Conversion to Twin Heaters

[lars86 42]

Hi guys,

I could use a bit of help from the electronic gurus, I'm more of a mechanical guy.

I redesigned the whole print head / hot end (again), and decided to use twin heaters for more uniform heat flow. I went with 6mm cartridges rated at 40w / 24V; which I calculated to have about 25w each @ 19v. I figured that running them in parallel for 50w would be plenty.

The cartridges are wrapped in aluminum foil to acheive a very tight fit to the block.

When I fired it up to test, the hot end had a very hard time reaching temperature. After insulating the heater block a bit, I was able to get up to printing temps (didn't see how high it would go), but it definitely struggled compared to the stock setup.

Any ideas as to what could be going on? Did I miscalculate?

[lars86 42]

Anyone?

I reduced the wall thickness on the heat brake and the seemed to help. I was able to get it running a print (still took a while to reach temperature). But only a few minutes in, the display started doing that "fade out" thing and gave me a "heating failed" error.

[KevinMakes 22]

Sooo...heres teh problem. First of all there is a formula to figure out how hot the heaters should get at a known voltage... if the thermistor does not read the temperature as being within...(3deg) (its been a while since I read through the firmware) then you will get an error.

Thought #2

You will have to change the firmware... Dont try to avoid it.... theres no way around it.

My suggestion. Use the second heater connection on the motherboard for dual extrusion that never came to be...(thanks a lot UltiHeadCheeses) and then use the dual extrusion firmware that is (very) well hidden in the cura software. install that firmware. then when you start a print turn that heater on in the options to the desired temperature (requires a second thermistor)

Any questions just reply here. Ill talk you through the painful stuff.

[KevinMakes 22]

Here's the long difficult solution to the heater block error

If you have solidworks premium with simulation this is a lot easier.

*************DISCLAIMER********************

-This sciency part is probably wrong

- This worked for me

- This MAY not work for you

- I am still working on my coffee

- This is a lot easier with solidworks simulation

- you literally have to add material types hot spots and get temp @ thermistor point

 

Find the thermal conductivity of the material used buy UM for their hot ends. (watts per meter kelvin (W/(m·K)

-Yellow brass I believe(

Then Find the thermal conductivity of the material you are using

(I suggest just using their material)

If you know it then compare the distance and dissipation of your heater bock

IE if you are using 306 stainless steel 26(tc) then you will have to place your

heater 2.5 times closer than the yellow bras with a (tc) of 67.

67/26

It should get you close enough to beat the "heater block" error 90% of the time

[KevinMakes 22]

also I wouldnt stick the aluminum foil in the heaterblock... if the foil fits you are doing something wrong prbly

If you want a good fit find some thermal compound made for a temperature that you are trying to achieve... on second thought... You could probably use some ceramic compound for that.

[lars86 42]

Sooo...heres teh problem. First of all there is a formula to figure out how hot the heaters should get at a known voltage...

 

Kevin, thanks for the replies! This forum is like a ghost town.

I'm not sure I'm familiar with a formula like that. There are very basic formulas that govern a circuit like this ( Joule's law, Ohm's law). I have done these calculations. But, how hot somethings gets, is much more complicated and involves heat transfer.

 

if the thermistor does not read the temperature as being within...(3deg) (its been a while since I read through the firmware) then you will get an error.

Thought #2

You will have to change the firmware... Dont try to avoid it.... theres no way around it.

My suggestion. Use the second heater connection on  the motherboard for dual extrusion that never came to be...(thanks a lot UltiHeadCheeses) ;)and then use the dual extrusion firmware that is (very) well hidden in the cura software. install that firmware. then when you start a print turn that heater on in the options to the desired temperature (requires a second thermistor)

I change the firmware quite a bit, so I'm not against it. What do you think needs to be changed? I'm sure that tuning the heater PID parameters will improve the thermal response, but on my first tests, the heater output was pegged at 100% duty cycle (which should be a constant 19V), yet still struggling, so no PID tuning will help that.

Using both heater circuits would give you more output capacity, but as the max output of my new configuration is only 10w above the original, I'm not sure that I would need to.

[lars86 42]

Here's the long difficult solution to the heater block error

If you have solidworks premium with simulation this is a lot easier.

                *************DISCLAIMER********************

       -This sciency part is probably wrong

       - This worked for me

       - This MAY not work for you

       - I am still working on my coffee

       - This is a lot easier with solidworks simulation

       - you literally have to add material types hot spots and get temp @ thermistor point

       

Find the thermal conductivity of the material used buy UM for their hot ends. (watts per meter kelvin (W/(m·K)  

-Yellow brass  I believe(

Then Find the thermal conductivity of the material you are using

(I suggest just using their material)

If you know it then compare the distance and dissipation of your heater bock

IE if you are using 306 stainless steel 26(tc) then you will have to place your

heater 2.5 times closer than the yellow bras with a (tc) of 67.

67/26

    It should get you close enough to beat the "heater block" error 90% of the time

 

My heater block and the Ultimaker block are made from aluminum, so no substantial difference in thermal conductivity. UM makes their nozzles from brass, but that won't affect heater / thermocouple placement.

I have done extensive thermal simulation on this hot end configuration.

[lars86 42]

also I wouldnt stick the aluminum foil in the heaterblock... if the foil fits you are doing something wrong prbly

If you want a good fit find some thermal compound made for a temperature that you are trying to achieve... on second thought... You could probably use some ceramic compound for that.

 

I have yet to see a compound that is rated for sustained temperatures like a hot end produces. If you have a suggestion, I'm all ears!

Using aluminum like that is pretty common. We are not talking much, just a couple layers. Many people suggest that even if the heater is a slip fit to the block, that it expands enough during heating to form solid contact. Adding aluminum foil helps to fill any airspace voids which would not conduct heat well.

[gr5 2,229]

Using both heater1 and heater2 on the ultiboard is a firmware nightmare - you would have to rewrite the PID code. I don't recommend that. Hopefully the single MOSFET can handle 7.7 ohm (50W) load.

Your calculations were fine.

Did you measure the resistance of the heaters? I would check them. They were supposed to be around 14.4 ohms but it sounds like they sent you the wrong heaters.

How much heavier is your aluminum block than the default one? Even if twice as heavy it shouldn't be more than twice as hard to heat, right? Also what is the surface area of your block versus the original?

[KevinMakes 22]

No problemo, I am not sure what the formula was exactly either I can take a peek in the firmware.

The basic Idea that I got from the formula was that there is a imaginary table (represented by the formula and rules)

and it creates an imaginary line graph with watts on one side and temperature on the other

and if for some reason that the watts input did not create the (degrees) that it "should" produce within 3deg (i think it was 3) then it will kick out one of the heater block or heater errors.

Could I get a few more details from you?

Picture of the heater block (to satisfy curiousity)

exact type and stats of your heating elements and thermistors

changes you made to the firmware (general)

the suggestion about using the other on board connections for the secondary heating setup is more pointed towards having better control, and so you can keep things (firmware) as "stock" as possible.

[KevinMakes 22]

Using both heater1 and heater2 on the ultiboard is a firmware nightmare - you would have to rewrite the PID code.  I don't recommend that.  Hopefully the single MOSFET can handle 7.7 ohm (50W) load.

 

There is actually firmware (inside cura) that wonderfully manages the secondary heater and thermistor.

[KevinMakes 22]

How much heavier is your aluminum block than the default one?  Even if twice as heavy it shouldn't be more than twice as hard to heat, right?  Also what is the surface area of your block versus the original?

 

The simulation should cover the area, density, and thermal conductivity. guessing and checking is a craps shoot from what I personally found.

[KevinMakes 22]

I have yet to see a compound that is rated for sustained temperatures like a hot end produces. If you have a suggestion, I'm all ears!

 

There is a few copper ceram. ones that are only "rated" for 200 ish but it wont hurt it any to go higher. I've been doing it for months.

 

Using aluminum like that is pretty common. We are not talking much, just a couple layers. Many people suggest that even if the heater is a slip fit to the block, that it expands enough during heating to form solid contact. Adding aluminum foil helps to fill any airspace voids which would not conduct heat well.

 

I have found that the heater is not slip to fit, and the "retaining screw" for the element and the thermistor really sucks. I used a tiny bit of paste and pressed the element and thermistor into the block until it escaped through the retaining screw(which I threw into the bin within a few days of owning the printer) hole. I don't think I would be able to fit a few layers of tinfoil in that gap. and when I say heater block I really mean (nozzle) sorry for the confusion.

[lars86 42]

Using both heater1 and heater2 on the ultiboard is a firmware nightmare - you would have to rewrite the PID code.  I don't recommend that.  Hopefully the single MOSFET can handle 7.7 ohm (50W) load.

Your calculations were fine.

Did you measure the resistance of the heaters?  I would check them.  They were supposed to be around 14.4 ohms but it sounds like they sent you the wrong heaters.

How much heavier is your aluminum block than the default one?  Even if twice as heavy it shouldn't be more than twice as hard to heat, right?  Also what is the surface area of your block versus the original?

 

You know, I should have done that right away! hahaha

I checked the heaters and they measured 8.2 ohms in parallel. After switching back to the stock heater, everything seemed fine. But a few days ago, I got another heating failed error. I went underneath and reseated the heater wires in the terminal block. Since then, I have run a couple 7 hour prints without issue. I'm wondering if that error was just coincidence.

The physical size / mass of the heater block isn't as important as the steady state characteristic of the whole thermal system. Basically, looking at all the heat flows (at the desired temp): into the filament, up into the heat sink, and radiative/convective loss. All that combined equals the steady state power consumption. Things like surface area can definitely have a big impact here. But mass / volume would just slightly slow the initial heating.

[lars86 42]

No problemo, I am not sure what the formula was exactly either I can take a peek in the firmware.

The basic Idea that I got from the formula was that there is a imaginary table (represented by the formula and rules)

and it creates an imaginary line graph with watts on one side and temperature on the other

and if for some reason that the watts input did not create the (degrees) that it "should" produce within 3deg (i think it was 3)  then it will kick out one of the heater block or heater errors.

Could I get a few more details from you?

   Picture of the heater block (to satisfy curiousity)

   exact type and stats of your heating elements and thermistors

   changes you made to the firmware (general)

the suggestion about using the other on board connections for the secondary heating setup is more pointed towards having better control, and so you can keep things (firmware) as "stock" as possible.

 

Interesting, I'll have to dig around in there!

I've taken a little break from this project as I've been busy with other stuff. I actually had another issue, where molten filament started to leak out from a press fit between the thermal brake and heater block! I did not expect that one.

[KevinMakes 22]

I've taken a little break from this project as I've been busy with other stuff. I actually had another issue, where molten filament started to leak out from a press fit between the thermal brake and heater block! I did not expect that one.

 

Oooh that sounds fantastic haha

shoot some pictures when you get a chance. I am an industrial designer by trade so I only use my 3d printer every once in a while so I understand getting pulled off a project

[gr5 2,229]

"heating error"? I think there are 3 or 4 different errors and they mean different things. I don't remember which is which.

There is a test in Marlin that if the heater PID is demanding 100% power for more than X seconds and the temp doesn't rise by at least Y degrees in that time then you get a failure and power is removed. Is that what you are talking about? There are about 4 different X and Y values depending on Marlin version. Typical is 30 seconds, 10 degrees.

This is to detect if temp sensor or heater slipped out of the block and prevent a melt down.

I may have skimmed your original posts too quickly the first time but I don't remember any mention of error messages so I may have led you down the wrong path on my earlier post.

[gr5 2,229]

More details here:

https://ultimaker.com/en/community/view/15720-3dsolex-installation-support#reply-106396

[lars86 42]

Thanks gr5, good to know!

I am not 100% on the error message, because it was very hard to read. The LCD was doing it's best impression of melting down. Not sure why that happens.

[gr5 2,229]

I had to to tweak my UMO Ulticontroller after about a year. There's discussion somewhere but it's pretty simple. Remove all the wood around it and inside is a potentiometer that you can rotate with a plastic screw driver. Play with that until display looks ideal. It adjusts brightness/contrast or something similar.