Heatbed/Hotend cool off before hitting ABS temps

I have a MendelMax 1.5 with a RAMPS 1.4
When I start my printer warming up the heatbed and the hotend both start out getting warmer, and will reach PLA settings and be quite happy. When I set it for ABS they start to get warmer, but when the bed reaches 80C it starts cooling off, and then about 30 seconds to a minute later when the Hotend gets to 120C it starts cooling off as well. There are no error readings in Printrun, in fact when the temp peaks printrun shows nothing different. I am new, and this is my first 3D Printer, so I am at a bit of a loss and any help is greatly appreciated!!

you can try to adjust the power voltage a little bit up, the situation you described seems caused by the lack of the voltage.
good luck

What you are seeing it one of a couple of things:

1) Your power supply is cutting out because it's overheating, a voltmeter check at the power output will tell you if this is happening.
2) Your heated bed is pulling to much current and the resettable fuse is cutting out. A volt meter check at the B+ side of the heated bed wiring will tell you if this is happening.
3) You have some really odd PID settings in your firmware. If you have the defaults, this should not be the case.
4) You have the wrong thermistor set in your firmware and the bed is getting *very* hot (not likely at all).
5) Your hot end is shorting out as it heats up, tripping the fuse. If you have a board with plated through holes and an aluminum plate mount this can happen.
6) You have a bad connection in one of your wires.

Most likely, your power supply is cutting out. If it's adjustable, set it to a lower voltage. That will make it put out less power / less current so it does not shut down as quickly. Better yet in that case, get a good industrial 12V 30A supply to replace it. There are a lot of good brands and models, the Mean Well SE-350-12 is one that is known to do the job. They cost about $45 US.

Wow, thanks for the great comments Uncle Bob and AAA!
I checked the voltage at the power supply, it starts at a steady 12v and doesn't change throughout the entire process. At the beginning the voltage coming off of the RAMPS 1.4 is 11v in both the bed and the hotend. when the hot end stops gaining temp you can hear the power supply start to work a little easier, and the voltage at the RAMPS for the bed drops to 0, while the voltage for the hotend goes up to 12v. The hotend continues to rise in temp for another minute, and then you hear the power supply work easier again, and the temp in the hotend starts going down again. The crazy thing is that as the temp is going down, the voltage at the board stays at 12v for the hot end! I ran this test 4 times and once the Hotend managed to stay at 222 for a long time (it was set to reach 230 and never did) until I turned it off and ran the test again. I figure this eliminated #1 and #2 above.
Then I got to the PID Autotune. It is explained here: [reprap.org] This was my first time using it, and I want to describe it for any newbie (like me) who reads this topic later to help them understand what it is, so please correct me if my explanation sounds wrong. PID stands for proportional-integral-derivative and it is an algorithm that predicts what the circuit board should do to heat up the hotend or bed, and what it should expect. The PID Autotune heats up the part you are testing 8 times in a row to the temp you select and takes the measurements coming back from the thermester and predicts what the bed will have to do next time it heats the component up. Enter in the code given in Pronterface and when it is done it spits out values for the PID, which you then have to go back into Marlin and plug them in. It worked great for my Hotend so I plugged in the new values in Marlin, the old values were CRAZY off. I pushed Marlin to the Circuit board, and tried to heat up just my hotend alone, and it heats up to 221 and then bounces back down to 213 and just rocks back and forth between them both, so even though the PID was very wrong, it wasn't the cause of THIS problem!

I just ran PID on my bed, and after heating it up to 123.61 @:103 (past my defined 110C) and holding it there it said "PID Autotune Failed! timeout" Given how far off my PID was for the Hotend, I am sure that part of my problem with the bed is the same, but I can't finish Autotune! Something else is preventing it from finishing Autotune. Any ideas?

I don't have plated thru holes and my circuit board is sitting on plastic.
I don't think the Hotend is shorting out because PID Autotune was able to work over and over 8 times no problem. Any other ideas?

If your hot end or heated bed is cutting out, the auto tune process will fail. It does not understand a broken system. The default settings should work ok for initial setup. The only time you really need to change it at setup is if you have temps that swing all over the place.

If the voltage going to the heated bed goes to zero, your fuse is cutting out. That's not a good sign. Fix that before you try anything else. If you are loosing 1V (12V to 11V) on the ramps board, check your connections. You may have something loose. If so you can melt a connector. Since both voltages drop to 11V I'd bet it's the power input connector. It's also possible that the wire you are using from the supply is to small. The same might be true of the wire to your heated bed or hot end.

I replaced the wires connecting the power supply with the RAMPS with ones twice as thick as the ones that came with the printer, and the heated bed makes it to 110C without stopping and will hang out between 108 - 110C as long as I like! I set my multimeter to a more sensitive setting (I'm just learning how to use this as well) and it looks like the ramps is loosing about .50 volts from the powersupply to the connection for the hotend and bed. I checked all of the wires, and none seem loose. Between the RAMPS and the heatbed I am loosing about .20 volts. Should I get thicker wire for that?
My chain of voltage is: Powersupply 12.22v, connection to RAMPS 12.00v, Leaving ramps to Hotend 11.59v, leaving ramps to bed 11.17v, attaching to bed 10.90v I can't get access to wires in hot end. Am I in danger of melting something? Would changing my wires to my bed help with the voltage dropping at the board?

So now the bed is behaving completely as it should, but the Hotend just won't go above 220.7C unless I am running PID Autotune. Autotune ran great on the hotend before I changed the wires, so I just ran it again and uploaded the new values to the board, no change. I am running the Hotend by itself to track the voltage, and as it reaches 220 C the voltage leaving the board and going to the Hotend drops to 5v right when the temp drops, and then climbs back up over the next 10 seconds or so. Any other ideas what might be the cause? Thanks again so much for your help!! I would be lost without the RepRap community!

If you have a normal setup, you are pulling 10 to 15 amps when everything it heating up. Each volt you loose at a connector or in a wire could be 10 to 15 watts. If it's all at one connector, 5 watts is plenty to get it to melt. If the 1V is in a short wire (say 50 CM), that wire will get quite warm, you will be at risk of melting the insulation. Getting things below 0.1V at each connector is a good goal.

If the voltage is dropping to 5V going to the hot end, there is something wrong. Track back on the board with your volt meter. If you still have 12V into the board, loosing 7V before you go to the hot end means that there is a lot of power going somewhere. One thing to be sure of:

There's a supply lead to the hot end, it will stay at +12 always.
There is a return lead to the hot end. It will be 0, +12 or something in-between depending on what's going on.

It's very easy to get the volt meter on the wrong lead when you are in a hurry. I have *lots* of personal experiance with that.

Thanks so much for the info on voltage! I will definitely try to hunt this down. One thing I don't understand as a newbee is why don't I have this problem when I run PID? It ramps all the way up to and beyond 230C over and over with no drop in voltage or temp. If the problem is voltage, wouldn't it occur every time I heated up the board regardless of what program is heating it up?

The PID mode should not impact what you see on the supply side pin of the heater. It will impact what you see on the return / switch side of the heater. The PID impacts how the MOSFET switches on and off, not how the power is supplied to the top end of the heater. If you are seeing a 7V drop on the supply side, it’s likely to be unrelated to the PID.

If you have a 20 to 40W heater, there is a *lot* of power being lost somewhere in order for the supply pin to drop that far. Most times in that case you will see smoke....

I have now replaced all of the big load carrying wires with 12 gauge... It really seems like overkill, but now I know it's not the wires. I loose less than .04 volts from one end of the wires to the other. the Bed is still behaving great, but the Hotend still heats up to 220 and then the voltage drops 7v or so and the temp and voltage cycle together between 210C - 220C never reaching 230C. How do I hunt down voltage trouble if it is on the board, or could it be in the heating element? I have literally had my nose overtop of the board every time I have run it, and I have never seen or smelled the "Magic Smoke" (I heard somewhere that if you can grab up all the smoke as it emerges from the board and inject it back in everything will work again :-) I guess that now I need to learn more about my RAMPS 1.4 board and where the voltage drain could be comming from. Is there some kind of circuit breaker on it? What else could I look for?

If the voltage is going to 7V into the heating element and it's staying at zero volts out of the heating element and the supply is at 12.0V here's what's worth checking:

1) Voltage at the Ramps board after the power supply connector, should always be 12.0V
2) Voltage on the Ramps board after the fuse. Should always be > 11V
3) Voltage on the ramps board at the supply side connector to the hot end (labeled +), should be within 0.1V of the fuse output voltage.
4) Voltage on the wire coming out of the supply side of the connector headed to the hot end. Should be within 0.01v of the other side
5) Voltage between the ends of the wire going to the hot end. Should be < 0.1V

That's pretty much it. If you have 7V at the heater and 12V on the supply it's got to be in there somewhere.

Ground works pretty much the same way:

1) Voltage at the ground wire on the board before the connector 0V
2) Voltage between wire and ramps < 0.1V
3) Voltage between ramps connector and MOSFET ground side < 0.1V
4) Voltage at low side of heater with MOSFET "ON" < 0.5V.

Once the MOSFET starts it's switching process it will go to 2 - 10V on the return side of the heater (labeled D10 (or 9 or 8) / top side of the MOSFET. That's normal with the heater stuff running.

Are you still running default PID settings? If you are running something else, switch back to the defaults. If the learning algorithm messes up, you can get some strange numbers.

Bob, you have been so much help it's been great! Thank you. I think I discovered my main problem, it's my inexperience with the multimeter. No matter how good your advice is, with flawed input from me you had no chance! So here is what I was doing. When I reported voltages before I always put one probe on that terminals + and one on it's -. Only now did I try testing them with one probe on the ground of the powersupply and the other probe on the terminal I was testing. When I test this way the voltage looks great, with very little loss, and it looks like what you told me to do in your last post. I'm so sorry that I got things wrong early on and that they sent us down this rabbit hole, and that explains why I have no magic smoke. Does this make sense to you?

It makes perfect sense. Sorry I was not more clear in my instructions. I often assume things that really should not be assumed.

Problem possibly solved. I e-mailed the maker of my printer kit, and they suggested I had the Fan aimed at the head. Like I said, I'm a newbee! So I took off the fan and I got above 230 no problem! Then I put the fan back on the heat sink only and again I wasn't able to get to 230. Should I be using the fan when printing ABS?

The fan should blow on the body of the hot end. Blowing a bit on the heated end works if you have a 40W ceramic heater cartridge.

Great, moving the fan around helped out a lot. I now get my Hotend onto 230C solid as a rock, and I can touch with my finger any metal part that comes into contact with the plastic parts of the print bed, so no worries of overheating. Now I need to fix the bed. In the past I got it up to 110C no problem, but now it's just getting up to 78 and then falling off. ARGH! I just went back to check it again before posting and I cant get the bet to warm up at all! I power cycled the printer and restarted pronterface and got it up to 43C before it quit. This is running the bed only, not the hotend. Any suggestions?

Back to the same set of measurements. I think that using the voltage drop on your main +12 lead as an amp meter would be a real good idea. It's not going to be 100% accurate, but you will have a much quicker troubleshoot process. About all that is involved is soldering a couple of sense wires onto the + lead.

Quick check - finger on each of the connection points when things are acting up. The one that's hot is the one that's the problem.

The voltage on the bed is 11.6V, and the temp gets up to different temps every time I use it. If I power cycle everything and let it all cool all the way down I can get the bed to get up to 90C, but if I try again shortly after it was warm I only get up to th mid 40s C. When it gets to its top temp of the moment, the voltage drops to .36V at the bed and of corse the temp drops. What can I do?

If the voltage at the top end of the heaters is dropping to 0.36V then your fuse(s) are cutting in. When they get to much current they open up. Everything you are describing sounds like one (or both) of the PTC fuses deciding that there is to much current flowing.

It's fairly easy to work out whether the fuse of ( edit -- or) the FET is limiting the temperature. If it's the fuse, both sides of the heater go to about zero volts (relative to ground). If it's the FET then both sides go to about +12 (relative to ground)..

Edited 1 time(s). Last edit at 11/26/2013 01:13PM by uncle_bob.

Did you mean fuse -or- FET?

Both sides of the bed are dropping to near zero, so it must be the fuse, right? Is that one of the two flat tall yellow diodes? Could something else be tripping the fuse?

Edited 1 time(s). Last edit at 11/26/2013 10:11AM by Nununugent.

The big flat yellow things are the fuses. Just about anything could be tripping them. It's not likely that they both go out at once. One goes to the heated bed.(11A). The other runs everything else. If the "everything else" (5A) fuse goes you pretty much can't do anything (no motors, no hot end, no light on the LCD).

Edited 1 time(s). Last edit at 11/26/2013 01:14PM by uncle_bob.

It worked just fine for a few hours, I got a few prints started and failed for other reasons, but then I turned it off and walked away for an hour and came back. Then I started it up again and the bed crashed again. I did some quick measurements and I lost voltage on one side of the 11A fuse, and it was fairly warm to the touch, where the other one wasn't. So what are my options now? Do I try to replace the fuse to see what happens? Should I just get another RAMPS 1.4 board? Should I get another Arduino?

power cycled it and let it cool off and it started up and the bed warmed up to 110C, but the A11 is hot to the touch, where the other yellow circuit breaker is not. Is this normal? As I ran it the A11 has cooled off some with a stable bed temp of 110C

You should put an amp meter on the heated bed or on the power supply. Either the heated bed is pulling to much power or the fuse is defective.

I finally got the meter hooked up and ran it several times and I couldn't recreate the problem. I had the meter on 10A and I got a measurement of 11.66 which slowly fell to 8.88 when it reached temp and then dropped to a reading of 0.00 when it reached temp and then went back and forth between the two to maintain temp. The 11A still got hot to the touch, but it was no problem. I even got started on a print another few times, and the bed seems to be handling everything well. Perhaps I fixed it somehow? Are these readings making sense?
Bob, if you are ever in North eastern Massachusetts I owe you a beer!

Well it's been months since I was last in that part of the country ....I can drink a lot of beer

11A is correct for the hot end supply. If it cycles as you describe that's ok. If the yellow 11A PTC fuse is getting hot to the touch it's probably your real problem. Once they heat up, a small change can make them trip out. I *hate* those fuses. I'd pull it off the board and replace it with a new one, our put in an automotive fuse. With 11A through it, I would not think it should get hot.

Edited 1 time(s). Last edit at 12/04/2013 07:49AM by uncle_bob.

So, an 11 amp 1000 v fuse? I'll get one on the way home.

I'd get a 15A 32V(DC) automotive fuse and a little plug in holder. If you get a three or four position one you can replace both of the yellow PTC fuses and have an extra location or two for accessories. All auto fuses are fine for 12V, the only reason to go for a higher voltage is if you eventually decide to do 24 V or higher. You want a DC (rather than AC) rated fuse since the arc suppression process is a bit different with a DC circuit. Most auto parts places should have a couple gizmos to pick between. They are a common item when you are adding accessories to a car.

When you pull the yellow fuses, just put a nice solid piece of wire in their place. What ever the biggest diameter that will fit in the holes should be ok.