Loving the ballscrew conversion! Part numbers? I probably missed that thread somewhere.

Here it is

http://www.soliforum.com/topic/2998/sd3 … all-screw/

I've been away for the weekend, so just catching up. Will post details for you re my setup later on.

I double-faced taped the fiber insulation cloth that was on the original heater pad, and wrapped everything in Kapton tape.  Then I clipped my glass surface on top.

The temp build up from 23°C to 95°C (according to Repetier) takes about 3 1/2 minutes. at that time the top of the glass read and average of 80°C.  After another 5 minutes (about 8 minutes in all), the top of the glass reads about 89-90°C, and it seems to stay there.  I have to actually crank the Repetier setting to 105° to get the bed to read 95°C.

I also took a couple of readings at the bottom of the print bed, now that I insulated it, and it seems to average 102°-106°C, so it looks like (some/most?) of the heat is getting trapped inside rather than radiated outward.

Nevertheless, I ran a couple of test print at the 95°C (Repetier) temp and the prints seem to stick quite well, so I'll leave it at that.  Never having done any temperature readings of the original bed setup, I have no reference of how hot the top of the glass used to get before, so for all I know it's hotter now than ever before.

I also printed a little switch holder for my power switch, and glued it in the lower left hand corner.  So at this moment it looks like the machine is working fine, but I'll probably need to get a replacement print bed frame one of these days.  Look how twisted the wood bed frame is on the front view.  The front right corner droops down about 1/4".  It's been like this from day one.  If you put it on a flat surface it just rocks back and forth, it doesn't sit flat.

Quality control at Soliddodle is not an exacting science, I'd say it's more like "close enough"

The reason is I just found out that autodesk gives away free student copies so I've been modeling some parts from my sd2 to try and learn inventor (and also play with simulation CFD which I know nothing about, but seems pretty cool). I added some files to the wiki for the extruder and the heatbeds even though I know no one uses the glitchy wiki.

In my case the "bottom" of the heater pad is attached to the bottom of the aluminum bed, and the readings are taken on the opposite side.  My pad heats up the aluminum bed on one side, and radiates the heat on the other.  I plan on putting the original insulation cloth back so that should help trap some of the heat instead of radiating it outward.

Once I get the bed fully installed, the glass will also go back on top, which should help even out the readings across the entire surface.

Anyway, I have a few question regarding your setup.  Did you use a QU-BD heater pad?  If so, I assume your pad has the thermistor near the top edge where the wires connect to the coils.  How accurate did you find the thermistor readings to be compared to the readings from your FLIR camera?

First of all, while the thermistor reading got up to 95°C in a little over 3 minutes, the top of the printbed never actually reached those temps, it actually topped out at about 91°C in about 6-7 minutes.  Additionally, while the readings around the center of the bed were around 90-91°C, the edges (about 1" in) read between 86-88°C.  I also took a few readings of the bottom of the heater pad, and the average was 113°C.

I cranked up Repetier to 100°C, but the actual bed temps only went up to 90°C around the edges and 92-93 in the middle.

With glass plate ontop:

With ceramic tile, where you'll note the HS1 mark is almost centered...

So yes, weird that you got readings completely different to multiples tests with multiple materials done over at http://www.soliforum.com/topic/2230/the … ed-setups/

Anyways - guess we'll have to agree to disagree here.

The following readings and findings seem to go against the conventional wisdom as it pertains to the installation of the thermistor, at least for a QU-BD heater pad.

I heated the pad to 70°C, and read the temps in a few different place, including the bottom of the pad, and top of the print bed. 

Take a look at the following 2 images.  The one with the high reading is taken at the spot where the QU-BD thermistor is, the other is in the center of the pad, where I placed the original thermistor.  Notice the difference between these reading which were taken a couple of seconds apart.  I also flipped the bed and took a reading on top, but forgot to snap a picture of it, the top read about 67.5°C.

The QU-BD thermistor is built into the pad near the place where the wires are soldered to the coils.  For this reason, the readings are much higher than the rest of the pad.  Not only that, but the temps build up a lot faster in that area compared to the rest of the pad, giving the false impression that the pad heats up super fast, when in fact it takes a lot longer for the entire pad to heat up evenly.

So, at least for my setup (and I think this would probably apply to anyone upgrading the original heater to a QU-BD model), reusing the original thermistor instead of the one built-into the QU-BD heater, is in my opinion a more accurate solution (sorry, Adrian).

And yes, I assumed (erroneously) you were putting it on top, as having the same sized pad as the build area means you'll foul the levelling screws. I see instead you got a 150 pad.. and thus its within the screws footprint. My bad. Most others have gone with the equal build size and gone on top in order to gain a directly heated larger surface area instead of relying on the aluminium to spread the heat..

As for feeling strongly - I've been more miffed by the desire to not just go with the 'as built' solution. And yes, I perhaps came across a bit strongly, and by no means assume you where no idiot with a soldering iron - that was what compunded my confusion in trying to (apparently) avoid a measly one line mod when you've done all the other work... it seemed entirely non-intuitive to me.

All to often lately people have been getting very , shall we say, cavalier, about thermistors. My attempt at highly discouraging yourself here is borne from having seen quite a few others passing it off as 'ok' or even, 'good'... so apologies if thats come out in these posts as being overly assertive... but I'm trying to discourage the perception that A: firmware mods are very very hard and therefore not worth bothering with and B: thermistors, meh, close enough's good enough right ? ....

So my apologies.

It will 'work'... it will even report a tempreature, and you'll be able to type 95°C and it'll 'report' 95°C...

Its just that the heated pad will be MUCH higher than 95... Because you have placed the thermistor so far away and on the otherside of a metal plate.... there will be anywhere up to 20°C difference between the top side of the QU-BD and the bottom side of the aluminium plate.

Are you thinking I'm putting the bed on top of the aluminum plate and the thermistor at the bottom? 

I'm not, I have it setup the same way as the original pad, at the bottom, with the thermistor sandwitched between the pad and the bottom of the plate

Not on my machine.  The thermistor was mounted dead center in the bed.  There's even a drilled n indentation at the bottom for the thermistor head to sit flush with the surface, it's just that they didn't route a groove for the wires, but they're fairly thin.

Wow...I can see you feel quite strongly about this.  I never said I won't do the firmware change, only that I wanted to try this setup .  I don't think I'm cutting any corners, it's just a different approach.  If I find the readings to be erroneous (after I check them with a heat gun), then I'll do the change.

I have no problems with firmware changes, but man, cut me some slack, I'm not and idiot with a soldering gun....

Check out the temp curve.  From 25°C to 96°C in under 4 minutes.  Not too bad, although I was kinda hoping for even faster performance based on the specs from QU-BD, which claims this pad can heat to 130°C in 1 minute.  The power supply is certainly capable of delivering the amps.

On the other hand, I don't have the temp gun here, so I don't know if the reading is correct.  It certainly feels hot, in fact it feels quite a bit hotter than it did with the old pad at the same temp readings.  I'll borrow a temp gun from work and check it tomorrow.

The only complaint I have so far is the PS fan noise.  I need to do something about it...

In the last image you can see the 150mm x 150mm QU-BD heating pad.  I'll first use the original thermistor, and if that works, I'll keep that setup, if not I'll use the thermistor that came with the pad and modify the firmware.

It will 'work'... it will even report a tempreature, and you'll be able to type 95°C and it'll 'report' 95°C...

Its just that the heated pad will be MUCH higher than 95... Because you have placed the thermistor so far away and on the otherside of a metal plate.... there will be anywhere up to 20°C difference between the top side of the QU-BD and the bottom side of the aluminium plate.

Seriously - you are cutting a corner that will end in a huge troubleshooting headache... Not to mention it will cause you QU-BD to run much hotter, and harder, than it should.. and will introduce a massively wide PID cycle as the distance the tempreature has to travel from where its created to where its measured is so much larger than it should be and thus the on/off cycles will be increased in duration... longer to (apprently) heatup and longer to (apprently) cooldown... 

Its just like on the stock SD Extruder, why you have to 'add 30°C' to the reported tempreature to know what the Actual Meltzone tempreature is.  That 30°C Delta is from a realtively short distance of brass, that transmits heat quite nicely. Do you know what the delta is between the top side of your QU-BD and the bottom side of the aluminium plate and the *side* (the standard thermistor is side mounted not center mounted) of it ? Its going to probably be a lot more than 30°C and you're the only one who can work out what that temp delta is because you don't have the benefit of the research that gave you that 30°C figure in the first place You're essentially suggesting that its 'ok' to set the Tempreature on an Airconditioner to control a room to a precise tempreature by employing a thermostat down the corridor and in an entirely different room

So unless you want the center of your bed running at 120°C, the sides at 105°C and the 'reported' tempreature to be 90°C... (which it will duly report) then deal with the fact you have to make a tiny change in firmware that 100's of people have made already to "great success" (to quote Bruno...). Or accept that you are cutting a serious corner that you really haven't examined the ramifications of and feel free to disregard the advice offered and go along your merry way  

But I just can't get my head around doing all this effort to skimp on one line in firmware... that will give you nothing but negatives in return ?

To clear the wiring coming out of the bottom hole, I used a 1/4" plate under the power supply to offset it from the surface.  All the wiring is done, but I need to make some type of holder for the power switch.  The switch is 1/2" diameter, and I don't want to drill a 1/2" hole in the printer frame, and undermine the rigidity , so I'll simply print a power switch enclosure.

In the last image you can see the 150mm x 150mm QU-BD heating pad.  I'll first use the original thermistor, and if that works, I'll keep that setup, if not I'll use the thermistor that came with the pad and modify the firmware.

I liked the Astec a lot, as it's physically smaller, but I didn't like that the fans are running all the time the AC cord is plugged in.  Even though the power switch is off, and the PS is not outputting any power, the fans are still running, and they're not exactly quiet (it's a server after all)

So I decided to go with the Delta.  Its fans only come on when the power is turned on, so once you turn it off, it's quiet.

Anyway, I will post a few pics as I go along with this rebuild.