Heat Bed comparison

Hi Guys
So i just built my prusa I3v and I am having trouble with prints warping when printing, the sides bend up. I hooked up my heat bed directly to my 12v 360 w power supply and it took about 20 minutes for the center of the bed to heat up. There was about 10 degree changes from the center to the side locations. I have 18 gauge wires running from Ramps 1.4 to the heat bed right now. I was thinking of upgrading those to some 12 gauge wire. I did notice that the bend i have MK1 has very tiny pads to solder to, will the larger wire still make a difference?

A chain is only as strong as its weakest link. 18 gauge should be enough, 12 gauge is overkill, but maybe 16 is worth a try. In my experience, the solder connection is usually the trouble point. Be generous with the solder, and if it a plated through hole, solder both sides. There's only so much surface area there - you want to take advantage of all of it - including the inside and back of the hole.

or you could run 6 wires to heated bed,3 for pos, and 3 for neg?

Edited 1 time(s). Last edit at 05/07/2014 03:30PM by jamesdanielv.

Per [en.wikipedia.org], moving from 18 to 16 gives you about double the copper, would halve the 21 mOhm/meter resistance, and about halve the ~ 2W ( 10A^2*0.02ohm) of power lost in the wires. To 12 gauge, you could save 90% of the lost through a meter of 18AWG, so it should get you about a 2% improvement in power delivered to the bed.

I'd look at the existing solder joints, maybe scraping back some of the insulation on the board per [reprap.org] , insulating underneath the platform and maybe putting a cozy on the top for part of the 20 minutes.

Does the heating pattern even out after some time?

I noticed that nophead puts the thermistor off-center to maybe get a more representative measurement.

I agree with DaveX that you may just need some insulation.

I ran some tests recently and found that my mk1 heated bed took around 20 minutes to reach 112 C when fully uncovered, which was a few degrees shy of the 115 I was aiming for. Placing a silicon insulator on top (designed for putting pots on) allowed me to reach 115 C in 15 minutes.

I also tested a Panucatt Helios with the same setup (excellent bed, but much pricier). It was able to reach temperatures a few minutes quicker than the mk1 - but it retains heat much better. I have since ditched the silicon pad for a piece of corrugated cardboard with an aluminium foil reflector. The board has a raised rim around the bottom so there is only contact with the bed along the edges. Underneath the bed I have covered the sides of my Y-carriage to trap air, and placed another reflector facing up. My current setup takes 10 minutes to reach 115 C.

---

[blog.bodgycorp.com]

I too endorse the Panucatt Helios - an excellent heated bed. I disagree with the "much pricier" label. The helios can be purchased from panucatt directly for $40. I find the MKII from comparable vendors for as much as $45. I know you can find it cheaper on places like ebay, but for me they're basically at the same price point and the helios blows away the MKII. It's a no brainer.

The Helios shipping to Australia was $29.85 - I was just lucky to be able to consolidate a few items with a shipping company to make it affordable.

And yes, the Helios is a quality piece of kit.

---

[blog.bodgycorp.com]

What's the measured resistance of the board? And what temperature are you trying to heat to that takes 20 minutes?

Resistance of the board is about 100k at room temperature. Trying to heat it to 110 and keep it there during the print. Would it be ok to run 14v into the ramps just for the heat bed connections, or would that fry the board?

Quote
wolffman122
Resistance of the board is about 100k at room temperature. Trying to heat it to 110 and keep it there during the print. Would it be ok to run 14v into the ramps just for the heat bed connections, or would that fry the board?

Resistance of a 100K thermistor is about 100k ohms at room temperature. Resistance of a 12V PCB heatbed is about 1.1 ohm so it can dissipate about 130 watts per P = V^2/R. (14V)^2/1.1 = 178 watts, so that would be about 36% more power, but it could put 14V/1.1ohm=12.7 amps through your wires.

RAMPS can take 14V just fine, but in some cases the MOSFETS (http://reprap.org/wiki/RAMPS_1.4#Q1.2C_Q2.2C_Q3_-_Mosfets ) and fuses sound marginal. If you have a good low resistance MOSFET, you can switch 24V at twice the amps with less of a cooling problem than you'd have with the the spec'd 18mOhm rDS-on STP55NF06L.

Edited 1 time(s). Last edit at 05/13/2014 11:06AM by DaveX.