I Can't Get A Break

[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.

[printbus]

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?

[Zulfe]

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?

[printbus]

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?

[AbuMaia]

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.

[Zulfe]

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.

[Mjolinor]

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

[printbus]

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.

[usarmyaircav]

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

[Mjolinor]

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.