Weird problem, has to be ramps or mega board

I posted this in the general forum but have no responses so im trying here as i suspect the problem could be hardware or firmware related.

I am using Pronterface, slic3r, and sprinter. Have a mega 2560 board with ramps 1.4 .

When i load a file in pronterface and hit print, it homes as it should then starts heating up. When it gets to within 2C below the target temp it will stop heating up and will never reach the target temp. No matter what i set the target temp to in slic3r it does this. If i manually enter a target temp in pronterface it will reach the temp as normal, just wont do it in auto while its waiting to print. So if i set the temp manually to say 5C above what i know the temp is in the g-code file and then hit print after the temp has reached above the g-code target temp it will then run a print. This is the only way i can print a part right now.

I connected to another computer and it still does it. I redownloaded pronterface and sprinter, reloaded sprinter firmware and same problem.

This problem popped up on me over the weekend, the first few times it did it i would shut down my laptop and restart and it would run. Next print it would do it again and would not run again after that. Now it wont run period without manual intervention.

Im stumped as usual. Anyone want to take a crack at it?

When you set it in Pronterface I expect it sends M104. When building I expect it uses M105 to set the temperature and wait. So it seems like a bug in the firmware to me. You could try typing those commands and see if that is the case.

Why it should suddenly occur I can't explain. If you are using PID it could be something has changed at the hot end that stops is settling at the right temperature but then why would M104 still work?

Edited 1 time(s). Last edit at 06/12/2012 09:57AM by nophead.

---

[www.hydraraptor.blogspot.com]

I just tried setting it in pronterface about 5c high so i could start a print and it did not reach temp. Stopped exactly 2c below the target temp as it does in auto mode. Im about to pull my hair out with this thing..

The problem with PID is it isn't guaranteed to settle on the target. That is why I use bang bang on my own stuff. It doesn't seem to work as well with Marlin so I used the auto PID tuning and that worked well. Perhaps you could use Marlin to get the correct PID coefficients if they are portable to Sprinter.

Or perhaps just increase the I term a bit.

Edited 2 time(s). Last edit at 06/12/2012 10:45AM by nophead.

---

[www.hydraraptor.blogspot.com]

I tried setting temps in Pronterface using M104 and M109 as you said and each time i tried it worked for both. I then set it using the set temp button and it worked as well. Heck i dont even know where to start looking. I checked all my settings in Sprinter and nothing looks amiss.

What do you mean by "Or perhaps just increase the I term a bit." ? What is the I term?

I just noticed you stated - "When you set it in Pronterface I expect it sends M104. When building I expect it uses M105 to set the temperature and wait."

Looking at the G-code for a sliced file it uses M109 to set the temp and wait. Not M105. I didnt try M105 i tried M109.
Was that a typo?

I just tried running a print after successfully setting temps using M104 and M109 and it still wont hit the target temp in auto mode. This is crazy.

Sorry yes it should be M109.

The I term is the I in PID, the integral coefficient. With just a proportional term there will always be an offset. The integral term is supposed to fix that by slowly ramping up until the error is zero.

Do you have a fan running in auto mode? Do you have a heated bed? You could try turning that on with M140 to see if it somehow affects it. As far as I know there is nothing different between typing commands and running g-code files.

---

[www.hydraraptor.blogspot.com]

Thats what im not getting here. Manually sending the commands seems to work fine but when you let it heat up in auto mode it dont get there. I do have a fan that runs non stop but no heated bed.

I am going to try the default settings in slic3r and see if it does anything different with the G-code. I dont know what else to try at this point.

I don't think there is an "auto" mode. Pronterface simply sends the g-code. If you type the first few lines it should do exactly the same thing.

---

[www.hydraraptor.blogspot.com]

I know theres not an auto mode. I have a hard time explaining things clearly so i was just using that statement as a reference to pushing the print button vs typing in commands and sending them.

I think i may have the problem solved. I read your post about a fan running and decided to turn off my fan. Whoala, it printed the first time i tried it. So i then turned the fan on using m106 s255, then it would not reach target temp and wouldnt print. So i tried this experiment a couple more times. Every time i tried to print with the fan on it wouldnt get up to temp, when the fan is not on while its trying to reach temp it worked evertime.

So maybe my hot end is getting weak and just not able to overcome the heat loss from the fan running all the time anymore. I dont know. Its never did this before so something has changed. but anyways i think its resolved.

Thanks, for the tips.

If you can set it to higher temperatures and it is always a bit lower than it should be then in means the PID coefficients can't cope with the fan rather than a lack of heater power. If it was power limited it would max out at some temperature and if you set it to less it would work and if you set it to more it would stay exactly the same.

---

[www.hydraraptor.blogspot.com]

I have no idea what the PID coefficients are, youre over my head but that aint sayin much. Its back to working so ill just change the slic3r cooling settings and live with it.
Thanks for the help.

It is this section if configuration.h:

//// PID settings:
// Uncomment the following line to enable PID support. This is untested and could be disastrous. Be careful.
#define PIDTEMP 1
#ifdef PIDTEMP
int pid_max = 255; // limits current to nozzle
int pid_i_max = 80;//130;//125;
double Kp = 2;//1.10;
double Ki = 0.01;
double Kd = 20.0;//0.01;

Kp, Ki and Kd are specific to each hot end.

Maybe Ki is too small or perhaps pid_i_max is too small for when the fan is on.

---

[www.hydraraptor.blogspot.com]

Ok, gotcha. Ive stared at that setting a few times and read the comment about untested and could be disastrous and moved on. I might play with it at some point. Whatever the issue it certainly has to do with the fan. A bit ago i was running a print and it turned on the fan for bridging and a minute or so later the temp had dropped to 85C and was still dropping. It ruined the print and i stopped it. But it got me to thinking. Ive never had any issue with this before so what have i changed recently. The only thing i have recently changed is all the wiring. My wiring was in dire need of being more organized so a few days ago i moved my mega and ramps board and rerouted all the wiring. Before i had the wires for the hot end, temp sensor and fan all seperated and loose. I now have them all running through the same loom and nice and neat. So im thinking i need to seperate the fan wire from the rest and see what happens. Its like the fan is putting out EMF thats getting picked up by the boards being closer in proximity as before or maybe getting picked up by the other wires now that they are all basicly wrapped around each other.

If you block the fan by taping some card over it or similar you will be able to tell if it is EMC or its cooling effect. I suspect the latter. Has the ambient temperature changed recently, or the position of the fan relative to the nozzle?

---

[www.hydraraptor.blogspot.com]

I have to get some prototype prints ran out for a customer as im a little behind shedule but ill play with it some more in a day or two and see exactly what the problem is. Ill post back when i get the riddle completely solved.