Heatbed, Powersupply and Z Axis issues.

Something I got back as a recommendation to also try is to move the fuses (rectangle yellow/goldish) away from one another. http://reprap.org/wiki/File:RAMPS1-3_DnF.JPG is what you are looking for. On mine, there were pressed together and laid over sideways a bit. The theory is that if you spread them apart from one another and provide some airspace there, that perhaps that would also prevent the bed from shutting off. I'm considering adding an indicator LED to mine so I can see if/when the bed shuts off, if it's the fuse having been tripped (aka a mechanical failure) or if it's a software issue (aka bed shutting off because of some other problem)... My heated bed shutoff was theorized to have occurred when the thermistor in my hotend failed... so I'm still waiting on parts before I can get my printer back up and running but I know it's an issue I'm gonna have to resolve too.

 

I will try that tonight then, I did notice that one of those fuses was bent twords the other. I have been just cruzing along with PLA until I get this heatbed thing figured out.
Now we are on to fixing some other issues, one of the other members posted about his X axis slipping and causing a drift like effect in his prints (think it was called "offset layers"). Well mine is doing the same thing, and that forum found a resolution for it . So it looks like i will be taking the printer apart to tighten the pulley on the motor shaft. An also making sure the Z axis is tight too. It does the same thing when there is a bunch of Z movement in one area.

 

in the software you can tell the extruder to retract move up a couple mm's and then move back down upon a position change. The setting is under the slicer settings and Printer settings you will see a box check that says lift extruder upon position change, extruder height 2mm lift . It was a default setting. Doing the 5mm calibration steps with .2 fill and a honeycomb pattern it moves up and down MANY times in a row while completing the blocks. The motor is making some odd noises. after a while of doing this it gets to about layer 12 moves up on a position change, then never moves back down. and keeps trying to print. So if it is not slipping what is that? Gear didn't feel like turning?
I fixed this issue by turning that setting off so the motor does not have to work as hard....

And tesseract I do mean both. X moves left and right.. it is skipping, there is some bad drift in the prints ive been doing. On the x axis, when it is doing a fill, it gets moving back and forth pretty fast (i tried turning down the speed, it helped a little though most of it is already on 35) the extruder will move to the right, then never move back left so its slightly off. the problem is worse at the base of the 5mm block step. It seems to go away the higher it gets into the structure.
The settings I used were from this post. (http://forums.robo3dprinter.com/index.php?threads/my-slic3r-settings.826/#post-4601) using the Iphone settings he had provided. Overall my prints have turned out okay using them.

This is the second time you have doubted me in my observations, and it turns out I was dead on with the heatbed not working in my first post. Lol give me some credit man.

 

Summary: Extruder cooing fan is causing a significant drop in bed temp to where the heater can't put out enough power to hold the set point. At first I thought the problem was caused by an electrical intermittent connection. Bed would hold temp in the Y home position, but then when I started printing temp would drop to about 55C over ambient. In my case with 21C ambient, bed would only hold 76C. I wired in a 0.010 Ohm current shunt and measured the on bed current to be 8.28A @ 11.2V. The bed PID does not PWM the heater. It cycles it full on or full off every few seconds as needed. The bed ON power never really changed from when I started the ABS print with a set point of 100C. ON was always about 93W even as the temperature was dropping during the print. Next I switched off the extruder fan, and the temperature start to "quickly" rise. This all kind of makes sense given the large surface area of the bed. During this time, I also taped a thermocouple to the top surface of the bed. It was reading about 16C lower than the PID thermistor sensor that is mounted under the bed. When I turned off the fan and let things stabilize, the thermistor and thermocouple (bottom and top of the glass) had almost the same temp.

Running cooling air over the board electronics will have the output FET run a little cooler. This might you a little bit of extra power to the bed.

Undertone* - "Yes that worked. Thanks for that suggestion." are you able to hold 100C during a print by increasing base air flow? wakari-masen

Not sure what it means that the bed won't hold 100C with the fan, and that temp drops after printing starts. Do other printers do this? 100C initial temp really helps my hairspray stick. Haven't tried any really wide ABS prints yet.

Is there a sticky posting some where containing one liner summaries and links to printer quirks like this? I spent multiple hours of experiments figuring this out; thinking I had a bad bed; only ending in a Duh. Be kind of nice to save other people from having to dig around the forum to figure all this out.

 

Jorbangr* I am currently using the retract setting as well. No problems. Try printing a cylinder an measure roundness. Check belt tensions. (My Y belt was set like a wet noodle). With steppers off, both X and Y should slide with little force and should not bind. X and Y should start to free fall when the machine is tipped on its side. Monitor extruder temp or run it hotter. I had some Gcode generated using PLA temp settings while printing ABS. Things got very unhappy. When you are done with a print, recheck Z by using command line G1 Z0 and check Z height to see if you lost any steps. Homing won't tell you anything and destroys the evidence. Can do similar for X and Y creeping up on them and listen for when and if the microswitches trips. Can also see microswitch endstop status with M119 from the command line.

 

Forgot about the thermal fuses on the Ramps. Cooling them will lower their resistance a tad under normal op. Key point is the measured ON bed power 93W ( 8.28A @ 11.2V) didn't change when the PID bed was ON. ON duty cycle went from like 75% to hold 100C to 100% with the fan on, and it still could only manage 76C. If in the best case by RAMPS cooling, we were able to get zero voltage drop across the supply wiring, fuses and STP55NF06L FETs, and the supply was at exactly 12V, and the resistance of the bed stayed about the same, the bed power would be 106W or 14% higher. Rough number say the bed could only maintain 84C at 100% duty cycle with the extruder fan on. Can't get there from here, with my bed resistance of 1.35 Ohms. Think I remember reading a post that the bed current budget was 9A, so this is in the ball park.

So if anyone else wants to investigate, carefully measure the voltage going out to the bed. You can't do better than the 12V supply. Before my current shut, I am seeing 11.3V (11.2+0.0828). In the mean time I'll ping the ROBO team.

 

There is an interesting RAMPS cooling fan mount on Thingiverse:
http://www.thingiverse.com/thing:112715

I'll probably be printing one of those out and mounting a small fan to it rather than trying to blow air under vs some other method of cooling.

 

May be usefull. If you got one of those which supply would you wire it to?

 

Matt - can you carefully measure the voltage going out to your heater wires on terminal block D8 when the bed cycles ON? How long does it take before the bed cuts out? Try to measure the heater voltage when the bed cuts out as well. On my board, the fuses don't have part numbers, but F2 is physically taller than F1 which makes sense since F2 supplies the bed which uses more power. Here's a data sheet MRF 1100 http://www.bourns.com/pdfs/mfr.pdf. These are not supposed to trip until 22A and can hold 11A @23C. They also have a rate life time of 100 trips at rated current. A fan might be good to get you going, but we should really figure out the root cause. LittleFuse 11A equivalent from Mouser shows their part should be good to 50C ambient and still handle 8.7A bed current.

Mouser - http://www.littelfuse.com/data/en/Data_Sheets/Littelfuse_PTCs_16R.pdf Part costs $0.80.

eBay - http://www.ebay.com/itm/5PCS-BOURNS...828?pt=LH_DefaultDomain_0&hash=item1e7a66de5c

 

tesseract - Agreed not robo3d specific. But we may be talking about two different things here. What do you mean by "cut out"? That was max temperature you could hold? or the resetable fuse was being tripped?

I like Matt's idea a heater LED.

Sounds like Matt is having his resetable thermal fuse cut out on his board. (...or is he getting a brownout if the power fet or fuse are getting too hot?)

In my case, my bed heater is going to max power (11.2V 93W) when the extruder fan is blowing over thermistor (opposite side of the board). Best case with 12V to the heater, it cannot maintain 100C under extruder fan. In the Y home position, it holds 100C all day, but it only when I move the extruder/fan near the center the part of bed opposite the thermistor, that bed duty cycle increases to max, ON voltage/power stays the same, and the thermistor temp drops. The controller starts to work really hard trying to heat the area above the thermistor, but it can't get there because the extruder fan is blowing on it. AND as the same time, other parts of the board over temp up to 115C. This is all ok and expected since there is only one bed heat zone and one sense point, and not a robo3d specific issue. Just unexpected to a newbie until you think about it.

Here's another possibility: The thermistor on my robo3d sits between two of the copper traces on the heated bed. I'll bet that if it was directly over one of the traces, it would think it sees 100C all day. Of course surface bed temp would be a completely different story.

 

tesseract - Hey man. Thanks for explaining. Double check you have two different sized thermal fuses, F2 should be taller than F1. Hour to recover... hmm. The fan on the power supply kicks in some time during the print right? I'll run a long print and put a thermocouple on F2 and on the top of the case over the ramps.

 

OK THIS IS PERSONAL NOW!

I had been running with the printer hanging off the edge of the table 'cause of space (I know) . Added thermocouples. Reoriented so base was fully covered. Walked away after 10 minutes. Came back and print died. ABS part came loose. Also cleaned off the nice thick layer of hairspray that had been building up on the bed. Thermal fuse for print bed tripped. Need to rerun, but last readings before I walked out were case temp of F2 90C, outside case temp above ramps 38.9C, ambient inside case 31.3C. External ambient 22C. Problem gets worse when extruder fan is on because at least for me, this cools the thermocouple and forces the bed heat is ON all the time, 100% duty cycle. We are talking about changing the trip point of F2 by blowing air over it. Easier fix might be to add a clip on heat sink to it. Let me get more data.

 

Print was no biggie. Good news is I can reproduce the same bed heater shut down failure as everyone else. Thinking about this, I changed airflow in the robo3d base, but I also reduced the ability of the Polyfuse to stay cool because I stuck the thermocouple on with a small piece of kapton tape. Reading more about Polyfuses. Had never designed one in before. Bigger devices like the MFR1100 we have are probably not going to be able to protect the power fet if there is a short circuit because of slow response time. Bourns data sheet is pretty limited, but NTE data sheet does spec response times. I hate having to add a fan to keep this board cool. Really want a vent to pull fresh air in from outside. The design is presently operating too close to its trip current at slightly elevated temp. Sometimes it can take hundreds of seconds for a device to trip. Other options are to use parallel fuses to increase effective trip current, add a heat sink, or go with a conventional one time 15A fuse.