I Can't Get A Break

If you've checked my recent posts, I've been having trouble with filament recently. I finally fixed the issue and printed off some quadcopter pieces. I noticed the next morning that they had layer separation on them, which was weird being that they were in a small, closed space (better environmental heat) and they were short (better inner heat). When I tried heating up my bed for tonight's prints, it wouldn't heat. The screw-down connectors for my bed are melted. The board has burn marks on it and the connector is done. I can't think of what to do next other than buy a new board or find a way to run the bed off of D09.

Any tips?
Attachment 3046

Take some better pictures.

It doesn't look too bad but it isn't good to see. Certainly I have seen a lot worse.

You should be whinging at whoever built the printer as that is caused by not tightening the wires enough.

It seems that for some reason a lot of these 3d printer makers decided they should solder the ends of the wires that fit into screw connectors. Tinning the wires for screw connection is an absolute no, the wire should be stranded and not tinned with anything if it is to fit into a screw connector. If you do tin them then this is what can happen.

I built the printer myself, but I know for a fact that I tightened all the screws down very much. I've broken screws in the same type of connector in the past due an issue of over tightening in the past. Maybe vibrations caused the screws to loosen over time? I could easy be the problem too, as much as I'd like to think otherwise.

Did you put solder on the wires before putting them in the connector?

Like Mjolinor said, the picture doesn't show us much. Another possibility is that the wires to the heat bed were allowed to flex at the screw connector and wire strands slowly started breaking off, leading to remaining strands overheating at some point.

Sorry to hear it Zulfe, don't be discouraged by these few issues you're having, stuff like this can just happens when you build your own 3d printer! You'll get it working!

Personally, I've tinned my wires that go into each screw terminal because in my experience, connections pull out more easily and are weaker when left stranded. Regardless, whether they are tinned or not, I don't think this alone is the root of your issue. It is more important to make sure that all of your wire slack is anchored to the frame so the wires aren't pivoting at the board connection when the printer moves. Your issue is that that one wire touched another while both were in the connector either by loosening slightly after being pulled on by the printer or by having too much wire exposed between the connection and the insulation during the installation. This also makes me want to go back and check that my screw down terminals are still tight!

Before buying a new board, does it still provide all other functions except the heat bed? If so, could you remove the burned connector and solder the heat bed wires directly to the board? To test this I would power down your printer first and remove the heat bed connector. Once it's cleaned up command the heat bed on with prontorface and measure the voltage between the two pins on the board where the connector used to be, also being careful not to touch the multimeter leads together. If everything's working the best practice would be to order a new screw down connector to replace the burned one. If you need help finding a replacement I'm sure we could help you look.

UPDATE BEGIN
I recently redid all the wiring and wire routing on my printer and after testing the strength of tinned wires and untinned wire terminations in the screw terminal, I 100% agree with printbus and Mjolinor that the wires in a screw terminal should NOT be tinned. Especially if you are trying to fit more than one wire into the terminal. Make sure that you twist the ends of your stranded wire before inserting the untinned end into the screw terminal. Many thanks to printbus and Mjolinor for educating me!
UPDATE END

There's probably a NASA workmanship standard out there somewhere that details it, but Mjolinor is correct in how stranded wire should never be tinned when the wire will be crimped or tightened under a screw head.

Yes, it appears to provide a more solid initial feel for the crimp or the screw terminal. The problem is that solder isn't as solid as it seems. When put under pressure (as in a crimp or screw terminal), the solder will slowly creep and flow out of the way, loosening the connection. It's just a matter of time - maybe months or more, but it will eventually lead to a poor connection.

Thanks for all of your suggestions, guys. It really shines some light on this crappy situation. I'm actually quite lucky I didn't burn my house down when this happened, as my printer is located in my room's closet. If the wooden frame caught fire, who knows what could have happened.

I'll take the board off the frame and analyze it some more and take some pictures when I get home. Maybe we can find a solution that won't require me buying a whole new $50-$80 board/shield. Funds are tight right now and I'm really counting on my printer being up and running to create pieces for some of my current projects. I don't really see the usefulness of a screw terminal on a vibrating machine since the potential of the vibrations loosening the screw is always a possibility. I would've preferred a pin connector like the rest of the board uses for the motors and thermistor. I also assume that it was entirely my fault for this happening since there was a very small amount of exposed wiring sticking out of the connector, creating the possibility for this issue to occur. My first solution is to modify the board's code to run the bed off of D9, which is completely unaffected. In the pins.h file, the bed runs off of '5', so I'm not sure how to go about moving it to D9. I'll eventually upgrade to a RAMBo board anyways, but I just need a quick fix in the mean time.

Thank you everyone.

EDIT [10/14/2014 @ 9:30AM]: Colin just replied with:
"I have seen that before if the heat bed is struggling to maintain the target temperature. There are a couple things that can help with this, first make sure you have insulation under the heat bed, next a heat bed relay will be the biggest help, if you get the heat bed relay you can move your heat bed to d9, if you do not get a heat bed relay d9 will quickly get damaged if trying to run the heat bed. The relay will let you reach higher heat bed temps at faster speeds, otherwise 110c may be difficult to reach.

From what I have seen in the past this is just due to very high current, not from shorting. The heat bed technically is shorting the two wires out so it can't be protected.


Here is the relay:http://www.makerfarm.com/index.php/p...bed-relay.html"

Zulfe,

See this link for the replacement part. Search "J2"

http://reprap.org/wiki/RAMPS_1.4

It seems like it has the exact same connector in black, or a replacement two part pin connector that will also fit in to the board. The second one seems like what you were hoping would have been there in the first place ;)

If you don't have a heat bed relay, I don't think you can move the bed heater to D9. D8 gets 12v power through an 11-amp polyfuse to protect the board (and your house) from a short in the heat bed circuit. This is the larger of the two yellowish postage stamp looking things. D9 gets 12v power through the smaller 1/2-amp polyfuse. You'll open the small polyfuse and shut down your printer quickly if you try to feed the heater directly off D9. Ignoring that problem, there's another. The MOSFET switch for D8 has a heatsink on it. The MOSFET switch for D9 does not. You'd likely overheat the D9 switch by drawing heat bed current through it.

FWIW, when polyfuses are subjected to overcurrent, they go into a high resistance state acting somewhat like a fuse. When the overcurrent goes away, the polyfuse resistance will restore to normal. So, I somewhat disagree with Colin. Since RAMPS has an 11-amp polyfuse, the board layout should be designed to handle at least that without any issue. I don't know if it was or not, but this makes me lean towards thinking the problem was in the wiring to the board.

If you do have a heat bed relay (or do as some have and wire in an automobile 12V relay to use for one), you should be able to swap functionality of the D8 and D9 connectors by swapping the values for the FAN_PIN and HEAT_BED_PIN in pins.h. Double check your pins.h file. For motherboard type 33, HEAT_BED_PIN should normally be defined as 8, hence the heat bed connection to the D8 port.

When a heat bed relay is used, only minimal current related to the bed heater is flowing through the RAMPS board. Also note that when you use a heat bed relay, the on-board polyfuse is no longer in the loop so it can't protect against overcurrent anymore.

FINAL EDIT? The reason why a heat bed relay can heat up the bed faster is that it is more efficient at getting power to the heater. It basically uses a mechanical switch connection that will have less voltage drop than the on-resistance of the RAMPS MOSFET switch. The RAMPS board also has voltage drop in the "on" resistance of the polyfuse. There may also be more voltage drop from traces on the RAMPS board and the push-on connectors that are generally known to be somewhat imperfect. These voltage drops equate to power that never makes it to the heat bed.

printbus,

I understand. My board is currently running on type 99, I think. Will switching to 33 affect anything else or is it even possible to change specifically the pin for the heat bed? I just ordered a heat bed relay to use D9 and for future use because it seems that I probably should have used one in the first place. I like the 20020109-H061A01LF that gmay3 pointed out, but will I be able to solder a connector to the heater cartridge and heat bed relay safely? Typically for those pins the connectors are attached to a very high gauge wire, like 22 gauge for example. Will the connectors and wires be able to support the level of current going through them?

And will the polyfuse being out of the loop introduce any new possible dangers? Should I enclose the heat bed relay in some sort of plastic enclosure so that in the case of it shorting or this issue happening again the enclosure will melt on it? I'm actually pretty paranoid because of this.

Quote:

---

Originally Posted by Zulfe

printbus,

I understand. My board is currently running on type 99, I think. Will switching to 33 affect anything else or is it even possible to change specifically the pin for the heat bed? I just ordered a heat bed relay to use D9 and for future use because it seems that I probably should have used one in the first place. I like the 20020109-H061A01LF that gmay3 pointed out, but will I be able to solder a connector to the heater cartridge and heat bed relay safely? Typically for those pins the connectors are attached to a very high gauge wire, like 22 gauge for example. Will the connectors and wires be able to support the level of current going through them?

And will the polyfuse being out of the loop introduce any new possible dangers? Should I enclose the heat bed relay in some sort of plastic enclosure so that in the case of it shorting or this issue happening again the enclosure will melt on it? I'm actually pretty paranoid because of this.

---

There is no board type 99... At least not in the latest Marlin code.

I just realized I was confused on where the RAMPS board used the push-on connectors. I thought they were on all the wiring ports. Glancing at the printer now I see only the power input uses the push-on type. Some of my earlier words about push-on connectors don't apply. The smaller polyfuse is also 5 amps, not 1/2 amp as I initially posted. 5 amps provides enough capacity for the cartridge heater, but not the heat bed. At least the help is free.

If it were me, I'd stick with the WM7857 terminal block approach to eliminate any flakiness with the push-on connector scheme. Some have had intermittent issues with those on the power input connections and removed them. Regardless, according to the datasheet linked through DigiKey, the push-on connectors have a current rating of 12 amps, so no issue there. The datasheet doesn't specify a wire size, but the 12-amp current limit implies it should be able to accept a pretty large wire gauge - maybe as large as #14 AWG? The MK1 bed and cartridge heaters are likely prewired with #16. If you go with the push-on connectors, remember that you have to order two parts - the part that gets soldered to the board and the part that the wires are connected to.

I don't want to offer you direction on the polyfuse. Running without any protection would be like 99% of us that use a heat bed relay. Without any protection, you're depending on what the power supply will do in a short circuit condition. One approach to a compromise would be to add an in-line automotive fuse in the heat bed power circuit with something like a 15 amp fuse installed. That way if there's an issue the fuse would open up likely before anything burns up. You'd have to replace the fuse after that, but no big deal, right?

Thanks for the plethora of information printbus, it's been a real help!

If I were to stick with the terminal block, how can I eliminate this issue from ever happening again? Would it even be worth fixing if I have just ordered a relay? Would it be smarter to just go to D09 with the relay until I purchase an upgrade board like the RAMBo?

Whether you use the direct solder screw terminals or the pluggable connectors won't make a difference in whether loose wire strands or something shorts between wires. The biggest difference you'd be making with the heat bed relay is that any high current issue related to the heat bed is kept on the heat bed relay, not the main controller. As mentioned earlier, adding an inline fuse would be one way to ensure some protection beyond what the power supply itself will do.

Whether you move the heat bed control to D9 sort of depends on whether you're ever going to want to have a print cooler controlled by the slicer with that controller. We don't even know that the D8 is unusable - the picture isn't good enough to offer much input on that. I'd pull the RAMPS board off the Arduino board and inspect both sides of the board any trace or component damage. If you don't see any, reinstall it and test it, maybe using a light bulb or voltmeter as an indicator rather than the heat bed. If that works I'd then wire in the heat bed just for the shorter heat up time.

Since the 10-inch i3v comes with a heat bed relay by default, this must be an 8-inch i3v?

Heh, I've had one of the screw terminal blocks on the heat bed relay from MakerFarm melt. I don't know what caused it. I just unsoldered the melted block and soldered in a replacement, no problems since.

Yeah, it's the 8-inch i3v.

I've taken some more images with a better camera, though it still had trouble focusing because of how close I had to be. I tried to avoid disconnecting all the motors, switches, and thermistors, but I feel like I've taken a few good images.

I can't upload the images here because they're too big, so you can find them here:
http://s2.postimg.org/c4lxy58jt/DSC_0004.jpg
http://s14.postimg.org/ykriyxifl/DSC_0003.jpg
http://s27.postimg.org/cidl51wmb/DSC_0001.jpg

If you're still having trouble seeing the issue, I can take down the entire board.

That doesn't look too bad to me. Does it still work?

I agree with Mjolinor. This looks like the heat bed wires just didn't have a good connection at the screw terminals. The bad connection(s) led to higher resistance than normal, and the higher resistance led to heat dissipation that softened the terminal block and made the connection worse. The bulging of the terminal block in the DSC_0001 picture tells me the heat was coming from the screw terminal. Other than the deformation of the terminal block, the rest of the board is likely fine.

In other personal and professional efforts, I've used terminal blocks like these but I don't really care for them. For a given terminal block, they work better with larger gauge wire than they do with smaller gauge wire. The ones here are better, but some also seem to be designed for use only with solid conductors, not stranded. Here's my approach to ensuring good connections to this sort of connector. These recommendations would apply to both types of connectors listed in gmay3's link to the RAMPS parts list.

• If you have the option, go with larger wire gauge to provide more copper volume for the terminal to grip.
• When I'm stuck with a wire gauge that is pretty loose in the terminal, I'll usually strip twice the normal length so I can fold the end of the wire back over itself. This doubles the gauge of the wire the screw is binding to. I've had some extreme cases where I've folded the wire end more than once.
• Strip the wire carefully. If you cut or appear to nick any strands in the stripping process, cut off the damaged end and try again.
• After stripping, I usually re-twist the wire but that may be arguable. This will help prevent strands from not going into the terminal or strands that aren't gripped tightly by the screw. If the end of the wire is being folded over, fold it after the twisting and use a pliers to squeeze the fold tight.
• Be sure that enough wire insulation was removed so that the insulation is not entering the terminal connection. The screw can't grab the copper properly if the insulation is in the way.
• Don't tin the twisted wire. It's fine for the strands to have been tinned by the wire manufacturer. If the twisted wire is tinned, the solder will creep over time and slowly weaken the screw connection. EDIT: Here's a reprap wiki post regarding this - http://reprap.org/wiki/Wire_terminat...crew_terminals
• Tighten the screw firmly, but don't over do it.
• Inspect your work. Make sure there are no loose strands that didn't go into the terminal and that the wire insulation is clear of the terminal. If you can see it, make sure the wire passes under the screw, not off to the side where it may not be gripped as well.
• For loose cabling like wires to the hot end and heat bed, add a cable tie-down somewhere to minimize any wire movement making it to the terminal. Wire movement can shift the copper strands and weaken the screw connection. The part of the wire without insulation is already a weak spot in the wire; any movement there can stress the wire and lead to strands starting to break.
• Maybe a day or a week later, go back and retighten the screw terminals. You'll likely be surprised to find the connections are looser than you left them. For those that have tinned the twisted wires, it is especially important to go back once in a while and retighten the connections.

Would it work/be a good idea to use something like these?
https://www.google.com/url?sa=i&rct=...13467573610232

Quote:

---

Originally Posted by usarmyaircav

Would it work/be a good idea to use something like these?
https://www.google.com/url?sa=i&rct=...13467573610232

---

No, there is no need.

What you do depends on whether or not it is working and what abilities you have. If it is working and seems OK, not getting hot, not intermittent, then just clean it all very well and use it. If it is not working or is flaky then you need the board out and see why. The most likely reason is that the solder joint between the terminal and the PCB has gone tits up. You may be able to clean and resolder or you may have to pick up the track preferably by soldering directly onto the transistor leg. The track from the transistor to the terminal is pretty thin for the work it has to do anyway so adding a copper link to the terminal can't hurt as long as you are able to do that.

The spade lugs also just won't work here. They're for use on flat terminal blocks like on the 12V power supplies. Here, the screw terminals have roundish holes intended for direct wire engagement.

There's another motivator for finding a way to reuse the connectors that are on the board. The ground pins likely solder to a full ground plane. Depending on experience and the desoldering equipment on hand, they could be challenging to unsolder without causing internal damage to the board.

Definitely loose connection. Have to disagree about not soldering though. I soldered all electrical connections for 20+ years for house, commercial. and industrial wiring and had zero issues for circuits up to 440v 40 amps. Only stopped doing it when Code was revised. Then, surprise! started having some connection issues.

A loose connection will heat the wire and the solder can melt, granted, but stranded wire will burn up as well when connection isn't tight.

Quote:

---

Originally Posted by Steven57

Have to disagree about not soldering though. I soldered all electrical connections for 20+ years for house, commercial. and industrial wiring and had zero issues for circuits up to 440v 40 amps. Only stopped doing it when Code was revised. Then, surprise! started having some connection issues.

A loose connection will heat the wire and the solder can melt, granted, but stranded wire will burn up as well when connection isn't tight.

---

Steven57, I agree with you that in some other types of connectors having the wires tinned is probably ok. Especially if you can really torque down on the wire with a more heavy duty commercial type connector.

I recently redid my wiring and after testing the strength of this specific RAMPS screw down connector's grip on both tinned and non tinned wire, I 100% agree with printbus and Mjolinor that for this connector on the RAMPS board non tinned wires are the way to go. I have updated my previous post to capture this change of heart. Big thanks to printbus and Mjolinor for educating me!