You are not logged in. Please login or register.
SoliForum - 3D Printing Community → Help/Repair/Maintenance → Melted Power Terminal
What does it mean when the 12v terminal melts? Is there something that I did wrong? The power had cut out once when I was turning the printer around. I moved the cable around until power came back on, then taped it down. Tonight it shut off again and wouldn't turn on. I thought the cable might have a break, and had a hard time getting one of the leads out, then saw why-
I wonder if this might be the problem - I found this in an ebay listing for a Sanguinololu
PLEASE NOTE The Sanguinololu design is excellent but if you are going to run a heat bed the traces on the Sanguinololu board are NOT sufficient to carry the current required to operate the heat bed. Jumper wires need to be soldered on the bottom of the board. If these jumpers are not there your screw terminal will heat up enough to melt. If you would like these jumpers installed before shipping let me know and I will add them for you FREE OF CHARGE.. here is a link to the information http://create3d.com.au/sanguinololu-ass … nforcement. If you purchase this board from me or elsewhere these jumpers will be required to run a heat bed.t
If I end up upgrading my heated bed, I was planning on running a simple relay to the heated bed, so that the Sanguinololu only switches on the relay.
I was looking at a 15A 12V power supply like this:
http://www.mean-well.com.au/shop/gs220- … 2-r7b.html
It's the same one the Replicator uses but with a 12V instead of 24V output. 15A of current instead of the 10A the Solidoodle power supply provides. The plug also has four pins, so I would solder in a new plug for this with one set going to the electronics and the other going through the relay and straight to the bed.
My cables were not fully in place after I got the printer. Wiggling the cord would connect/disconnect power as you describe. I ended up stripping a bit more of the rubber insulation to get the cord in there properly. On paper, the smaller the amount of contact you have in the joint, the more heat will be generated as the current flows through it.
I believe lawsy has the best solution with the extra relay to control the HBP.
Do you think dielectric grease might help?
I couldn't get the terminal off, so I went ahead and ordered an Azteeg X1. It is equivalent to the Sanguinololu, also using the ATMega 1284. I got to look one over at Deezmaker, the store opened up by the guy who developed the Bukobot (http://www.kickstarter.com/projects/dee … -source-3d)
He said the motor drivers are better quality than the Pololus, and don't run as hot. Also it comes with pins for running a gcode controlled fan, and headers for the SD card. I don't think it would work with the panelolu, but the developer is working on a LCD panel. They had a prototype at the store, and it looks pretty slick.
The other important thing is the bed and extruder heaters connect to screw terminals, and are more robust than the Sanguinolu.
Got my hands on a complete Ramps 1.4 kit that I will try to fit next weekend. Spent this weekend making a adapter for fitting it on the solidoodle 2 without drilling new holes and stuf. Figured out most of the wiring too, but then, off to work for a week 
Really wanted a lcd and fancontrol from a mos rather than just running it from the 12v rail...
The big question is should us future solidoodle owners consider soldering in the jumpers if this is a known sanguinololu issue?
The big question is should us future solidoodle owners consider soldering in the jumpers if this is a known sanguinololu issue?
It would probably be better if solidoodle did this themselves, otherwise people will be burning boards and they'll have to be replaced at solidoodles cost.
I wonder if this shortcoming of the Sanguinololu is mitigated by the power supply. I can get the bed up to 95 on a good day, it only hits 100 if it is mostly covered, or the garage ambient is something like 95F. People often go 110-120 with other printers, and it might be that 100+ range that is a problem for the board.
It may be the power supply is not strong enough, or it may be the board traces are not strong enough to handle it, or a combination of both. That is why jumper wires may help out some. But since I have not gotten my printer yet, it's kind of hard to tell 
I don’t know how much current is needed for everything running, but that type screw terminal that melted are usually rated at 6-8 amps max.
The consensus at the RepRap forum, where I also asked, was that the melting was likely caused by a bad connection which caused resistance and a buildup of heat. This particular problem wouldn't be the heated bed, but that was brought up as an issue with Sanguinololus. One person thought that the block looked like the kind that is only rated for 6-8A, and not the 10A you would get at full heat and speed. The power cord wasn't taped down to the back of the printer, so turning the printer around to get at the back may have pulled one of the wires loose enough to start building up heat.
A went ahead and bumped my order to an Azteeg x3, which is equivalent to Ramps. It supports a second extruder, whose MOSFET could be used to run a controlled fan instead if desired. There are also an additional 4 expansion headers designed to be used for fans or LEDs, there is a built in SD card reader, and additional areas for expansion. They've been experimenting with adding bluetooth, using a radio they found for $8.
One other feature I really like about the X3 is it supports 6 endstops, so you can put switches at both ends of each axis.
devilman2075 wrote:The big question is should us future solidoodle owners consider soldering in the jumpers if this is a known sanguinololu issue?
It would probably be better if solidoodle did this themselves, otherwise people will be burning boards and they'll have to be replaced at solidoodles cost.
I agree completely it should fall on solidoodle to add necessary jumpers, but if they don't and itll add some added security that you won't burn down your house, I think itd be a worthy addition on the users expense. I can't imagine it really being an issue if its not supplied this way to begin with. They have got to be running their printers pretty extensively and if it hadn't become an issue for them maybe it was just a loose cord
The Azteeg x3 looks like a much better built PCB for handling higher current.
boosting the tracks is not going to help with that problem, thats a connector failure due to high resistance for sure.
I suggest when first using the device and on occasions, test the terminal with a patented heat monitoring device known as an index finger and see if its getting hot. if it is, make sure the screw is tight..
I will be finding a spare 12v relay and getting that 10A off the pcb entirely if I can, is it a slow switching thing or does it use PWM to hold the temperature ?
I believe the current algorithm is on/off only.
The officialMarlin has just released a PID loop bed temp update and I'm investigating before I merge it to my Solidoodle 2 firmware.
I have a PID controller on my IR oven for soldering. I imagine the heated bed is going to run similar loop times and a relay is fine then. cycles are usually 10 seconds or more, an external mosfet would be much better anyway, no annoying clicking.
I will be making small pcbs for little upgrades like this and throw them on ebay pretty cheap, i already do that sort of thing now. I was thinking the LCD upgrade could do with a pcb.
I have no problems with keeping it open source and will chuck eagle files up here later.
Stoney, what do charge per PCB for small production runs?
I've been doing some Arduino projects and am interested to know.
it's a very well known issue that the power components of the board are under spec'd.
it's been discussed on the Reprap forum, and the solidoodle google pages.
over at reprap there are some guys who will only sell kits, (not assembled products) specifically because they consider them dangerous.
if you were going the route of printing a board, might be better to re-design the whole board with beefier components?
Just started wiring up my Ramps board and everything went well until the power connectors. The thing has a seperate 12v rated at 11 Amps input that I need to provide in order to run the heat bed. So I guess there is some hocus pocs going around in the heat bed of the solidoodle, or is the design different when it comes to the heatbed?
Have an extra Ramps heat bed laying around that I can fit, so thats not a real problem.
If I want to power the steper drives, and the heated bed I need atleast 16Amps. The PSU wit the solidoodle provides 10A.
It's still a few days before my new board comes in, so I thought I would try to replace the melted connector, but I can't get it off. I have a 30watt solder iron, braid, a solder sucker, but I can't get the solder holding the connector to melt enough to remove. It gets a little shiny, and a little malleable, but it won't flow. Is there a trick to getting components off of a pcb?
Heat both terminals at the same time. Try to apply a little solder to get more heat transfered
Depending how the board is made (internal power planes and trace width), your 30 watt soldering iron may not be able to transfer enough heat to get the terminals off. Adding some extra solder will help some. Worst case, trying cutting the terminal block apart so that you only have to heat one pin to get it off instead of two pins at the same time.
I wonder if this might be the problem - I found this in an ebay listing for a Sanguinololu
PLEASE NOTE The Sanguinololu design is excellent but if you are going to run a heat bed the traces on the Sanguinololu board are NOT sufficient to carry the current required to operate the heat bed. Jumper wires need to be soldered on the bottom of the board. If these jumpers are not there your screw terminal will heat up enough to melt. If you would like these jumpers installed before shipping let me know and I will add them for you FREE OF CHARGE.. here is a link to the information http://create3d.com.au/sanguinololu-ass … nforcement. If you purchase this board from me or elsewhere these jumpers will be required to run a heat bed.t
This is truly disturbing. We shall look into this.
Regards,
John
IanJohnson wrote:I wonder if this might be the problem - I found this in an ebay listing for a Sanguinololu
PLEASE NOTE The Sanguinololu design is excellent but if you are going to run a heat bed the traces on the Sanguinololu board are NOT sufficient to carry the current required to operate the heat bed. Jumper wires need to be soldered on the bottom of the board. If these jumpers are not there your screw terminal will heat up enough to melt. If you would like these jumpers installed before shipping let me know and I will add them for you FREE OF CHARGE.. here is a link to the information http://create3d.com.au/sanguinololu-ass … nforcement. If you purchase this board from me or elsewhere these jumpers will be required to run a heat bed.t
This is truly disturbing. We shall look into this.
Regards,
John
a few months ago now I made a post on the google groups that there was an issue with the sanginololo electronics.
I link to some posts by nophead in the reprap forums.
I was going to put a link to that post here but I really can't seem to find it.
no matter what my search times, even if I search for all posts made by me, I can't find it.
can't seem to find it now. (actually, it's a bit frustrating that a post, (especially a post saying keep an eye on your machines as the electronics are confirmed as underspeced and possibly dangerous) -has just vanished.
Anyway.
As said above, it's a well known issue, neither traces nor connectors are substantial enough.
An idea to improve the electrical safety of the machines would be to point to point solder a wire between the spurce pin and the bed plug such that the board (and it's under-spec'd trace) is bypassed completely.
or connect a relay to the bed heater control and have a power relay switch the power directly.
SoliForum - 3D Printing Community → Help/Repair/Maintenance → Melted Power Terminal
Powered by PunBB, supported by Informer Technologies, Inc.
