Heatbed driving me crazy, Silicon replacement?

Hi all
So the heatbed on my i3 is driving me insane, its a mk2a running from a dedicated 12v 30amp psu with mains power rated wire and it takes forever to heat up (at least 20 minutes from cold to hit 110 degrees C, cant get it to reach 120 degrees) Ive tried insulating it but doesnt make much difference, so either theres something ive missed or possibly i have a faulty heatbed.

So i started looking around for a replacement and came across this [replicatorwarehouse.com]
Ive never seen one like this but they claim if heats up incredibly fast, so i thought id through it out there to you pro's and see if any of you have heard or used this type of heatbed, thanks

Your existing heat bed should heat up in about 5 to 8 minutes. If it does not, something is wrong.

1) If it's mounted directly to an aluminum base plate, it needs a thermal break. Cork, wood, and nylon spacers all work. Having some sort of cover behind the board does help.

2) Check the voltage out of your power supply with a volt meter. It may be sagging under load. Power = resistance times volts squared. A small voltage drop is a big deal.

3) Check the voltage drop on your wires and connections at the same time. Same issue there.

4) Connect the heat bed directly to the power supply and see how fast it heats. You may have a firmware issue in your control board.

5) If you have a good ohm meter, check the bed resistance. It should be about 1 ohm. Alternately you can check the current drain at 12V. It should pull 10 to 12 amps when directly connected.

6) check the wiring to the bed. If it's a 12V / 24V version, you may have it wired for 24V. (should be a 2b if it is dual voltage, but who knows...).

7) Check your thermistor type and the setting in firmware. If you have the wrong thermistor set it could be getting hotter than you think.

8) Check the thermistor wiring and it's attach to the board (same issue as 7).

It's much easier to fix what you have than to replace it.

@uncle_bob

My hot bed takes 25mins to get to 115 deg. I presumed this is normal.

60 deg for PLA happens in about 5 mins, but for ABS it takes ages.

I presumed this is normal for a 12V PCB heat bed.

I have 5mm glass on top of my heat bed, and insulation under the PCB

20 minutes is not unusual - it probably depends how well its insulated from its undercarriage.

---

_______________________________________
Waitaki 3D Printer

If you have a fan on the hot end, make sure it does NOT blow on the heat bed. Doubles the heat up time for me. Adding a sheet of paper on top of the heat bed while heating up also helps quite a bit.

TG

PS: Power = Voltage squared / Resistance

The fan on my hot end pretty much blows on the heated bed. The air runs into the back of the X carriage and then goes straight down. I have not seen any issue with getting mine to 110C in 7 minutes with a very conventional 12V supply. My heated bed measures right at 1 ohm (as it should). The bed sits maybe 5 mm above a piece of plywood which acts as a back side insulator. The glass on top is 2.5 mm. I have not seen any noticeable difference heating it with or without glass. Based on all the "heated bed never heats" threads, I watch the heating time pretty closely each time I fire up the printer.

uncle_bob Wrote:
-------------------------------------------------------
> My
> heated bed measures right at 1 ohm (as it should).

This is likely the issue. There is a wide variance in resistances for these heated beds, especially the cheaper ones. I've seen ones higher than 2 Ohms before, which would explain the excessive warming times.

---

Help improve the RepRap wiki! Just click "Edit" in the top-right corner of the page and start typing. Anyone can edit the wiki!

Twice the resistance would give you half the power. That would at least double the heat up times.

If you go into the design specs on the heat beds, they are very clear. The board should be made with rolled copper (not plated / electroless copper). The board shoud come out at 1.0 to 1.2 ohm. The people who sell the 2 ohm boards are selling scrap / junk parts. They have half the coppeer on them that they are specified / designed to have. They aren't just high resistance, they also likely have hot spots. This is all directly from Prusa's posted notes on the design. It's also very rational from a pcb design standpoint.

That said, there's no reason why a silicone replacement woud not have the same issues. If you have a bad board, buy one from a reputable supplier who will stand behind their product. The good ones are only in the $20 to $30 (US) range.

Thanks @uncle_bob

I just measured the resistance of my heat bed, and including the wiring (short wires), its reading 3.2 Ohms at the RAMPS board.

So it looks like my kit was shipped with one of those cheap heat beds. I'll order another from one of the suppliers in the Wiki and see if its any better ;-)

If you are at 3 ohms, you don't have anything even close to the proper design on the board. I suspect that you have a plated board that got very poorly done. Unless it's a very unusual process, you are likely to have significant hot spots on that board.

Hi uncle_bob

I bought my kit from Blomker.com, and on the whole the components seem to be good quality, however in this case they have have supplied a cheap or possibly even partially defective heat bed.

I have a sheet of 5mm glass on the bed, which is probably evening out the hot spots due to its thermal mass.

I've just ordered 2 x new heat beds from one of the suppliers listed in the Wiki (http://www.ebay.com.au/itm/RepRap-MK2b-Dual-Power-PCB-HeatBed-RepRap-Mendel-Prusa-and-other-3d-printers-/171039960983 )
I've order the dual power version in case I want to upgrade to 24V in the future.
(I'm building a second reprap, which is why I need an additional heat bed ;-)

Hi uncle_bob
Thanks for your suggestions, heres what i have found

i ran the heatbed to 120 degrees C and it took:
18 minutes to reach 115 degrees C
24 minutes to reach 120 degrees C

I checked the temp with my multimeter and the reading was within 3 degrees of what the printer was reporting so thats pretty close

I hooked the heatbed directly to the power supply and ran it again and it gave:
15 minutes to reach 115 degrees C
21 minutes to reach 120 degrees C

As you can see it heated up quicker but still very slow

i had the probe at the center of the heatbed and when it reached 120 degrees C and moved the probe closer to the edge and the temperature dropped to 113 degrees C, not sure if this is normal but thought id mention it

The board is definitely the mk2a version that is spaced approx 10mm above the Y carriage

i also measured the voltage, the readings were:
12.00v on the terminals on the power supply
11.33v on the ramps connectors for the heatbed
11.11v on the connectors on the heatbed

and lastly i measured the resistance of the heatbed and it came out at 1.4 ohms although it seemed to jump around so not sure what was going on there

any suggestions?

kind regards
Chris

UPDATE:
So after checking heating times, resistance, voltage etc and not really getting anywhere i decided to remove the heatbed from the printer and wanted to test it when it was completely insulated. Now i didnt have anything to insulate it with so i got 2 kitchen towels and folded them in half to give 4 layers of towel and ran the heatbed upto 120 degrees C as i did before, here are the results (in minutes):
115 degrees C in 11.20
120 degrees C in 13.40

as you can see in comparison to when the heatbed was attached there is a massive difference in heating times (24 minutes when attached to the printer, 13 minutes when insulated)

So it would appear as though the frame is sucking all the heat out of the heatbed, even though there is a 10mm gap between the bed and the printer, so im gonna buy some cork and see if i can insulate it

Definitely insulate under the bed.

Both wool and cotton are OK as insulators, both have ignition points well above 120 deg (according to the sources I've seen online)
BTW. "Oiled cotton" has a much lower ignition temp, but I suspect this is mainly the ignition point of the oil).

I'm not sure if it helps, but I've also used aluminium cooking foil under my bed for insulation (as its supposed to reflect radiant IR).

I currently have a sandwich of aluminium between to sheets of cotton wadding. So if you can sacrifice 1 of your kitchen towels, you'd achieve the same results, but putting the aluminium foil between a folder tea towel.

I'm not sure of the relative merits of cork.

There are also plenty of references to people using corrugated cardboard from boxes.

Edited 1 time(s). Last edit at 10/29/2013 08:57PM by rogerclark.

Rather than measuring resistance it's perfectly rational to measure current. If you *know* your power supply is putting out 12V, the current will tell you the resistance. Hook the bed directly to the power supply and see what it reads. If it's 1 ohm you get 12 amps if it's 1.2 ohms you get 10 amps. Once you know the bed is right, hook it to the Ramps. Leave the heater off and hook your amp meter (DVM) between the return wire on the heated bed and ground. You should still get 10 to 12 amps. If not, check wires and connectors.

-----

You have bolts at the edge of the heated bed. There *should* be a gap between the copper on the board and the holes. The holes should not be plated through. They should be rough not shiny. The gap is there to provide insulation. Don't let springs or washers bridge the gap.

Other than that, my heated bed is no different than any other. They are very simple beasts. Put X power into them, don't put a thermal short on them, and they heat up. The most obvious problems all revolve around not getting enough power into them. If you want to rev one up, insulate the back side with something good. That should let you heat up in about 3 or 4 minutes. Again, if not, something is wrong and it should be fixed.

-----

Putting the bed inside towels kills convection. You have a *lot* of surface area on a heated bed. It's horizontal, but it still counts for convection. I'm not at all surprised that the bed heats faster wrapped in towels. It should heat to 110 C in about 2 minutes wrapped in towels if all is well. If it's taking 11 minutes, something is wrong. Check the voltage and current.

------

If you have a DVM with a limited scale on DC amps the easy solution is to use a shunt. You can buy one for big bucks (maybe $10) or you can make one:

Take a chunk of #18 AWG wire (1mm wire if you are somewhere other than the US). Measure out about 1 meters / 3 feet. The main wire is your shunt. Attach two leads to your shunt to run over to the DVM.

The wire is 6.385 miliohms per foot. If you have about 3 feet that's going to be about 19 miliohms. Put 10 amps through it and you get 190 mv on your meter. Is it perfect - no. Is it plenty good enough to troubleshoot a heated bed - you bet. Is it cheap enough to leave connected to your printer all the time? Just how cheap are you.... ....

You can get pretty good with some careful measurements. The wire does have a temperature coefficient and at 20 amps that will be an issue. At the 10 A level you should be ok. If you want more current, use bigger wire and adjust accordingly.

@uncle_bob

I have various measuring devices including a current clamp that reads DC. I just checked how much current was being drawn, and its around 6A. @ 12.3V.

Where as if the board was 1Ohm, I'd expect between 10 and 12 A depending on the resistance of the wires from the RAMPS to the bed and from the PSU to the RAMPS.

I'm pretty sure my pcb heat bed is not a good quality one.

If you use a clamp on DC meter, be sure to zero it first. Often they are set up for some monster current (like 200A) and the zero can wander a couple of amps.

-------------

If you have half the current, you have half the power. That's not a good thing. The current plus voltage measurement is pretty definitive. You can mess up measuring one ohm with normal gear, but current and voltage are not that crazy. I'd get in touch with who ever sold you the board. If they are good people, they should replace it.

Does the current fluctuate ?

6 amps seems not that much for when its first heating up, my setup is different to yours but i draw like 20 amps to start with, unless of course its a small bed?
12 volt right?
id be trying to increase the current

The point of the current measuring exercise is to check the power going into the headed bed. If your bed (or wiring) is high resistance, you will not get enough power into it. A properly manufactured bed will pull about 11 amps at 12V. If the wiring is ok, and your bed pulls 6A at 12V, the bed has been manufactured wrong. There are some simple issues with pcboard manufacture that could easily explain the defective boards. Quick summary - it's a *lot* cheaper to make them the wrong way than to make them the right way.

Edited 1 time(s). Last edit at 10/31/2013 12:43PM by uncle_bob.