Heated build plate not hot enough - I think.

Well, Ive been working with my REPRAP about a year now, and its great! I seem to have better results with ABS as far as quality goes.
However, even with a HBP, I still tend to get warpage on the larger/longer prints. My thermistor shows the HBP temp at 60 when its set at 100. In fact, no matter what I crank it to it still only will achieve around that temp. Now Ive read and seen guys cranking these up to 120c - why cant I?
Im using ramps 1.4 and sprinter. Is it a software thing or something else? The hotend heats up to temp just fine so I dont think its my power supply (200w ATX).
Any ideas?
Thanks!
Jeff

Not enough amps?

hmm - apparently im using a 165w atx and the 12v is rated at 10 amps. Really need more to drive that?

Ok,
replced that with a 300w atx that has 14a on the 12v line output. let my bed heat up for 10-15 mins and did a print. Still maxed out at 73 deg c.
My pronterface is set to 110 for the bed. so it almost seems like a software issue? The light on the bed never goes out so I imagine its still heating.
according to the wiki it only needs a 10 amp line. the bed doesnt feel like it has broken traces as its evenly hot.
NOW -
With that being said, I do have a 1/8 piece of glass over the bed and the thermistor is taped to the top of that on a corner. So maybe its hotter in the middle? But glass carries heat pretty good it should be relatively the same.
so - any ideas? Id like to be able to control this temp a bit better.
Thanks!
Jeff

Two things.

a. have you checked the temp with a meter with and without the glass?
b. If the temperature went up when you supplied more amps then maybe its a power issue?

no - I dont have a meter - yet.
also, it seemed to go up, but not that much.
So how much amps do you need to run this? the wiki only says 10. I KNOW theres guys out there getting the correct temp with modest power
supplies.
Jeff

It may be that your not actually getting the expected output from your Atx supplies. They can be a bit of a crapshoot, some require significant load on the 5v lines to get the rated output from the 12v line.
I'd put a meter on the output to see if your actually getting 12V and see what the current draw is.

I would buy something like this:

[www.ebay.co.uk]

Measure what Volts you are getting at the output of the ATX and at the bed, if you have a volt drop say only 10V at the bed then the cables supplying the bed are not big enough. most ATX supplies require a load ie about 5R power resistor on the 5V rail so that you can get the rated output on the 12V rail. You could also measure the R of the bed to see if it is too high.

I had trouble with my HBP taking ages to heat up. I tried chaging the cabling from the psu with no effect. In the end I swapped in a mark 2 PCB to replace the one I had and there was a immediate improvement.

>With that being said, I do have a 1/8 piece of glass over the bed and the thermistor is taped to the top of that on a >corner. So maybe its hotter in the middle? But glass carries heat pretty good it should be relatively the same.
>so - any ideas?
I found quite a difference in the temp between the centre and the edge of the PCB itself. Around 20 degrees IIRC.

Hi Jeff,

I assume that you use a Prusa designed PCB HBP. Heat is running out from the bottom side of HBP, so some heat insulator should be required if you wish to squeeze out the efficiency. According to my test results, good insulating material equalizes the heat spreading at the top side. Beside of balsa wood or cork sheet, I recommend you glass chopped strand mat (FRP layer material) as it's light, thin and cheap. Cut 20cm square, layered double, wrap up with aluminum coated thin film and put some Kapton tape.

I put this heat insulator mat together with 2mm cork sheet just under the PCB HBP, placed 3mm glass (weight 270g) on the top, and logged the heat increasing curves using four thermocouples. A DC power supply kept to feed 12V during the test.



In the following images, Green - bottom center, Black - midst between bottom center and edge, Red - top center, Blue - midst between top center and edge.

[etherpod.org]

In this case, I couldn't get 120 deg.C at glass top surface even if waited for 30 minutes.

Next, changed 3mm glass to 2mm aluminum plate (weight 210g). Temperature rising is obviously fast, better than 3mm glass.

[etherpod.org]

I hope these results become some reference.

Genie

Edited 3 time(s). Last edit at 05/28/2012 10:56PM by Genie.

I had the same problem, and ended up running the wiring different.

I put the negative lead to the heat bed into the negative on d9 on my ramps and connected the positive directly to the positive on the power inputs on the ramps.

I can now get up to 120, and havn't tested it higher but Im sure that it would get there if I tried it.

My bed would not get passed 80 so I flipped my Prusa designed PCB HBP. So instead of the trace being on the bottom of the pcb then the glass, it is between the pcb board and the glass. I now hit 100. Use some sugru to insulate the contacts on the pcb.

Do you not get a problem with the centre doming downwards when it gets hot? Normally it presses against the glass.

---

[www.hydraraptor.blogspot.com]

No bowing down of the pcb that I noticed....I would figure that if it would bow, it would bow downward more if the trace was on the under side at metal expands faster until the temperature evens out. At present I have with the trace between the glass and the pcb , I think it works nicely transferring the heat directly onto the glass and it does a nice job of maintaining heat. I have version 1 which does not have the center hole for the thermistor so I kapton taped it onto the bottom middle of the pcb.. So the reading is actually measuring thru the pcb and not directly by the trace.

Normally the hot side of a sheet expands more than the cold side, so you might expect it to bow the other way but what actually happens is the solid copper side expands more than the zigzag side. People that mount it trace side up have found they need to support the middle underneath.

---

[www.hydraraptor.blogspot.com]

Oh, we may be talking something different here...I have no solid copper side. I have a MK1 pcb with etched Cu trace (the zig zag) with the connection point and spot for an led/resistor if I desire but there is no solid Cu side. It's just has silkscreen writing. Now if you are talking about the acrylic bed part bowing, I have a silicone bake sheet as an temperature buffer between it and the pcb with the glass clamped to the pcb for rigidity but I haven't printed enough to see long term affects. I still trying to make the first layer stick and without curling...hit and miss...but my cube looks nice after a few mm up

The MK1 should have solid copper on the top under the solder resist. If it doesn't you only have half an MK1.

---

[www.hydraraptor.blogspot.com]

Ok, I think I need to examine my heated bed closer. So under the trace there should be solid Cu?

I checked my board an cannot see a solid Cu backing. From the [reprap.org] , it does not have the solid Cu as well so is there a newer design somewhere?. From the reprap pcb wiki pic link, it looks like I setup mine very different from their pic. I enclosed my side profile of my heatbed setup. From the top is glass then pcb with trace up, 1/4" spacer, silicone baking sheet, acrylic, springs/etc, ending with the last acrylic sheet. The assembly thru the web are very different when I started back in February and I ended up just assembly what I thought...and this makes leveling the bed difficult as I will need to epoxy the thru bolt top for the springs so I can adjust with out pliers and wrenches. I'm looking for metal M3 wing nut. My may hog the bed setup out to M4 as I have wing nuts at that size.

I can see it in this picture.

It doesn't go right to the edge of the board because it would not score properly, so I can't see if from your photographs.

Edited 2 time(s). Last edit at 08/01/2012 07:44PM by nophead.

---

[www.hydraraptor.blogspot.com]

Ok. had to take it apart to redo the spring setup. Yes, I see under the silkscreen which looks like a solid sheet of metal(Cu). Is this the part that heats up as well? I thought only the trace heats up thereby heating up everything else. So the solid Cu heats up first then the trace, ok, I see why it would bow on the side of the solid Cu. Once I rework the springs, I will look at so silicone mold to support the platform. I still like the trace touching the glass as I got better temps than with the silkscreen side touching the glass.

Only the traces heat up. The solid copper is there to spread the heat. The top surface is cooler because the heat has to be transferred through the fibreglass. However it seems to expand more because it is a solid sheet.

---

[www.hydraraptor.blogspot.com]

Ok. So if we need more heat we should have the trace up but we would need to have more support under the middle of the pcb to account for bowing due to a greater expansion of the solid metal surface.