Replicator 2 HBP Thermistor Wiring

2,665 views
Skip to first unread message

John McKisson

unread,
Feb 26, 2013, 8:58:12 PM2/26/13
to make...@googlegroups.com
I've been assembling a glass HBP for my Replicator 2 and have a question about the wiring of the thermistor.

My setup uses a 120V silicon heater pad switched thru a solid state relay via 24V control from the
MightyBoard, and allowing the onboard PID to control temp thru the HBP thermistor input.

I understand from looking at assembly instructions for other bots that there may be a voltage divider
involved on my side of the cgrid connector, but what exactly is this circuit?  All the links from other
pages to MBI wiki supposedly describing this are 404'd.

Would appreciate any help!

Dan Newman

unread,
Feb 26, 2013, 10:34:14 PM2/26/13
to make...@googlegroups.com
At thingiverse, look at the schematic for, say, the Rep 1's heated build
platform (requires Eagle),

http://www.thingiverse.com/thing:16061

You will want a 100K NTC thermistor with a beta(25/85) of around 4066.
(The thermistor voltage to temperature conversion tables in the MBI
code are set up for such a thermistor. Oh, and with a resistor of
the resistance shown in the above schematic.)

Basically, you have a resitor sitting between
Vref and Vout which is there to help make the
thermistor measurements be most linear in the
temperature range of interest. The thermistor
itself sits between Vout and Ground. Sometimes
there's a resistor put in parallel with the
thermistor so as to limit the self-heating
experienced by the thermistor. However, it also
reduces the A/D resolution. The MBI boards do
not have this parallel resistor.

Dan

Eric

unread,
Feb 27, 2013, 9:58:34 AM2/27/13
to make...@googlegroups.com

I would be curious to know how you make out with this; i've been debating adding a HBP as well. 
Message has been deleted

John McKisson

unread,
Feb 27, 2013, 10:21:25 AM2/27/13
to make...@googlegroups.com
Dan

It is a bit interesting to try to follow these discusssions given the lack of specificity. 

On the Replicator 2 the thermistor circuit includes a 4-contact connector and an existing 100K resistor. What remains unclear from your description is what the connection for the thermistor must be to present a complete circuit that the Replicator 2 expects.

In the schematic, there is no reference to Vref. Do you mean to say the 5V on pin 1?
Also the A/D connection to the processor is named HBP-THERM (pin 1), is this what you refer to as Vout?
You don't specify the value of the "resistor sitting between Vref and Vout" - is is possible that you mean the 100K resistor that is connected from 5V to HBP_THERM? (Pins 1 and 4 on the connector?)

Would it be accurate to say the following?

The thermistor for the heated build plate is connected as the lower resistor in a two-resistor voltage divider. The upper resistor (100K) is already present in the Replicator 2 circuit and connected between pins 1 (+5V) and pin 4 (the input to the A/D. The thermistor should then be connected from pin 4 to ground (either of pins 2 or 3).



As a point of reference it is the resistor often placed in parallel to the thermistor that linearizes its response in a limited temperature range, not the upper resistor in the divider. The parallel resistor is typically chosen to equal the value of the thermistor at the temperature of interest, and is placed in parallel with the thermistor. This parallel connection also *halves* the nominal value of that parallel pair at the temperature of interest, so that the other resistor in the divider must also be changed. The need for linearization depends entirely on the approach used in software to convert the output of the voltage divider to an engineering value (temperature) It is possible to use a non-linear lookup table to correct for the nonlinearities. 

John McKisson

unread,
Feb 27, 2013, 10:24:06 AM2/27/13
to make...@googlegroups.com
When I get everything working I'd be glad to post a few pictures along with diagrams and a BOM.

Clinton Hoines

unread,
Feb 27, 2013, 10:25:27 AM2/27/13
to make...@googlegroups.com
Better to use one of these if you want to be able to use your max print area.
 
Clinton

On Wednesday, 27 February 2013 08:17:06 UTC-7, TheMakerGuy wrote:
Might you not be able to squeeze one of these in?
 
Also this is open source so you should be able to find the wiring
 
On Wednesday, February 27, 2013 9:58:34 AM UTC-5, Eric wrote:

Dan Newman

unread,
Feb 27, 2013, 4:08:02 PM2/27/13
to make...@googlegroups.com
John McKisson wrote:
> Dan
>
> It is a bit interesting to try to follow these discusssions given the
> lack of specificity.
>
> On the Replicator 2 the thermistor circuit includes a 4-contact
> connector and an existing 100K resistor. What remains unclear from
> your description is what the connection for the thermistor must be to
> present a complete circuit that the Replicator 2 expects.
> In the schematic, there is no reference to Vref. Do you mean to say
> the 5V on pin 1?
I was talking in general, not specifics. You will need to pull your
RevG mightyboard and trace the
circuits yourself as well as consult the firmware to see what exactly
you want to hook to. (A
downside to the board not having publically available schematics.)
For both
the Rep 1 and 2, the firmware uses one of the AVR's onboard a/d pins to
read the thermistor voltage. So,
the thermistor should be tied in to the appropriate analog voltage
pins. Since the Rep 2 firmware
uses the same code as the Rep 1 firmware for reading the thermistor,
the circuits should be
largely the same modulo which AVR analog pin to tie into. (They are
different AVR pins
between the rep 1 and 2. See PLATFORM_PIN in the Configuration.hh files
for the two
bots.)


> As a point of reference it is the resistor often placed in parallel to
> the thermistor that linearizes its response in a limited temperature
> range, not the upper resistor in the divider.
Yes, sorry I had things backwards. The parallel resistor is for
linearization.

Dan

John McKisson

unread,
Feb 28, 2013, 10:37:37 AM2/28/13
to make...@googlegroups.com
I've completed assembly of the HBP system, here are some pictures of the assembly.  I obtained most of the parts from McMaster Carr and/or Newark.

With regards to the assembly of the thermistor, the MightyBoard already supplies a 100k resistor on-board so connecting the thermistor from Pin 4 to GND (on the HBP Thermistor cgrid header) immediately worked.  Calibration was spot on when tested with another thermocouple.

The insulated plate is G10 fiberglass with holes machined to fit the plate leveling screws.  1/8" aluminum stock was cut to fit the silicon heater (6"x10) to act as a heat spreader under the Pyrex glass.  The glass was cut to 10" from 12" stock from MMC.

We added a cutout on the side of the aluminum plate to accommodate for the thicker area of the heater, as well as a hole machined in the center where the thermistor is pushed up to contact the glass.  Copper plating is screwed into the side of this plate for a place to solder the green wire as a safety ground.

A switched & fused IEC power connector was added to the bottom plate under the bot and a piece of DIN rail was mounted to hold the relay mount.

All 120V wire is 18 AWG we found laying around the lab and the 24V control is 22 AWG.
IMG_1489.jpg
IMG_1490.jpg
IMG_1491.jpg
IMG_1492.jpg
IMG_1493.jpg
IMG_1494.jpg
IMG_1495.jpg
IMG_1496.jpg

Wingcommander whpthomas

unread,
Feb 28, 2013, 10:46:05 AM2/28/13
to make...@googlegroups.com
Nice job - next thing you will need to build is an enclosure perhaps?

flackmonkey

unread,
Aug 16, 2013, 1:36:00 PM8/16/13
to make...@googlegroups.com, john.m...@gmail.com
Why hasn't anyone make a kit for this yet??  I would love to add a HBP to my replicator 2.  I just struggled to print a 2.5x2.5x1 object at 100% infill with out warping and I finally gave up.

Can you comment on you success with the HBP on the rep2?  I assume that you have been using it for a while now.

Thanks,
Mike

Jetguy

unread,
Aug 16, 2013, 2:15:19 PM8/16/13
to make...@googlegroups.com, john.m...@gmail.com
Why hasn't anyone make a kit for this yet?? 
 
 
Ok, maybe the title wasn't catchy enough. Basically, a complete kit. 

Bottleworks

unread,
Aug 16, 2013, 4:44:45 PM8/16/13
to make...@googlegroups.com
I should have used a grander title...

If you're interested see: http://www.bctechnologicalsolutions.com/hbp/index.html

worldburger

unread,
Jul 29, 2014, 8:11:31 PM7/29/14
to make...@googlegroups.com, john.m...@gmail.com
I have a Replicator 2 and have just recently done a similar install (HBP, 100k thermistor except I went with a 12v relay + separate HBP power supply).

John (or anyone else who can answer this), I see from IMG_1491.jpg that you've got the red wire from the thermistor on the MightyBoard in the right most pin, where does the black wire from the thermistor go on ?

I've wired a 100k thermistor with the red in the right most pin socket (by the J5 on the "HBP_Thermistor" part of the board), and the black pin in the adjacent point, but a few issues have arisen for me:

I can't get the extruder AND the HBP to preheat simultaneously.  When I go to preheat the extruder and the HBP, the HBP shows up and will heat, but then the extruder says "waiting".  However, if I only preheat the extruder, the extruder will preheat.

Also, the platform reading in the Replicator 2 screen seems incredibly slow to ramp up to a close temperature (granted my thermistor is beneath the PCB heatbed).  Is there any way to calibrate/adjust the speed of the reading?  My IR thermometer shows a 20C delta from the Rep2 screen.

David


On Thursday, February 28, 2013 10:37:37 AM UTC-5, John McKisson wrote:

Dan Newman

unread,
Jul 29, 2014, 8:51:03 PM7/29/14
to make...@googlegroups.com, john.m...@gmail.com
On 29/07/2014, 5:11 PM, worldburger wrote:
> I have a Replicator 2 and have just recently done a similar install (HBP,
> 100k thermistor except I went with a 12v relay + separate HBP power supply).

Keep in mind that if you're using your own thermistor you should match the
beta(25/85) value to that used by MBI. It's not critical, but it's in
general it's a good idea to get reasonably close so that the firmware's
resistance to temp. conversions are reasonably correct. MBI used
3974K for their beta.

> I can't get the extruder AND the HBP to preheat simultaneously. When I go
> to preheat the extruder and the HBP, the HBP shows up and will heat, but
> then the extruder says "waiting". However, if I only preheat the extruder,
> the extruder will preheat.

This is normal. It's because MBI used and underpowered power supply and it
cannot, from cold, heat up everything at once. So it makes the extruder
wait while heating up the HBP.

Dan

worldburger

unread,
Jul 29, 2014, 10:44:05 PM7/29/14
to make...@googlegroups.com, john.m...@gmail.com
Thanks Dan :)

Is the thermistor beta a number I can alter in MakerBot or ReplicatorG?
25/85? 3974K?  I have to admit, I have no idea what those numbers are or how I apply that knowledge in a "next step" action...

Dan Newman

unread,
Jul 29, 2014, 11:01:17 PM7/29/14
to make...@googlegroups.com, john.m...@gmail.com
On 29/07/2014, 7:44 PM, worldburger wrote:
> Thanks Dan :)
>
> Is the thermistor beta a number I can alter in MakerBot or ReplicatorG?

No, it's a key specification for when you order the thermistor from whomever
you are buying them from. Since you seemed to suggest that you had already
made your own HBP, it's probably too late at this point but I thought I'd
get the info embedded in this thread for anyone else who comes across it in
their research.

At this point all you can do, if it becomes an actual issue, is to build
your own firmware with the proper conversion table matching your thermistor.
Or, better yet, just use the control temp which gets you the desired result
and not worry it any further.

Dan

Rich Webb

unread,
Jul 30, 2014, 8:59:58 AM7/30/14
to make...@googlegroups.com, john.m...@gmail.com
Thermistor beta is a function of the material and its resistance at two specific temperatures, in this case, 25C and 85C. Its units are kelvin; the K isn't a multiplier. Beta can be used as a parameter in to convert between T and R (but there are better estimators). Mostly, think of beta as just a magic number which, along with the nominal resistance at 25C, is a way to match the performance of two thermistors.
Reply all
Reply to author
Forward
0 new messages