Very bad Onyx heating (max 80°C)

[mhackney]

Basically in a resistive heating element like that in an Onyx, power is converted to heat. Increasing power will get the temperature to a set value faster - as long as the device can handle it. So, using a bit of electronics calculations, you need to figure out how to apply more power to the Onyx (again, as long as you don't over power it and melt the traces, etc). The equations used are Ohm's Law and the power law:

Ohm's Law is V = I*R where V = voltage, I = current and R = resistance. The Onyx's resistance is fixed at 1.3Ω. Rearranging the formula, R = V/I or 1.3Ω = V/I. This tells you that current will be proportional to voltage. So, if your power supply can supply the current, increasing voltage from 12 to 15 to 24 volts will be met with a proportional increase in current. With this knowledge, you can then apply the power law: P = I*V where P is power. So, having calculated V and I with Ohm's law, you can calculate the power by plugging those into the power law, or you could use an online calculator that calculates power for you using these equations. With this calculator, enter the Onyx's resistance and the voltage and it will calculate the current and power. Try it for 12, 15 and 24 volts and you will see that power increases (as does amperage).

12V - 9.2A - 111W
15V - 11.5A - 173W
24V - 18.5A - 443W

The question has been asked, is the Onyx engineered to handle 17A of current? And, of course, the RAMBo version 1.1 is limited to 15A 12-24V so you won't be able to pump more than 15A to the Onyx. Plug 15A into the equation with the resistance and you get 19.5V and 292.5W. I've talked to Steve at SeeMeCNC and he thinks the board was designed for 12A or so with a few amps margin of error. He believed it should handle 15A but told me to try at my own risk. I have been running my Rostock on a 24V 14.6A (350W) supply for a few weeks now. I have over 30 hours of "heat time" on the Onyx - admittedly not a lot of time but folks here seem impatient and thirsty for information so I'll tell you what I've learned!

I see no evidence of fatigue, blistering, excessive warpage or any other physical damage to either side of the Onyx or the electronics connector pad. The LED works properly too - I am running the 24V through the RAMBo to power everything.

SIDEBAR: I saw an earlier post somewhere about running steppers at 24V. There is absolutely no problem with doing this. Read section 7 of this great overview on steppers and this article on Stepper Motor Voltages Explained. The SeeMeStepper is this one 42BYGHW811 and the rated voltage spec is 3.1V. So, running at 24V is only 8X, which from the Gecko article:

An empirically derived maximum is 25:1, meaning the power supply voltage should never exceed 25 times the motor’s rated voltage

makes it still well within spec. I can not tell any difference in my steppers running at 24V, they don't seem to run any hotter - basically, everything works just fine and heating is much faster.

With my 24V supply, the naked Onyx heats incredibly fast to 100°C. Putting anything on it like aluminum or glass or both adds thermal mass that needs to be heated via convection. This takes power, so of course it takes longer to get to temperature with either or both of these for the same power input. Currently, I have my 1/8" thick aluminum heat dissipator and the borosilicate glass plate on top of that - a fair amount of extra mass to heat up. Below is my temperature-time curve from Repetier that shows heating the Onyx from 21°C to 80°C with the aluminum and glass in place. It took just a hair over 5 minutes. I've done the auto PID calibrate on this and you can see the temperature rises and stabilizes pretty quickly. The temperature - measured with a thermocouple - at the center of glass surface after stabilizing was 80.1°C and several points at the very edge of the glass were 79.2° and higher - very even temperature distribution.

I have also done some quick heat up tests with just the aluminum and just the glass plate. The time decreases by about 1/2 in both cases - the aluminum is a little quicker to heat than just glass.

Only time will tell if there is any damage to the Onyx. I am not running mine at greater than 80°C so I can't say what would happen increasing power to get up to say 100°C for an extended time. I'm quite happy with the performance of both the heated bed and the rest of the system. I have added a 24 to 12 volt convertor to run my auxiliary fans and LEDs. The fan controlled via software runs fast at 100% so I just turn it down to 50% or so.

Finally, the last thing I'll add is, I have purchased one of the 24V silicone heat pads. I intend to test it as soon as I have time. It has a built in thermistor too. The challenge with it is it is not perfectly flat. There is a bulge where the wires come in that has to be accounted for. I haven't decided how I want to address that yet. I will likely attach it to the bottom side of one of my aluminum plates with Permatex Copper around the edges to get good physical contact and then elevate the assembly with standoffs but without the snowflake spacer included in the Rostock kit.

Shewwww, sorry for the long winded post!

[dbarrans]

You certainly don't need to apologize for being long winded. Information is power (sorry, I couldn't help myself!), and I appreciate you taking the time to document what you've learned.

- dan

[aehM_Key]

Thanks for writing everything so nicely down

The Onyx's resistance is fixed at 1.3Ω.

This is only right for a cold heatbed. The resistance increases to 1.7Ω @ 120°C. So you have less power at higher temperatures.

I've used mine now for many hours with 23V @ 125°C without any problems. Maybe the aluminium plate keeps the traces from flying into outer space

[mhackney]

Thanks guys. aehM_Key you are absolutely correct that resistance changes as a function of temperature. One thing the aluminum plate may do is act as a big heat sink (which it is of course) that actually helps minimize local hot spots that could result in failure of the traces. Just a hypothesis but it seems reasonable. Running the Onyx at 24V with only glass or nothing might lead to failure but until someone does it, we won't know.

[geneb]

Can the original RAMBo board that shipped with the Indigogo kits (which is what I have) take a 24v power supply? If not, would a 15V supply be a good choice? My Onyx won't heat past 83C.

g.

[aehM_Key]

http://reprap.org/wiki/Rambo -> Heated Bed, 15A 12-24V (version earlier than 1.1 limted to 14A 16V due to fuse rating, see rambo_development for hacks)

The early kits shipped with version 1.0e, as far as I know.

I changed the fuse: http://forum.seemecnc.com/viewtopic.php?f=42&t=1289

[Polygonhell]

Can the original RAMBo board that shipped with the Indigogo kits (which is what I have) take a 24v power supply? If not, would a 15V supply be a good choice? My Onyx won't heat past 83C.

g.

It's a little different depending on if you just use 24V for the heated bed or intend to use it for all the power.
You can't use 24V to the logic circuit because the Arduino voltage converter is only rated for 6-20V and in practice it isn't recommended to exceed 12V.
You can however power that from USB and use 24V on the other lines with caveats.
24V to the hotend circuit is almost certainly going to blow the standard 5A fuse on the 1.0E boards, you could swap it for a 10A fuse, I did this when my 5A fuse failed (on 12V).
The 1.0E boards use a resettable poly fuse for the bed, basically if it gets too hot it stops conducting. Whether you could push the almost 20A through it you'd see from connecting the Onyx to a 24V supply is probably a crap shoot, it's the same part used on RAMPS boards and I have one RAMPS board here I can't even connect to a standard heated bed at 12V without it over heating without active cooling. I think you'd end up having to replace it with a higher rated fuse, which brings up the question is the rest of that circuit up to the larger current draw and that I don't know.

[cjdavis618]

Here are some things I have found with my factory Rostock Kit and some general observations.

At first when I got the machine built, every thing worked great except the Onyx. I could get readings of the temp on the Onyx, but regardless which firmware that I used, it wouldn't turn on. I checked pin.h, configuration files but regardless nothing worked. No red light was coming on with the manual or Gcode entry of turning on the headbed. After some careful measurements with my fluke meter, I found that the 15amp blade fuse holder was not correctly soldered down on the board and was only making contact on one side which was the input side. Output to heatbed was feeding air. Fixed that and then I was able to get my bed heating up.

At the time, I had taken the 6pin connector from the 450watt powersupply and cut that off and fed the whole board with it. So I only had one yellow and then one black feeding the heatbed input from that set of 6 wires, mosfets used a pair and the board was fed the other pair. I was able to get the heated all the way to 100C but it took nearly 20 minutes. With the discussions here, I decided to use all 3 yellow and all 3 black, tie each of the color groups together, solder them and then provide more wire into the input side. I was already using some thin stranded 12Ga wire for the output to the heatbed so no changes needed there.

What happened after that change was that I couldn't get the heatbed above 72C after 30 or more minutes after that change. I then opened the door on the rostock to make sure my wires were not loose or anything and I noticed the cap below the blade fuse just fell off of the board on to the platform. So obivously there was ANOTHER soldering issue to deal with. I corrected that and then checked other localised components and then put it back together. No other issues were readily found.

After I put it back in, I still have the wire groups soldered together, but I was then able to get the heat bed up to 93C but not any higher. Even after setting for an hour. As for the power inputs, I am now going to remove the other 2 sets of wires and see what happens and see if indeed the 2 18ga input wires can again return to the 100C mark like was working.

I am still calibrating my machine now but I have learned some things about it that may help some and I will post them in another thread. I saw there were so many build threads, I decided that I would be re-inventing the wheel, but I could add some pointers that the novice like me may not know and not easily figure out. Maybe this information above will help someone else.

[Eaglezsoar]

I think the best way is the way Cambo3d did it in his build. The 24v goes into the output of a DC-DC solid state relay to the Onyx.
The input to the solid state relay is the 12v output for the bed. When the Rambo decides it's time to turn on the bed it actually
turns on the solid state relay which then passes the 24V to the bed. Works great and no damage to the Rambo because no 24v
passes through it.

[aehM_Key]

Works great and no damage to the Rambo because no 24v passes through it.

This implies, that the Ramo will get damaged, if used with 24V. But this is not the case. I'm using it with 24V. Only the fuse has to be changed, if the board revision is <1.1 . No need for a SSR.

[Eaglezsoar]

Works great and no damage to the Rambo because no 24v passes through it.

This implies, that the Ramo will get damaged, if used with 24V. But this is not the case. I'm using it with 24V. Only the fuse has to be changed, if the board revision is <1.1 . No need for a SSR.

Point taken but I would rather not take a chance on a $200 card. It's great that we come up with different ways to accomplish the same things. If the Rambo wasn't boxed with the kit, I would have
gone with Ramps. I do not like the motor drivers built into the card, the old "all your eggs in the same basket". I'm glad you got yours working without problems.
Carl

[cambo3d]

yes you can run the rambo at 24v, just have make sure all the other factors are taken into account.

[TheMakerGuy]

Here is a link to my google group where I am working with my newly built Rostock Max. I have wired the power supply that came with the machine three different ways and this is what I decided to do. I am using a CNC power supply. I can reach 100c in 20 minutes. Feel free to join the group if you like. I am posting things as I figure them out and it is not just for Rostoc. I hope this helps. Thanks.

https://groups.google.com/d/msg/3dprint ... o4LNTdhRAJ

[Eaglezsoar]

yes you can run the rambo at 24v, just have make sure all the other factors are taken into account.

You have that massive 24v supply that can output 20+ amps. Have you connected it to the Onyx and applied the full 18+ amps
without any problems to the Onyx?

[cambo3d]

yes you can run the rambo at 24v, just have make sure all the other factors are taken into account.

You have that massive 24v supply that can output 20+ amps. Have you connected it to the Onyx and applied the full 18+ amps
without any problems to the Onyx?

I've not gotten around to taking measurements of current draw at varies temperatures. I assume that the onyx is using the full required current because of how fast it heats up. Will confirm/unconfirm when I get some actual measurements. edit: i'll do this today and post in my build log.

[cassetti]

Does anyone have a picture showing how they are wiring the original power supply to the RAMBO? I have snipped the 3 pairs of yellow/black wires and connected to the 6 pin connector block as shown in Geneb's manual. But from reading comments here, I should add an extra pair or two of yellow and black wires to the hotend set of wires?

I have a pair of 18guage speaker wires running from my RAMBO to the Onyx, but from reading various comments here, it sounds like I need to add more power from my power supply to the hotbed source input? Kinda confused - a nice higher resolution shot would help. I'm not interested in swapping out the power supply just yet for something beefier. Though I will admit I was shocked to find this printer came with a standard computer power supply..... You'd think they would have sprung for something more powerful.....

[geneb]

Step 22.4:

"Note that the picture only shows one black and one yellow wire in the “Heater Bed” positions. To get the most out of your power supply, it’s recommended that you put three yellow and three black in the “Heater Bed” (the two terminals on the right) positions to get more power to the heated bed. I recommend twisting three together and then soldering them to ensure good contact. (DO NOT SOLDER THE BLACK AND THE YELLOW TOGETHER. MUCH SCREAMING AND FIRE MAY ENSUE.)"

[cassetti]

Step 22.4:

"Note that the picture only shows one black and one yellow wire in the “Heater Bed” positions. To get the most out of your power supply, it’s recommended that you put three yellow and three black in the “Heater Bed” (the two terminals on the right) positions to get more power to the heated bed. I recommend twisting three together and then soldering them to ensure good contact. (DO NOT SOLDER THE BLACK AND THE YELLOW TOGETHER. MUCH SCREAMING AND FIRE MAY ENSUE.)"

But if I put the 3 yellow and 3 black in the heater bed position, what about the other power for the RAMBo? What i'm confused about is there are 3 yellow and 3 black. If i put all 3 pairs in the heated bed, where do I get the other 2 pairs for the board?

That's why I'm asking for a picture

[cambo3d]

Step 22.4:

"Note that the picture only shows one black and one yellow wire in the “Heater Bed” positions. To get the most out of your power supply, it’s recommended that you put three yellow and three black in the “Heater Bed” (the two terminals on the right) positions to get more power to the heated bed. I recommend twisting three together and then soldering them to ensure good contact. (DO NOT SOLDER THE BLACK AND THE YELLOW TOGETHER. MUCH SCREAMING AND FIRE MAY ENSUE.)"

But if I put the 3 yellow and 3 black in the heater bed position, what about the other power for the RAMBo? What i'm confused about is there are 3 yellow and 3 black. If i put all 3 pairs in the heated bed, where do I get the other 2 pairs for the board?

That's why I'm asking for a picture

from the other yellow wires in your power supply harness. you can use all and any of those wires, all yellow wires are 12v, all black wires are negative.

3 yellow 3 black for your heat bed power input
1 yellow and 1 black for the other inputs.

[geneb]

Now THAT explains the confusion! Thanks!

What you need are three yellow/black in the heater bed position and one yellow/black in the other positions. The only difference between this and the photo in Fig 116 is that you're adding two additional wires to the +12v (yellow) and GND (black) heater bed end of the connector.

Would this text make the whole thing more clear?

"In the end, you should have three wires yellow wires (12v) in the in the Heated Bed “+” position and three black (ground) wires in the Heated bed “-“ position as well as one each in the other positions, for a total of five yellow wires (3 in 1 position, 1ea in the other two positions) and five black wires (3 in 1 position, 1ea in the other two positions)."

tnx!

g.

[cassetti]

Now THAT explains the confusion! Thanks!

What you need are three yellow/black in the heater bed position and one yellow/black in the other positions. The only difference between this and the photo in Fig 116 is that you're adding two additional wires to the +12v (yellow) and GND (black) heater bed end of the connector.

Would this text make the whole thing more clear?

"In the end, you should have three wires yellow wires (12v) in the in the Heated Bed “+” position and three black (ground) wires in the Heated bed “-“ position as well as one each in the other positions, for a total of five yellow wires (3 in 1 position, 1ea in the other two positions) and five black wires (3 in 1 position, 1ea in the other two positions)."

tnx!

g.

Thanks guys! That clears it up! If I recall, there is another plug with 4 wires - 2 black and 2 yellow. Maybe that's why I was having a hard time heating up to 75c! I don't care if it takes 20 minutes to heat up, I just would like higher temps to combat warping with ABS. Eventually I might upgrade to a heavier duty power supply, but for now I think I'll stick with the original.

Geneb - yes that text helps

[geneb]

git repo for the docs updated with the change, thanks!

https://github.com/geneb/Rostock-MAX-Docs

[JakeCShake]

Thanks geneb....I don't think we ever tried it that way. We had the original two wires and added to those. Been away from the machine for far too long (at my brothers house) but am looking forward to adapting the machine to all these new changes. Hopefully will get myself one in the near future......

[cjdavis618]

Just wanted to come back and update this page with some results on my tests.

As I said in the beginning, I was able to use one yellow and one black wire from the 6pin plug to feed the onyx and was able to reach 100c in about 30 minutes. After I added the other 2 pairs of wires to that, (So 3 black soldered together, and 3 Red soldered together) I was not able to get to temp at 100c. So I ran it like that for a while and used it at 85c. Still took 45 minutes to get there though.

I removed the pairs of wires and went back to one and then again was able to get it down to 25 minutes or so on 100C.

All of this was with factory power supply that came with the kit.


I then went to amazon and ordered this.
http://www.amazon.com/gp/product/B006QS ... UTF8&psc=1

Put that in place and adjusted the DC voltage from 24 down to 19.8 and ran it there. I connected it to the heatbed power input with no other changes to the Onyx wiring. Used a 14Ga wire from power supply to rambo, and then another 14ga wire from Rambo to Onyx. Now I can get the onyx up to 85C in less than 3 minutes. Working very well so far.

Hope this helps.

[cjdavis618]

Wanted to add one more thing to this.

Before I put the 24V power supply in, I had a suspicion that my borosilicate was not flat. So, I took the borosilicate plate off the top of my onyx and tested it both hot and cold. It was as flat as could be, but the onyx was far from it. My machinist level showed that the onyx was sagging severly in the center 6" of the heatbed. After pulling it off and looking closer, the issue was due to the heat softening the support plate below it. The support plate below the onyx had sagged nearly 1/8" down in the center from the heat after only about 24hrs of total print time at 85C. I took 2 small 1/2"X3/4" expansion springs that were rated for about 6 lbs and cut them in half with my dremel and cut off wheel. Then placed them below and around the middle circle of the "Snowflake" plate at equal distances apart, but right at the edge of the opening. This helped the onyx stay flat against the borosilicate and not sag under long prints. There is now enough pressure from below to push the onyx up against the plate, but not so much as to bow it during printing.

Just thought I would share this with you..