensive protective measures built in, so use those when possible- the most you have to do is enable and configure them. Some are probably already active and configured by default.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

How fast those keenovo AC heaters perform let's say from 30c to 100c?

Then with the AC heaters, we typically could go slightly thinner wires probably 18 or 20awg right?

My main problem with the pcb dc heated bed was that of course, underpowered and often beaten by a fan blowing onto it thus cooling the whole bed down even further. I'm waiting for a cotton insulation sheet to insulate underneath the pcb

Many (perhaps most) PCB bed heaters are under powered. One of the problems is that the copper thickness on cheap PCB heaters isn't well controlled. Another is that they are often used with cheap 12V ATX power supplies, which have poor regulation so that the 12V output drops under load, and cheap RAMPS electronics which has too much voltage drop in the bed heater circuit. The PCB heater on my Ormerod with a good PSU and Duet electronics will ultimately reach 120C. It's been a while since I heated it to 100C but AFAIR it took about 10 minutes.

How fast a heater will get the bed to 100C depends on the power density and the thickness (which affects the thermal capacity) of the bed. My Delta uses a 330mm diameter round bed with a 350W silicone heater and 5mm thick aluminium plate. This works out at 0.4W per square cm. There is insulation under the bed, mostly to protect the electronics and help support the heater, but it also helps a little to reduce the heating time. The heating time from 22C to 100C is 5 min 24 sec. At full power the bed will ultimately reach around 190C (I stopped it at 180C when testing it).

You can use a higher power density if you want faster heating times, but then a thermal cutout becomes mandatory so that you don't exceed the heater temperature rating (typically 200C for silicone heaters).

If you use a thicker bed plate then the heat up time will be a little longer due to the increased thermal capacity, but the ultimate temperature reached at full power will not change.

HTH David

Edited 3 time(s). Last edit at 01/27/2018 03:14AM by dc42.

---

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].

Quote
DC42
The Duet firmware doesn't currently support averaging of temperature inputs (although it does support multiple bed heater zones

Thanks for the clarification as I did misinterpret what I read.

Quote
the_digital_dentist
Using a TCO in thermal contact with the heater/bed to interrupt power to the heater is the ultimate in runaway protection.

I agree the TCO is the best route, but was looking for some "some" additional safety at the software level. If I remember correctly, Repetier-Firmware allowed use of a secondary temp input to detect failed and separated thermistors. I've found my Duet 0.85 to be extremely stable having run many small batches of parts and not require periodic resets between prints like the old 8 bit hardware.

Quote
sarf2k4
How fast those keenovo AC heaters perform let's say from 30c to 100c?

Then with the AC heaters, we typically could go slightly thinner wires probably 18 or 20awg right?

My main problem with the pcb dc heated bed was that of course, underpowered and often beaten by a fan blowing onto it thus cooling the whole bed down even further. I'm waiting for a cotton insulation sheet to insulate underneath the pcb

I have recently changed my heatbed over to A/C mains power - it's only 200W but it can get the bed up to 100C in around 10 minutes, where it used to take up to 30 minutes with the old 24v PCB heater (and it wouldn't get there at all if the room was cold). Any attempt to run a print cooling fan would knock the bed temperature right down almost immediately, especially on the initial layers where the extruder was close to the bed. Now, no problems ( I have taken it up to 120C within 15 minutes, but that was only for the sake of testing).

My heatbed is now 5mm aluminium tooling plate - I still have insulation under the bed, and I do have a 150C thermal cut-out installed. The bed is driven by an SSR, and there are a LOT of protective earths in and around the heatbed and printer metalwork. You do have to think hard about earthing/grounding all metalwork, and also about restricting physical access to keep fingers away from high voltages at all times (barriers, enclosures, etc).

One bonus is that I was able to change to a much smaller PSU for the rest of the printer's needs and it doesn't have a noisy fan; my machine is a lot quieter these days.

Quote
gen2eng

Quote
DC42
The Duet firmware doesn't currently support averaging of temperature inputs (although it does support multiple bed heater zones

Thanks for the clarification as I did misinterpret what I read.

Quote
the_digital_dentist
Using a TCO in thermal contact with the heater/bed to interrupt power to the heater is the ultimate in runaway protection.

I agree the TCO is the best route, but was looking for some "some" additional safety at the software level. If I remember correctly, Repetier-Firmware allowed use of a secondary temp input to detect failed and separated thermistors. I've found my Duet 0.85 to be extremely stable having run many small batches of parts and not require periodic resets between prints like the old 8 bit hardware.

There is such a feature on the Duet/RepRapFirmware too, see [duet3d.com]. Amongst other things, it allows you to use a second thermistor to flag a heater fault and turn it off if a preset temperature is exceeded.

---

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].

I guess this is where AC heatbed will start to take over. I've found keenovo's silicone heating pad in aliexpress and considering how much stuff I need to include just for the sake of using ac heatbed can be a bit overwhelming.

How much power in amps does the ac heatbed pulls if it is rated at around 500w and powered using 240v?
Any temperature drops if for example a big standing fan or the printer's nozzle fan blowing at full speed and how fast will it catch up?

If I don't rely on the TCO (If I understand it right it is Temp Cut Out module?) to cut off the overtemp but to use the marlin's thermal runaway protection, would that suffice?

I understand grounding is very important, I saw maker's muse got shocked due to improper grounding in his whole circuit, can I connect a wire from a printer's frame to the dc psu's ground terminal directly?

I wanted to reduce power outlet usage, can I wire the ac heatbed into the psu's ac input terminal then to the relay?

Any other alternative than using bulky SSR?

Very informative so far from you guys. I didn't know AC heaters already existed, or maybe I never noticed or didn't look it up

Quote
sarf2k4
I guess this is where AC heatbed will start to take over. I've found keenovo's silicone heating pad in aliexpress and considering how much stuff I need to include just for the sake of using ac heatbed can be a bit overwhelming.

How much power in amps does the ac heatbed pulls if it is rated at around 500w and powered using 240v?

That would be 500/240 = 2.08A.

Quote
sarf2k4
Any temperature drops if for example a big standing fan or the printer's nozzle fan blowing at full speed and how fast will it catch up?

Assuming you have sufficient heating power (see my earlier recommendation for 0.4W.sq cm), how fast it will catch up depends on how well the PID is tuned.

Quote
sarf2k4
If I don't rely on the TCO (If I understand it right it is Temp Cut Out module?) to cut off the overtemp but to use the marlin's thermal runaway protection, would that suffice?

No. If your heater will reach dangerously high temperatures if it is left on at full power, you should always use a TCO to guard against electronics or firmware failure. The alternative is to choose the heater power carefully so that it is powerful enough to reach the temperatures you want quickly, but not so powerful as to reach a dangerously high temperature when left fully on for a long time. That is what I attempted to do, but I should probably have used a slightly less powerful heater e.g. 0.35W per sq cm instead of 0.4.

Quote
sarf2k4
I understand grounding is very important, I saw maker's muse got shocked due to improper grounding in his whole circuit, can I connect a wire from a printer's frame to the dc psu's ground terminal directly?

Connect a wire from the printer's frame to the mains ground terminal. Also connect a wire from the negative output terminal of the DC power supply to mains ground terminal to prevent the issue that maker's muse illustrated. But beware of ground loops of you connect the electronics to a PC via USB, see [duet3d.com].

Quote
sarf2k4
I wanted to reduce power outlet usage, can I wire the ac heatbed into the psu's ac input terminal then to the relay?

Please explain which power outlet usage you are referring to.

Quote
sarf2k4
Any other alternative than using bulky SSR?

The additional bulk of the SSR is much less that the bulk you save by using a smaller PSU because the PSU is no longer supplying the bed power.

Edited 2 time(s). Last edit at 01/27/2018 06:01AM by dc42.

---

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].

Quote
sarf2k4
How fast those keenovo AC heaters perform let's say from 30c to 100c?

I have 750W heater on a 300x300x8mm MIC6 plate. It goes from 20C to 105C in about 4.5 minutes. There's no insulation on the underside. The "speed" depends on the heater power and the mass and material you're attaching it to.

SSRs come in a lot of different shapes and sizes, so it's possible to reduce the bulk a bit. This one is good for 5A (about 600W)

Insulating the bottom of the heater will allow it to heat up a little faster, or allow a lower powered heater to get to a slightly higher temperature, but it slows down cooling of the plate, too. Just something to think about if that sort of thing matters to you.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

The Keenovo heat pads are very good, and the company gives good customer service - I've had emailed queries answered in 30 minutes or so. Not bad for communications from UK to China!

I used this SSR from Rapid: Kudom SSR. It's tiny at 38.5mm x 28.5mm x 20mm, but will handle 10A - my 200W heater takes less than 1A. I also got a transparent plastic cover for safety's sake (NB: don't fit the cover until you've finished wiring, as it's a pig to get off again - guess how I found out). Both were cheap enough, and proper industrial quality (i.e. not dubious ebay specials).

BTW: don't rush to get a high-power heater. My 200W one is probably a good and serviceable minimum wattage (for a 200mm square bed) as long as you put insulation underneath. There's a trend to go for super-powerful heaters, but IMHO they're not really necessary. Of course, others may hold different views! Oh, and if you do get one, follow the fitting instructions closely - once it's on, it's almost impossible to remove without damaging it.

Edit: specified the bed size for the wattage I was discussing!

Edited 1 time(s). Last edit at 01/27/2018 09:32AM by David J.

Here is my rough sketch of using a single main outlet instead of two, and how frame are wired as well. Basically the heated bed will leech the power from the same wire as the psu

[CoSketch.com]

Edited 1 time(s). Last edit at 01/28/2018 03:00AM by sarf2k4.

That looks right, except that SSRs are normally single pole so the neutral wire from the mains goes directly to the bed heater, not through the SSR. Use a mains inlet connector with a builtin switch, fuse and indicator light.

Edited 1 time(s). Last edit at 01/28/2018 03:05AM by dc42.

---

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].

I didn't notice Tom used the AC silicone heater back in 2015 since in the video he didn't specify exactly the type of silicone heater he used here [www.youtube.com]

So that got me thinking that he might just use a regular dc silicone heater and wired through the ac like in the drawing he made.

A 650W heater isn't a low voltage heater.

Heaters don't care about DC or AC, but you have to be in the ballpark with the voltage they're designed for. Don't even think about connecting a 12V heater to line power (117 or 240VAC). It will burn up in an instant, the same way a 12V light bulb would burn up if you plugged it in 117 ro 240V.

Don't forget the electrical fuses that weren't shown in your sketch. It would be a good idea to use at least two- one for the printer's power supply input and one for the bed heater circuit. Some power supplies have fuses built in- you might want to check it.

Do connect the power supply input ground to the line input ground/neutral. I would not connect the power supply negative output to the printer's frame with the line input ground. The voltage reference in the supply comes from the output side of the supply and if you tie it to the power line ground you're liable to be adding power line noise to the DC ground. But if you search the web for engineering documentation from switching power supply makers, it isn't hard to find justification for either connecting it or not. In the end, you probably have to evaluate it on a case by case basis. If you get weird behavior with the output of the DC supply grounded, open the DC ground connection, and the reverse. In a home environment with a relatively quiet power lines, it may not make any difference. In terms of safety, grounding the output to the printer's frame means you've got one side of the DC power supply everywhere and it is pretty easy to accidentally create a short. If you leave the supply floating, you don't have that easy accidental connection to DC ground.

This is a overall wiring diagram of my printer. Note the different ground symbols used for DC and AC ground. I keep the two separate. Note separate fuses for the main power supply and the bed heater circuits. The third fuse is now used for the enclosure heater.




If you end up getting one of the common SSRs with exposed screw terminals, you can print a cover that will prevent accidents. I designed and printed a cover for the bed heater SSR in my printer. The design file is here: [drmrehorst.blogspot.com]



Edited 1 time(s). Last edit at 01/28/2018 09:40AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
... I would not connect the power supply negative output to the printer's frame with the line input ground. The voltage reference in the supply comes from the output side of the supply and if you tie it to the power line ground you're liable to be adding power line noise to the DC ground. But if you search the web for engineering documentation from switching power supply makers, it isn't hard to find justification for either connecting it or not.

The issues with not connecting the negative output to mains ground are:

1. Some PSUs have a high leakage current from the mains input to the output, to the extent that you can get a shock from the output if you don't ground it.

2. There is a risk that the insulation in the PSU will break down and the output will become live.

See [www.youtube.com]. ATX PSUs already have the negative output connected to mains ground.

Connecting the negative output of the PSU to mains ground does not of itself create a noise issue. But if you then connect the electronics via USB to a PC (other than a laptop running off batteries) and you don't use a USB isolator, or a RAMBO board that has a built-in isolator, then you create a ground loop which can be problematic - see [duet3d.com].

Edited 1 time(s). Last edit at 01/28/2018 10:40AM by dc42.

---

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].

Quote
the_digital_dentist
Do connect the power supply input ground to the line input ground/neutral. I would not connect the power supply negative output to the printer's frame with the line input ground. The voltage reference in the supply comes from the output side of the supply and if you tie it to the power line ground you're liable to be adding power line noise to the DC ground. But if you search the web for engineering documentation from switching power supply makers, it isn't hard to find justification for either connecting it or not. In the end, you probably have to evaluate it on a case by case basis. If you get weird behavior with the output of the DC supply grounded, open the DC ground connection, and the reverse. In a home environment with a relatively quiet power lines, it may not make any difference. In terms of safety, grounding the output to the printer's frame means you've got one side of the DC power supply everywhere and it is pretty easy to accidentally create a short. If you leave the supply floating, you don't have that easy accidental connection to DC ground.

I don't think I understand this part, to summarise things up, don't connect psu's DC negative to the frame while the frame connected to the AC ground at the same time?

Yes. Leaving the DC output floating may prevent some of the USB ground loop problems. As DC42 points out, in some power supplies the DC output may already be connected to the line input ground.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

What does this TCO looked like? What about the price and what is it that I should be looking for in terms of the spec?

TCOs come in different configurations, but they're generally cheap. The first spec is to make sure the TCO can handle the bed current. If your bed will take 10A use a 12-20A TCO. The next spec is the operating temperature which is the temperature at which the TCO will open. The current through the TCO creates some self heating, so you when you select the operating temperature you have to allow for that. The design info for the type I used said that at the current my heater uses, the operating temperature should be 24C above the highest expected heater temperature. I expect that I may want to go as high as 160C bed temperature for printing high temperature materials, so I used a 184C TCO. The type of TCO I used has a "hot" case, and I attached it to the bed plate, so I wrapped it in kapton to electrically insulate it from the grounded bed. I used high temperature insulated wire (glass/silicone) to connect to the TCO. You don't want to use PVC insulated wire (or heatshrink) for that- it's going to be getting hot. If the insulation melts or burns off, it doesn't insulate any more.



I mounted the TCO on the bed plate under the assumption that the heater's adhesive will keep it stuck to the plate forever, which may not be a good assumption. It would be better to attach the TCO to the heater itself, using high temperature silicone. That way if the heater should peel off the plate, the TCO will stay with it and do its job. The assumption there is that the high temperature silicone doesn't let go... There are always assumptions...

This is the part I used. I would avoid buying this sort of thing from the Chinese suppliers on ebay. You want to know that the thing is going to work. Get one with proper certifications based on testing and qualified manufacturing processes. Even fully specced parts are cheap. Buy a couple spares and keep them with the printer, just in case.

I would not use a self-resetting TCO. If the bed gets to a temperature that trips the TCO, there is a serious problem that should be fixed before letting the bed heat up again.

Edited 1 time(s). Last edit at 02/06/2018 09:06AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

So this TCO is a temperature based fuse and it is a one time use only right? If I got only one TCO, it got blown or cut out due to exceeding the temperature, I got to replace to a new one right?

Yes, the type of TCO I used is a one-shot device like an electrical fuse and has to be replaced if it ever does its job. In my printer the TCO as a safety device, not a temperature regulation device. It should never operate when the printer is working properly. At the TCO operating temperature I selected, a blown TCO indicates a serious problem that should be fixed (dead SSR or insane controller board) before powering the heater on again. The last thing I would want is for a self-resetting TCO to turn the heat back on a few minutes after it turned it off.

Maybe you have other ideas...

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]