Thermocouple hot end upgrade - build thread

A place to talk about user-made mods and upgrades to their machines
User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Fri Jun 06, 2014 2:07 am

After watching the nozzle temperature walk all over the park over the last few months I had a bit of a dummy spit about the accuracy and stability of thermistors and went and ordered myself a thermocouple and interface board to play with, as well as a few of those 40W cartridge heaters.

Through this thread I'll be posting progress as I wire it in, just in case someone else wants to give it a try without the trials and tribulations.

So, let's start with "which thermocouple"?

You're after a K type thermocouple with an interface board. These guys come either as a bonded wire pair or as a screw-in jobbie, usually 6mm thread. And the interface chip can be an older ADA595, MAX6675 or the newer and better resolution MAX31855.

I went ahead with the MAX6675 because it's good enough for me and half the price of the MAX31855. I grabbed a screw-in probe because it's what came for nix with the board. Figured I could screw it into the resistor hole vacated when I swapped the resistors for a cartridge.

This turned up today in the mail (not the hot end, that's my spare), I'm excited!
Image

First thoughts are that:
1. The hole's too big for the screw in, and just pushing it in there leaves the tip exposed to air. I'm going to turn up a threaded blind sleeve which forms an interference fit into the resistor hole, then just screw the thing into that. I've also ordered some plain wire thermocouples for a couple bucks each so, if the insert doesn't work, I have something to jam into the thermistor hole instead.

2. Jeez this is going to end up tidy compared to the silicone mess on my current hot end!

3. More wires required. I'm going to need 5V, ground, clock and data lines - and the data lines kinda need to be shielded or more electronics (differential drivers) and twisted pairs required. We'll see.

4. The thermocouples put out tiny signals, no way am I going to run that all the way to the RAMBO. So the chip is going to live on the effector platform. I'll jerry rig it for now but it could soon be time to transfer the chip to a new PCB that breaks out the heater, fans and thermocouple neatly into a nicer loom. We'll see.

Stay tuned!

User avatar
Jimustanguitar
ULTIMATE 3D JEDDI
Posts: 2618
Joined: Sun Mar 31, 2013 1:35 am
Location: Notre Dame area
Contact:

Re: Thermocouple hot end upgrade - build thread

Postby Jimustanguitar » Fri Jun 06, 2014 8:10 am

Awesome! I'm excited about doing this myself, so keep the updates coming. Great work!

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Mon Jun 09, 2014 5:50 am

Still waiting for the heater cartridges but had an hour free today so I turned up the little adapter plug for the thermocouple, threw it in the freezer, put the heater block of the hot end on the stove for a while to heat it up and put the two together. Brought it all back to room temperature and it's not coming apart ever again.

Image

Installed. The thermocouple tip is firmly pressed into the blind end of the tube, which is thermally coupled beautifully with the rest of the heater block.

Image

Next step the loom and adapter board - I'm off to our local electronics hobby shop tomorrow sometime to try and grab enough cable and the connectors to neaten things up. There's a lot of wire involved.

User avatar
Eaglezsoar
ULTIMATE 3D JEDDI
Posts: 7185
Joined: Sun Apr 01, 2012 5:26 pm

Re: Thermocouple hot end upgrade - build thread

Postby Eaglezsoar » Mon Jun 09, 2014 7:58 am

Where in the world did you find the little adapter plug for the thermocouple?
This is the first time I have ever seen one and I didn't know they even existed.
Thanks for bringing these to our attention.
“ Do Not Regret Growing Older. It is a Privilege Denied to Many. ”

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Mon Jun 09, 2014 9:29 am

In my lathe, dude.

Bit of scrap aluminium drilled 15mm deep with a 5.5mm bit then tapped with an M6x1.0 tapered tap, then turned down on the outside to 6.6mm *note walls are only 0.3mm thick* before being parted off and tapered in the linisher.

Took about half an hour, 15 minutes of that was finding the bit of scrap.

Kinda glad that the lathe and other assorted tooling in the shed aren't a complete waste of space now that I print a lot of my gizmos instead.

I have ordered a couple of simple wire ended thermocouples too, so when they turn up I'll see what we can do with the standard thermistor hole for people who aren't so fortunate to have the other toys.

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Mon Jun 09, 2014 9:33 am

Hmm, I'm also thinking of maybe replacing the PEEK block with a bit of threaded SS tube (1.75mm internal bore) and the top of the mount with a custom finned job. It might be nice to be able to print nylon without the expense of new hot ends. And another excuse to make up for my neglect of my lathe over the last few months :)

User avatar
JohnStack
Printmaster!
Posts: 852
Joined: Thu Jan 03, 2013 7:07 pm
Location: Carlsbad, CA
Contact:

Re: Thermocouple hot end upgrade - build thread

Postby JohnStack » Mon Jun 09, 2014 10:31 am

Nice!

What is going to be your process for testing?
Technologist, Maker, Willing to question conventional logic
http://dropc.am/p/KhiI1a

bdjohns1
Printmaster!
Posts: 238
Joined: Sat Jan 25, 2014 9:50 pm
Location: Madison, WI

Re: Thermocouple hot end upgrade - build thread

Postby bdjohns1 » Mon Jun 09, 2014 7:43 pm

Very cool. I've worked with Pt100 RTDs before - generally even more accurate than K-type thermocouples - I've got some in my homebrewing system. They're also our standard device of choice at work. I haven't researched into what form factors I could get one in, but it would be another nice option - a standard Pt100 should be good to >500C.

Since it's just a resistance device like a thermistor, it should be possible without an interface board. I thought I read somewhere that Up! printers use an RTD.

User avatar
McSlappy
Printmaster!
Posts: 809
Joined: Wed Dec 11, 2013 9:11 pm
Location: Queensland, Australia
Contact:

Re: Thermocouple hot end upgrade - build thread

Postby McSlappy » Mon Jun 09, 2014 7:49 pm

Captain Starfish wrote:Hmm, I'm also thinking of maybe replacing the PEEK block with a bit of threaded SS tube (1.75mm internal bore) and the top of the mount with a custom finned job. It might be nice to be able to print nylon without the expense of new hot ends. And another excuse to make up for my neglect of my lathe over the last few months :)


Sounds like a low cost but sweet upgrade you could offer Rostock users....
I loved my Rostock so much I now sell them in Oz :)

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Mon Jun 09, 2014 8:52 pm

Testing: same as I use to calibrate Mr Thermistor, bung the multimeter thermocouple under the nozzle and run it up to temperature, then let it cool down and watch the two temps to ensure they are tracking together nicely. I expect a little lag between the block and nozzle but not much.

Hot end upgrade: I'll need to find someone local with a FLIR unit so I can watch what's happening across the heat break or, failing that, run a thermocouple down the filament tube and check temperature vs depth. Making them for general distribution? I don't mind making one, that's quality shed time = relaxing, a little mental and dexterity challenge, good stuff. More than one? That's production instead of prototyping and it smells too much like work. Sadly, as with most of the stuff I throw together here, it might cost $3.00 in raw materials but by the time I add my hourly rate to it you end up with a $300 upgrade to a hot end which probably won't work anywhere near as well as an E3D or Prometheus hot end. That's assuming it works at all.

atoff
Printmaster!
Posts: 277
Joined: Sun Mar 30, 2014 5:24 pm

Re: Thermocouple hot end upgrade - build thread

Postby atoff » Mon Jun 09, 2014 10:21 pm

Thanks for sharing your progress. I'm planning on doing this as well, though with E3D's, so I'll be picking up a thermocouple without the bolt. Saw that kit you picked on Ebay for $12, not bad, considering everything's included.

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Tue Jun 10, 2014 1:14 am

Trying to rationalise and future proof my cabling requirement to the effector platform and what I'm seeing is a lot of gear hangs off the 12V2 rail which is nice.

So it's looking like I'm going to (eventually) be rocking out with three fairly high power lines:
- 12V2 rail
- Heater 0 return
- Heater 1 return

A few low power rail or switching lines:
- Fan 0 and Fan 1 returns (from 12V2)
- 5V rail
- Signal Ground

And a few signal lines:
- Z switch (just in case one day)
- SCK, SDA lines for the thermocouples
- CS line (with a discrete transistor inverter on the effector board to toggle between two thermistors) in case of a 2nd hot end one day.

Interestingly I can buy some VGA cable which has three fatty coax lines for RGB (might use for power?) and nine other lines (signals, power, whatever) plus a shield in the one bundle with maybe a neat setup for connectors. Or I can tape the existing four core heavy duty rig onto a nine core shielded line and have a slightly less neat but much flexier cable.

Any thoughts?

bdjohns1
Printmaster!
Posts: 238
Joined: Sat Jan 25, 2014 9:50 pm
Location: Madison, WI

Re: Thermocouple hot end upgrade - build thread

Postby bdjohns1 » Tue Jun 10, 2014 2:12 pm

Since I immediately switched to an e3d hotend after building my kit, here's how I'm wired:

18AWGx4 wire - 2 for the hotend, 2 for the e3d heatsink fan. Hotend is on an XT60 connector, fan is one of those polarized locking connectors like the RAMBO board.
Cat6 cable (24AWGx4 pairs) - one pair for thermistor. 3 pairs available for future expansion (cooling fan, LEDs, etc).

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Tue Jun 10, 2014 7:27 pm

Thanks, bdjohns1. The VGA cable was a washout - didn't sell by the meter and I wasn't prepared to drop $100 on 30m of the stuff just for an experiment.

I'm trying to keep extra cable free for a second hot end, LED ring and a Z probe. So the high power stuff could just be the existing four core for the hot ends, and the rest for fans, switches and signal. Which takes us up to lines pairs, so the cat 5 might be a winner. Although it'd probably be better to run the thermocouple signal lines in a separate shield with fans and stuff not really mattering that much. Hmmm. Maybe that's the best way, keep the existing couple of 4 cores (one thick power one, one thin fan one) and add another 4 core shielded job for the thermocouples.

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Sun Jun 22, 2014 4:42 am

Spent some time on it today. No time spent tinkering is really "wasted", having said that, today was a waste of time: a lesson in assume making an arse out of you and me. Well, me.

SD and SCL line labels on a micro are generally a good indication that they are I2C protocol compatible so I spent the morning neatly wiring up my 6675 board to the I2C breakout port on the RAMBo.

Buh bow...

6675 ain't I2C, it's SPI (a different digital serial comms protocol) which needs an extra pin on the cable, an extra pin on the connectors and, because of the way SPI works and is pinned out on the RAMBo, it means I'm going to have to hack the LCD adapter board to get at the SPI pins and change pins.h to move the chip select line to something also on the LCD board. None of this is deadly difficult, I just have to go through and find a free pin in the expansion header that's not presently used by the LCD board and solder into that.

Annoying but closer than I was this morning. When I have it all together I'll post up photos and code changes required.

Still waiting on my eBay heater cartridges too, that's getting a bit old.

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Mon Jun 23, 2014 12:33 am

Life just got hard.

I have two issues.

First was finding a chip select pin to use. The SPI interface is a bus that a whole bunch of things can hang off (in our case it's how the RAMBo talks to the SD card and the digipot that sets the current on the steppers). Each chip on the bus has a separate digital chip select line that gets triggered by the master (ie the 2560 or brain of the RAMBo) to say "YOU, YES, I'M TALKING TO YOU!". If that line isn't set, the chip ignores the bus and won't write to it. Obviously only one can be on at a time else things get messy.

Anyhoo, need to find a random free digital pin off the 2560 that I can get to. By going through the schematics of the LCD and RAMBo boards and beeping out the adapter board for the LCD it looks like I can choose between pins PH7, PH2, PD6 and PD5 from the extension port for this chip select line as they are unused by the LCD. Going through the netlists for the RAMBo turns out PD5 is used for an LED so that's out. I'ma go PH7 I think, just because pick one = pick the first.

Question for RAMBo nerds: Have I missed something here, or is PH7 truly a free pin?

Moving along to 2nd issue, is a choice. I can just solder the VCC, Gnd, MOSI, MISO and CS wires up to the effector platform directly onto the back of the LCD adapter board (which I will do to debug) but this kinda sucks, I think. What's the concensus - is it worth making up a new adapter board which breaks out another little connector with SPI and a couple of extra lines for chip selects (two of 'em in case someone wants twin thermocouples)?

John et al, if you pick this post up: do you guys make these little adapter boards or are they RepRap/Ultimaker gear?

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Mon Jun 23, 2014 4:55 am

Fffffffun and games: none of the spare pins on the EXT port mapped as Arduino pins so some custom code in Extruder.cpp is required. No worries, got that done today and put it all together: except now the blasted thing won't upload the firmware whilst the LCD adapter is plugged in!

Hmmm. Enough frustrage-on for one day. Will have another look tomorrow.

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Tue Jun 24, 2014 1:15 am

And here we come to a screeching halt.

With the cable running up from the SPI interface to the top of the Max and back down to the effector plate, the unit refuses to program, refuses to let the LCD drive, refuses to get temp. This is with or without the adapter board connected at the effector end.

Clip the "SCL" line at the LCD adapter board and it all comes back to life - except for the temperature, obviously.

My take on this is that we've exceeded the maximum length of the SPI clock line and either noise or transmission line ringing are just botching up the whole interface. SPI was designed for very, very short runs eg from RAMBo to the SmartController LCD. I2C has the same issue. I could run a longer thermistor cable but then we have a tiny signal running parallel to power and hard switching which sounds like a recipe for disaster. Or I could run a different protocol - RS485 maybe - between the effector platform and the RAMBo but that would mean rewriting the temperature getter code, finding an available serial expansion port line pair and adding new circuitry to both ends for the line level conversion.

Gah.

Time to rethink this - either with an I2C compatible chip (needing a rewrite), with the Olimex board (exxy) or with the AD595 which spits out an analogue voltage related to the temperature, and is probably the easiest way to go forward.

Time to chase parts again, then...

User avatar
Eaglezsoar
ULTIMATE 3D JEDDI
Posts: 7185
Joined: Sun Apr 01, 2012 5:26 pm

Re: Thermocouple hot end upgrade - build thread

Postby Eaglezsoar » Tue Jun 24, 2014 11:16 am

Captain Starfish wrote:And here we come to a screeching halt.

With the cable running up from the SPI interface to the top of the Max and back down to the effector plate, the unit refuses to program, refuses to let the LCD drive, refuses to get temp. This is with or without the adapter board connected at the effector end.

Clip the "SCL" line at the LCD adapter board and it all comes back to life - except for the temperature, obviously.

My take on this is that we've exceeded the maximum length of the SPI clock line and either noise or transmission line ringing are just botching up the whole interface. SPI was designed for very, very short runs eg from RAMBo to the SmartController LCD. I2C has the same issue. I could run a longer thermistor cable but then we have a tiny signal running parallel to power and hard switching which sounds like a recipe for disaster. Or I could run a different protocol - RS485 maybe - between the effector platform and the RAMBo but that would mean rewriting the temperature getter code, finding an available serial expansion port line pair and adding new circuitry to both ends for the line level conversion.

Gah.

Time to rethink this - either with an I2C compatible chip (needing a rewrite), with the Olimex board (exxy) or with the AD595 which spits out an analogue voltage related to the temperature, and is probably the easiest way to go forward.

Time to chase parts again, then...


If you suspect noise on the lines, put a ferrite core on the wires. They work great to reduce the noise. There are several topics on this forum on the ferrite, search and ye shall find!
“ Do Not Regret Growing Older. It is a Privilege Denied to Many. ”

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Thu Jul 03, 2014 6:44 pm

Thanks for the tip. Unfortunately what it looks like is ringing causing the jitter, not noise. Purely a function of line length and non slew-rate limited signalling into high impedance inputs where the change in a digital signal propagates to the end of the cable, bounces back and can look like a second pulse to any inputs close to the initiating end.

Anyway. Today's tip: avoid the "newfrog" ebay seller when buying cartridges. A month and a half later I've had to ask for a refund before the PayPal cover expires, and have ordered another set of them as a second chance. Whilst I'm waiting for those guys, I'll also be laying out a new PCB to sit atop the extruder and clean up the cable mess as well as housing the AD597.

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Sun Aug 31, 2014 9:17 am

Things kinda got stalled for a while, waiting for the cartridges and then with a new job and some other more pressing projects too, just haven't had time to touch it.

Anyway, during the week I installed a new cartridge with a screw from the top of the heater block clamping it in place. Then the best bit of bad timing ever today: about 5 minutes after deciding "Nah, thermocouple is overkill, it's doing ok now that I have the thermistor dialled in nicely" I went to switch from 0.5mm to 0.35mm and discovered a plug around the nozzle. Pulled it all out of the effector platform for a wander down to the shed to drill out the plug in the press and _BANG_ there goes a thermistor lead.

Crap.

Seems like a cosmic hint to finish this project. So I've got the SOIC (rectangle, not round) AD597 on a breakout PCB from Sparkfun. The thermistor wires needed silver solder tinning first before I could solder them to the board, which I've done. Now just need to add a PSU decoupling capacitor and a connector to be able to remove the whole head (thermocouple and heater) from the loom for whatever reason. Given the nice long wire length I reckon I'm safe to solder the heater wires this time around, but we'll see how that goes :)

User avatar
Captain Starfish
Printmaster!
Posts: 955
Joined: Tue Dec 10, 2013 5:24 am

Re: Thermocouple hot end upgrade - build thread

Postby Captain Starfish » Fri Sep 05, 2014 10:52 pm

Other Aussies may remember "Upside down, Miss Jane".

And that caught me today, soldering the breakout onto the loom and I managed to fry the AD597.

Next step, get another board preassembled and ready to go: that's two weeks away. But I need to do some printing.

So the money is probably going instead to a shiny new E3D extruder from McSlappy - email on its way, dude.

This project is now shelved, many lessons learned but too many obstacles for (in the end) dubious advantage. If I need to go over the thermistor temps on the E3D I'll fire it up again, but otherwise I think I'll walk away from this one whilst I still have a modicum of dignity remaining.

andrewpaul
Noob
Posts: 1
Joined: Thu Nov 09, 2017 9:19 pm
Contact:

Re: Thermocouple hot end upgrade - build thread

Postby andrewpaul » Wed Dec 06, 2017 10:49 pm

Agreed! The type K sensor is a good overall selection.
For more information, please visit https://www.pyrosales.com.au/


Return to “Mods and Upgrades”

Who is online

Users browsing this forum: No registered users and 1 guest