Multi heater bedquestions
Tpmoses last edited by Tpmoses
I have a printer at work that was built by someone who no longer works here. The bed was heated by a 1000x1000mm 120V silicone mat controlled through a SSR that has recently died. I've gotten the go ahead from my boss to go another direction with time in mind, and the best option I can find is to get 4 smaller mats as replacement.
I believe I have found a suitable option for the mats themselves but now the question is how to wire them. Having a zoned build plate would be nice, but not a priority.
The question I have now is, with four 120V 1300W heating mats each with their own 1000k thermistor what's the recommended way to wire them?
The board I am using is the Duet 2 Wifi
So saying I didn't need it to be zoned and I just wanted all four to work together as one, could I just wire them all in parallel? Can the power be run through a single SSR? I have an open heater port on the board as well as an open thermistor port, so if I needed to make it two zones I could. I see that on a different thread there are options to wire the thermistors together either in parallel or series, so its more the power that I'm worrie about.
Any Advice would be greatly appreciated. The heaters are on order and should be here at latest Tuesday. I'm just trying to think in advance and see if there is anything else (IE another SSR) that I need to order.
peterartechno last edited by
I would take into account that you have 4x1300W of power. Do you have a heavy duty power group? Normally the fuse would snap with this wattage.
I use two silicone mats that are put in parallel which give 530W each.
If you have a 1m2 bed with 5200W of power, that is more than enough. I have a bed of 1.4m2 with 1060W which gets my bed up to 50 degrees.
I would suggest doing it in parallel as that does not change the voltage of your flow for each heater. If one of your heaters fail and it is in series, they all stop. Also I don't know if your heater can work on 30V (120/4) when placed in series?
@peterartechno I have been doing some math. A single heater would take 10.83A, and is essentially an 11.08ohm resistor.
In parallel they would act as a 2.77ohm resistor and require 43.32A, so too high and would just pop a breaker.
In series, they would act as a 44.32ohm resistor, using 2.70A and probably take forever to heat up.
However, if I wire them up in Series-Parallel (two parallels in series) they act as an 11.08ohm resistor, and require 10.83A.
SO, in Series-Parallel it should be a safe current requirement but still have enough oomf (technical term) to heat up in a reasonable time frame.
o_lampe last edited by
A single mat is 250x250mm sized and has 1300W??
I don't think they take forever to heat up. Unless the bed is a monster chunk of metal.
Just an example: my printbed is 300x200mm and has only 180W. I'd call it a bit slow with heatup times of ~5min. (6mm MIC6 plate)
@o_lampe A single mat is 510mmx510mm at 1300W and should take about 20min to heat, so for all four in the setup should take about 40min to get to 100C. The plate they'll be heating up is a 1170mm x 1170mm x 6.35mm aluminum plate. The previous bed heater was a 1000mmx1000mm custom silicone mat that took about 2 hours to get to temp.
I have found that the second you get above about 5000mm^2 of build volume the warm up times get loooooong. The plate is just such a large heat sink it takes a lot for the proper heat soak.
And just for reference, I am using this printer in a manufacturing setting, making parts that are taken and used as positives for mold making for fiberglass parts. The extruder uses pellets instead of filament, and so far its exclusively ABS, Hence the need to heat above 100C
peterartechno last edited by
@Tpmoses You take the amperage value of the heater when they are all in series. When you use series-parallel, that would make it 20A right? Not 10. Still the heater has to be able to work at 60V. I don't know if you have an isolated heater but I got electrocuted once by my aluminium heated bed so make sure you have it grounded.
I have a 10mm aluminium plate which takes up to 30 min to reach 50 degrees.
@o_lampe If he says so, 5200W is a lot of heating power.
@Tpmoses Why above 5000mm2? It is just a matter of the amount of power that you put combined with the thermal energy consumption of your material. I have my heaters set so I can decide if I want both heaters or just one. Waste of energy to power them both if I only use half the area.
jens55 last edited by jens55
The problem you will be running into is that you have way more power requirement than you had before. The average 120V outlet can supply 1500W (15A really but 1500W is safe). You will need 4 separate 120V circuits to run that puppy and you will need 4 SSR's.
A single signal line paralleled to the SSR's will work. You can choose one of the temperature senders as the controlling sensor and you can either not hook up the others or display them separately on the temperature graph.
I do not know if you can just parallel them ....
Edit: I wish I could learn to read all replies before commenting
Anyways, the issue with a series/parallel heater arrangement is that if you run a 120V heater on 60V, you cut heating power to one quarter. Each pad would deliver 325W and the four combined would give you 1300W which is not enough for that kind of a bed.
Being in a manufacturing environment, you might have 240V outlets available to you. If you run two pads in series on 240V, each pad will (in theory at least) get 120V and produce full power. Parallel two series circuits on two 240V power lines with two SSR's will get you full heating on only two circuits.
Even better, if your 240V outlets are three wires plus ground, then you can split that one power line into two 120V lines which can each be run by their own SSR to power a single heater.
The problem with running two heaters in series on 240V (ie not split into two 120V lines) is that you depend on the resistance of the pads to be exactly the same. If one pad has a higher resistance then the other, it will heat differently and you could have different heat zones. This is, IMHO, not a good way of doing things. If you heat each pad with it's own 120V then you can still get different heat zones but the differential is not as much since you are using a fixed input voltage. If you are hooking them up in series, you might have one heater having 140V across and the other 100V across (as an example) while in parallel you are guaranteed 120V across each.
One more thought ... it is recommended that you supply positive pressure on the heat pad rater than depending on the glue to hold the pad in constant contact with the bed plate. With the amount of power involved (fast heating) and the amount of surface involved, I would make sure that I was providing that pressure onto the heat pads.
@jens55 The big question is I am worried about is if I were to run them in parallel if the current draw would be too much for the breaker. 240V would be nice, but my boss would prefer not to have to do that. The bed and frame are all grounded to common....... because the guy who built the machine didn't ground them and I got shocked. We would have gotten another large single heater, but one that large wouldn't be even available for another month, something my boss was not willing to wait for. The whole printer is enclosed, I'm most worried about being able to hold temp, not so much about warm up time.
Hence, why I came on here to ask for what the best way to wire it would be.
The pads will have positive pressure, the bed is made up with two aluminum plates a top and bottom plate, the pads attached directly to the top plate with cork sandwiched between. the whole assembly is clamped together.
@Tpmoses , sounds like you got things pretty well covered. As far as power is concerned, you really have limited choices - 240V or 120V. You could put heavy duty 120V circuits in that van handle 30A and run two heaters off each circuit but in my mind it doesn't make sense to run 120V for those power levels. The choice is between running 4 dedicated circuits of 120V or running two dedicated circuits of 240V. You could also run a single dedicated circuit of 240V 30 amp.
As I explained before, you do not have the option of running two pads in series because if you run the series/parallel configuration you would have 1300W total heating power.
I guess at this point it would make sense to ask how much wattage you had on the old heat pad .... that and the previous heat up time will tell you if 1300W will do the trick.
@jens55 From what I can find the previous bed mat was a 220v 2000w, being run off of 120v, so 1000W. I can't be 100% sure because there are no markings on it.
I have been seeing that generally, 0.6W/cm^2 is the ideal when calculating what you need, so if y bed is 110cm^2 1300 should be plenty? let me know what you think.
@Tpmoses, as previously noted if you apply half the voltage you get 1/4 of the power so if it was indeed a 2000W heater, it only supplied 500W if run at 120V.
If it was hooked up that way and you got sufficient heat in the bed (after how many hours of heating) then you could run two parallel strings of two series heaters on one SSR on 120V.
It is not a configuration that I would recommend.
Yes, 1300W would likely work if you are very patient.
jens55 last edited by jens55
Just as a reference I run a single 120V pad on my 500x500 heat bed. I believe it to be rated at around 1500W plus or minus 200W. It gives me good response. I would not be happy with 1/4 of that power. I think my 300x300 printer is 700W.
If we assume 0.6W per cm^2 and your bed is 110cm x 110 cm then according to my math you should be providing 7260W for that size of a bed.
I must admit that this seems high so maybe I messed something up ..... but it does suggest that 1300W will result in verrrry long warm up times and you might not reach the bed temperature you desire based on insulation around the printer. If it is a well insulated unit you will get enough heat but with long warm-up time. If it is not well insulated you might never get what you want.
o_lampe last edited by
@Tpmoses and @jens55
To mention all options: there is also the option to use star-delta conversion. (at least in Germany)
I think we all agree, that you need a 120V 3-phase outlet at least.
You can connect to eg. L1 and L2 instead of L1 and N and get 1.7 x 120V = 206V
Connect two mats in series and use L2 and L3 for the other two mats.
I wouldn't be too concerned about driving them in serial, you could read their resistance with a multimeter and find two pairs that match best. The big chunk of metal will spread the heat evenly over time.
About the thermistors, you can wire them anyway you like. For RRF it doesn't make a difference as long as you specify them correctly in the config.
But if you are concerned about uneven heating, you can specify four independent thermistors. One for the heater object model, the others as your reference.
Update, after talking to my boss, I'm switching the plan. We instead are going to go with four 1000W heaters, run them in two parallels with two separate 120V sources, and 2 SSRs with the same signal wire. That way the current draw wont be too much for the current breakers, and each of the heaters will get 120V.
What' size breakers are you using ?
Tpmoses last edited by Tpmoses
@jens55 They are 20 amp breakers, one parallel pair of the 1000W heaters will draw a max of 16.6666667A. Found another person who made a similar bed size with the same heaters, he had run into the same issue I was having with too big heaters, and had to go to this size.
@Tpmoses , sounds good. I asked because standard circuits are 15A and wouldn't have handled two heaters at a time. You are ok with a 20A circuit.
Okay just to make sure, it doesn't hurt to ask. The temp sensors on the heaters are NTC 100K 3950 thermistors. From what I have seen/been told I should wire them in parallel so it reads the highest temp coming from all four, and I will have to reduce the T parameter because the resistance will be lower. Is there anything else I should keep in mind?
If you have the option I would add a separate thermistor to the bed plate itself so that you could get a more accurate PID loop going for the bed surface temp. The thermistors from the pads themselves could still be used as you're suggesting and used as a thermal limit monitor on the heater pads.
@Phaedrux That's the plan long term, however in the meantime (waiting on parts that are on backorder) I was going to use the built in ones as a stopgap to be able to print until then.