Geeetech G2S Pro Delta 3D Printer - The Experience (2017)

[Hands0n]

each daughter board will handle 2 thermocouples I personally would convert the Bed to a know Thermistor such as http://www.ebay.co.uk/itm/Modular-Screw-on-M3-Stud-Metal-Thermistor-for-Reprap-Prusa-3D-Printer-Hot-End-/221892120007?hash=item33a9cd0dc7  or one of E3D's semitec 104 thermistors  there are lots of options there and stick with the Hotend Thermocouples

Ha!  I didn't think of only updating the Heat Bed to Termistor and only use a single daughter board for the incumbent hot end Thermocouples!     Brilliant!  Ta!       

That'll save a few quid...

Thinking that'll be the upgrade I'll go for first.  And a 24v upgrade if I can make it all work together - or at least a much higher wattage 12v supply than the one it shipped with.

[Dougal1957]

24 Volt all the way Danny board will handle it easily (I run mine on 24 Volt as well)  don't forget to do that you will need to swap out the hotend Heaters and the fans to 24V as well

The Board can handle upto 15Amps on the Bed heater circuit as well.

If you are thinking of swapping the Bed then a piece of Ecocast Ally with a silicon heater under it is the way to go.

[Hands0n]

Its been a little while.  I'm getting closer but still no Beer   

Heat Bed Faulty?
Well, perhaps not the heat bed itself, but thats where my problems currently reside.

Yes, yes, Doug, don't beat me up about 24v vs the incumbent 12v.  I've got rotorx doing enough of that    I get it, really I do, insufficient heat density at 12v and so 24v is the requirement.  I'm working on that.

So as every picture tells a story, here's mine



I have run into a mental brick wall about the Thermistor heat readings by the controller firmware and CAM software (both Repetier).  It was no different with Marlin by the way.  Something really strange going on.

At room temperature the heat bed sits around 20C (I run a warm house), certainly reflective of the ambient temperature in the room.  I have checked this against two other thermometers (one of which is in the video).   So far so good.

Yet when I apply power to the heat bed it starts to warm up and within about 3 minutes or so it reports the 110C that I have set it to.  Yet the surface of the borosilicate glass is only warm to touch.  Some time later, maybe 5 - 10 minutes, the surface of the glass steadies out around the high 70C (in this video it had been on for some 30 mins).  Typically the temp is some 30C below that being indicated.

I do not understand how the Thermistor can be reading 110C at the heating plate yet neither the aluminium plate or the glass plate get to that temperature, ever!

I've checked the Thermistor against another two and the resistance readings are the same ...

Next Steps
I've already got a 24v power supply, originally intended for my CNC upgrade but I've persisted with 12v so its a spare ...

I've ordered me one of these -->  https://goo.gl/CNX2l3  from China, half the price advertised here ...



I will use that to switch 24v to the heat bed in response to being driven by the original 12v from the main controller board. 



Happy days 

[Dougal1957]

Danny

Sounds very much like the Thermistor setting in the Firmware is probably wrong

It is probably a std Chinese thermistor with a B value 0f 3950 at 25/50.

Try getting a semitec from e3d online and setting the firmware to suit.

I find that there is normally about a 10 degree difference across Glass

HTH

[Hands0n]

Hey Doug, thanks for the info.

My first thought was Thermistor ... so I grabbed hold of a pair of these NTC 3950 Thermistor 100K 1% :  http://www.ebay.co.uk/itm/182110843706?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

They and the original Thermistor provide identical resistance readings for a given temperature.  Which would indicate that the one supplied by Geeetech is okay.

I did make a change to the firmware to use the NTC thermistor table (11 if I recall correctly, it was late) rather than the default generic table - but that has made no appreciable difference either.

I've got a few articles to read when I get home, check against the tables in the firmware source code i'm using.

Otherwise will grab one of them thar Thermistors from e3d you suggest

They're selling loads of these Delta ... there's going to be a lot of annoyed/desparate people out there ... Not me.

[Dougal1957]

Most thermistor are quoted as 100K at 25 degrees and the beta value equates to what you get (usually quoted as at 100 degrees) however those 3950 ones are at 25 and 50 degrees respectively and it is that that throws them out.

Finding the correct table is a pain in the bum but I do have the full spec of one if you want to write your own table?

Doug

[Hands0n]

Now thats interesting.  The one I chose is supposed to be 100K @ 25C - which I have to say was pretty close given the ambient temp where I measured all three.

If you don't mind sharing the spec I'd be delighted.  For sure, all that I'm finding in terms of replacement Thermistor doesn't give very much information at all.  My only thought regarding the one in the link above is that it is sold for use with RepRap from a 3D print supply outfit.  Ever the optimist, me. 

Some reading I've been speed-reading (dig in tonight).

https://nutz95.wordpress.com/2014/04/13/marlin-firmware-thermistor-3950-table/

https://cae2100.wordpress.com/2014/03/08/beta3950-thermistor-table-for-marlin/

http://microfabricator.com/articles/view/id/531ad7e59aad9d3131000000/beta-3950-thermistor-table-for-marlin

[Dougal1957]

Sent via E-Mail danny

[Hands0n]

Thanks Doug, email received and will be digested and worked on.  I think this is entirely crackable now I know what is the likely root of the temperature display issue.

So tonight's first exercise was to see if I can drive the surface of the heat bed to 100C.  Short answer is Yes, I can.

On the basis that the displayed temperature is some 30C short of the actual I tweaked the firmware to allow me to set the bed temperature to anything up to 150C.

The results speak for themselves ...



A test print I did (20mm cube) has stuck to the heat bed like the proverbial brown stuff to a blanket. This is what I needed.  I'd have to break the print to get it off the bed right now as it is cooling down - it'll pop off the bed by itself when cold.

Thermistor Temperature Tables
Thank you for the data. Using this I should be able to construct a suitable table for this printer and load it into Repetier Firmware and compile it up.    I'll chronicle the results on here

I feel a corner has been turned.  W00t!

... And then ...
Once the solid state relay arrives I'll convert the heat bed to 24v so that it heats up quicker than the 20 minutes it just took 

[Dougal1957]

Which SSR have you ordered Danny?

[Hands0n]

Hi Doug - I ordered this one specifically, it is the same as mentioned above, just cheaper ... £4.59 delivered.

http://www.ebay.co.uk/itm/182402240902?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

(click image for full size)

[Dougal1957]

At least it is a mosfet and not one of those Fotek DC-DC SSR modules you see on Ebay for a few quid but they drop around 1.6 volts across the output.

That was the reason for the Question>

Hopefully that one will do the job?

Doug

[Hands0n]

Yeah - I'm trying not to convert the entire machine to 24v with all that might entail.  That may come about in due time but not just yet awhile 

Tonight I shall mostly be bending my brain working out the Thermistor Table requirements for the heat bed.  Of course Repetier does it slightly different to Marlin ... Deep joy, many fun. 

[Dougal1957]

Yeah - I'm trying not to convert the entire machine to 24v with all that might entail.  That may come about in due time but not just yet awhile 

Tonight I shall mostly be bending my brain working out the Thermistor Table requirements for the heat bed.  Of course Repetier does it slightly different to Marlin ... Deep joy, many fun. 

The spread sheet I sent allows you to adjust the Beta Value to get somewhere closer to a correct reading at the upper end and I think I got mine to be around 4008 when I had a 3950 but I swapped it out for an E3D Semitec one?

Doug

[Hands0n]

Hi Doug - I've been looking at the spreadsheet today and think doing so at work was not one of my greatest ideas   Got a headache now 

Are the output values going to be Marlin, Repetier or does it simply not matter?

[Hands0n]

Temperature Reading and Control Solved.

Thanks to Doug for putting me on the right steer, I've resolved the issue of the G2S Controller reading the wrong temperature of the Heat Bed - optimistic by 30C across the range.

You'll recall that I am using Repetier Firmware in the Controller and that makes for an easy way of configuring the software (Configuration.h) and submitting it to Arduino to compile and upload into the controller.

After lots of research and pouring over numerous datasheets for the NTC 100K 25C Thermistors I finally settled on using the Generic Thermistor Table in Configuration.h

Iteratively I set the B Constant value starting at 3950 for the thermistor used by Geeetech which made the problem worse.  I then upped the values by 100 or so at a time, finally settling on 4900.  I may just wind it back to 4850 as it does slightly under-read by about 3C

So finally I have regained control of the temperature of the Heat Bed in the controller.

The lesson here is that not all NTC 100K 25C resistors are equivalent.  Not by a long chalk.

In the pictures that follow I have taken the temperature at the metal surface, at the rim and then on the glass surface at the centre.