Printrbot Talk Forum

by **APS** » 2013-Mar-Fri-12-Mar

Hi, this afternoon my Printerbot Plus caught fire a little.

I bought the PB+ kit around three months ago and have been using it quite regularly since then, and I've been happy with the results. In the past few weeks I have done several large 8hr+ prints, usually overnight, with good results. This afternoon I set off another such print, with an estimate of 9hr40, and around 5mins in, whilst I was still watching, I noticed a crackling and rather acrid smell. I first thought it was some impurities in the ABS, but quickly realised there was a flame underneath the printer. I was able to switch the power off straight away and the only damage is a few burnt wires and a small burn mark.

See pics below. The wire from the power supply was sparking and has burnt almost right through. At first I thought it may have burnt my garage down if I hadn't been there, but I think perhaps it would have burnt through the power wire and shut down, not really causing any more damage than it had. Either way, I'm fortunate I was on the scene.

IMAG0419.jpg

IMAG0418.jpg

IMAG0420.jpg

So, my questions are:

[1] I am assuming this is a problem with the power supply. Will it be resolved if I simply buy a new power supply? Can anyone recommend a good (more suitable) model?

[2] Is there anything that could have caused this, other than a random fault with the power supply? Is there something about my hardware set up or maybe my printer settings that could have caused it?

[3] Before I get my beloved machine up and running again I'm going to fire proof my homemade box (MDF!). Ideally I think I want some kind of thin fire resistant tiling, can anyone suggest anything suitable?

I'm going to refer PBHQ to this post so they are aware of the issue.

Any advice/assistance in diagnosing and resolving these issues would be most welcome so I can get printing again!

Thanks

APS

**Sponsor**

by **RetireeJay** » 2013-Mar-Fri-13-Mar

Yes, you are fortunate that you were there and alert to the problem!

APS wrote:[1] I am assuming this is a problem with the power supply. Will it be resolved if I simply buy a new power supply? Can anyone recommend a good (more suitable) model?

The short answer is no, your power supply did not cause this problem. Many people have upgraded their power supplies, but not for this reason.

When electricity flows through some kind of resistance, it creates heat. Even wires, which we think of as having "no" resistance actually have some finite resistance, and when too much current flows through them, they can get hot. But especially connectors can get hot, because the metal-to-metal contact is not always as perfect as we'd like it to be. Sometimes there are assembly defects where the metal insert in the connector was not properly crimped to join the wire, creating resistance right at that point. From your picture, that seems to be where the most heat was being generated - exactly where the wire met the connector.

In the photo, you show only one yellow/black wire from the power supply connected to the 'bot. But maybe you had made some changes during the emergency situation... The 'bot comes with a Y power connector cable, and you are supposed to connect TWO wire pairs from the Power Supply to the two plugs on the Y, while the tail of the Y connects to the Printrboard. When you do it this way, the current from the power supply gets shared between the wires and connectors, reducing the stress on each individual wire and connector.

You must replace ALL of the wires that were damaged by heat. I'm sure PBHQ will send you all the replacements you need, when you identify them.

By the way, take a close look at the connector & wires for the heated bed. Many people have observed overheating there. If it has NO signs of overheating, then that would be good news.

by **APS** » 2013-Mar-Sat-04-Mar

Thanks for replying, I think you've just diagnosed the problem - the second connection to the Y cable! It's hard to know exactly what I unplugged in my panic, but I'm pretty sure there was only one connection. I did initially have both ends connected and all I can think is that when I moved the machine out of the house a couple of months ago I didn't connect it properly!! That being so, is there anything reason this happened now and not sooner, has this just been waiting to happen for weeks?!?

I've checked the heated bed area and there's no sign of any heat or problems.

So, I need a new power supply and Y cable. The only other damage (pictured) is some slight melting of the plastic on the Z motor wires. You've suggested replacing these wires as well, obviously they're connected to the motors, will it be adequate to cut out the damaged length of wires out and re-connect them?

Thanks for your help.

by **RetireeJay** » 2013-Mar-Sat-10-Mar

APS wrote: is there anything reason this happened now and not sooner, has this just been waiting to happen for weeks?!?

In my work in industry, I've seen plenty of failures that result from slow-acting mechanisms like corrosion & vibration. Naturally, the failure also happens where there was also a weakness to begin with.

APS wrote:I've checked the heated bed area and there's no sign of any heat or problems.

The problems that people have been having are most commonly located at the connector where the wires for the bed plug into the Printrboard rather than at the heated bed itself (although there have been reports of broken wires at the bed, or overheated wires between the bed and the Printrboard).

APS wrote:So, I need a new power supply and Y cable.

Agreed. Technically, the power supply itself might be OK if it still has undamaged yellow/black wires. But if the failure was due to poor assembly of the connector you were using, then the other connectors are suspect too. So I agree you should use a new power supply.

APS wrote: You've suggested replacing these wires as well, obviously they're connected to the motors, will it be adequate to cut out the damaged length of wires out and re-connect them?

If you are experienced with splicing wires, that could be OK. You need to solder the splices and then cover them with heat-shrink tubing. I had failed to consider that since the wires are permanently attached to the motors at one end, splicing is the only way to repair the wires without also replacing the motors. So I agree.

by **plexus** » 2013-Mar-Sat-14-Mar

PBHQ recommends using the PSU Y adaptor and connecting to 2 rails on the PSU. typically ATX PSUs will run 2 parallel 20AWG wires to each 12V rail connector. read on why using 1 rail alone will create a Firebot...

My PB+ draws about 16A at 11.1V when everything is running (heat bed, extruder and motors). that is 180W so call it 200W to be safe. a wire typically has a specific thickness or gauge. the most common measurement is AWG. this is the chart i use. the wires from my PSU and to the PBrd are 20AWG as supplied from PBHQ. 20AWG is rated for 11A. so doubled up you get theoretical 22A of capacity. however the doubled up wires are close to each other and so they will heat up because if my PB is drawing 16A then thats 8A per wire (2 wires) and the rating is 11A - 8A is very close to the rating so the wires will get warm. put one warm thing next to another warm thing and the ability of each to dissipate heat is reduced (eg. gerbil pile). and so doubled up 20AWG wiring expected to take 16A total is likely going to burn at some point.

The best thing is to use more parallel wires or better yet a single thicker wire rated well higher than what you expect to use. on my PB i have the 2 12V rails of the PSU joined together to a single XT60 plug. that means i have a total of 4 20AWG wires on each + and - for a capacity of 4 x 11A = 44A. they still get warm on full bore but at least not nearly warm enough to melt. but you can see, even 4 20AWG wires with a draw of 16A are still going to get warm. if you want them to run cold you have to basically go with 4x the current you are drawing in amperage capacity. 16A x 4 = 48A = 10AWG. that is one thick wire!

to the heat bed again the stock config is 2x 20AWG. now if you have these bundled together or in a covered sheath (which can include wrapping of any kind), then the ampacity will be significantly reduced. here is a chart showing open air vs enclosed ampacity. you can see the 20AWG wire is reduced to 7.5A when enclosed. so doubled up thats 15A. you might as well call two wires sitting next to each other running across a table "enclosed" so 16A through 2 wires with a total rating of 15A... poof. :twisted:

even 16A through 4 rated for 30A is still going to get warm but at least we are far enough away from max that hey wont burn.

connectors are also a concern and have a max current rating as well. so you have to consider that. on my PB I ran 14AWG silicone insulated wire from the PBrd to an XT60 jack on the back of the Printrbot. then i bundled up the 4 + and - wires from the PSU to another XT60. XT60 connectors are rated for 60A. this way i can easily disconnect the main power to the PB. externally i have 30A total of wire current capacity and internally its 32A with 60A connectors. i don't want these overload problems.

to my heat bed i removed the paired up wires and used 14AWG wire again to an XT60 connector going to a mating XT60 connector on the PBrd. when you are dealing with power you have to ensure all your wiring and connectors are well over typical operating currents because unless you are around 4x over operating currents with open air wire, the wires are going to still get warm. the heat reduced the current capacity further, further heating them, further reducing capacity, etc etc until :twisted:

Final note on insulation... PVC is typically used and is flamable at lower temps. i prefer to use silicone wire. silicone has a much higher flash temp and so will prevent fire. i choose a PSU that uses silicone wire on the outside. silicone coated wire is a littler harder to find but its worth it.

Wire: RetireeJay points out that stranded wire with a high strand count (more individual wires bundled in the conductor) is preferable to wire with less strand count (and thicker individual wires). high strand count provides more flexibility in the wire which is important for wires that will incur movement and stress due to motion, such as those connected to the heat bed and the X axis. This where I get my wire: http://www.hobbyking.com/hobbyking/stor ... Guard.html . their silicone wire is pretty decent and has the thinner conductors i mentioned. its mainly used in radio controlled models which are always vibrating and crashing so the wire has to stand up to this.

Please don't compromise on these things and if you feel your Printrbot is in danger of becoming a Firebot, replace all the high current wiring with at least 14AWG or more silicone insulated wire and upgrade your connectors.

by **Mooselake** » 2013-Mar-Sat-19-Mar

Don't forget to have one of these close to your Printrbot.

0316132011.jpg

Kirk

by **Michael@Oz** » 2013-Mar-Sun-17-Mar

This is probably a good spot to remind viewers of Safety in general.

by **RetireeJay** » 2013-Mar-Mon-08-Mar

It is possible to run a much safer operation.

I don't use any acetone at all. Nor floor wax, hair spray, etc. No flammable liquids are anywhere near my 'bot.

I run my bed at room temperature. The heater is unplugged from the Printrboard (and will be completely removed the next time I have any cause to work on the bed). Thus, I will never have the heated bed plug overheat. In fact, the load on the wires from the power supply is significantly reduced so the failure noted at the top of this thread is less likely.

Although I print with nylon exclusively, it's possible to use the same technique with PLA or ABS.

The answer is to print on canvas.
viewtopic.php?f=74&t=2791

by **plexus** » 2013-Mar-Mon-16-Mar

I have tried canvas as well to print ABS and PLA and it does work unheated. however the reason we use a heated bed is not for adhesion (at least I don't), its to prevent curl when the lower layers start to cool as the upper layers are added hot. by heating the bed you help to keep the print warmer and reduce curl. i have not been able to get ABS and PLA (let along nylon) to completely adhere to the canvas without the application of heat to keep the bottom layers contracting and lifting.

by **RetireeJay** » 2013-Mar-Mon-19-Mar

Let's continue this in the thread for printing on canvas (in the Materials forum)

by **plexus** » 2013-Mar-Mon-21-Mar

cool. ill move these over there... see you in Materials. lab unit 36.

by **plexus** » 2013-Mar-Tue-00-Mar

Just and FYI, the wires coming off my PSU are 18AWG silicone coated, there are 4 per positive and negative. they are not bundled up but sit close to each other. i measure 42C if i put the sensor in the bundle of 4 wires where they bundle up towards the connector. the ambient temp is 23C. according to the chart 4x 18AWG = 64A. printer draws max 16A when full bore so you can see how wire rated for 4x the current still gets 20C over ambient! :twisted:

by **David@Printrbot** » 2013-Mar-Tue-18-Mar

@APS - Please send me a PM with your mailing address, email, and a list of the parts that need to be replaced.

by **plexus** » 2013-Mar-Wed-16-Mar

This was just posted to the Marlin github. pardon the poster's english skills, i copied it verbatim. this is a good point - if there is a problem with the hotend thermistor such that it gets a cold reading from an otherwise energized and hot hotend, it will keep power to the hotend until... catastrophe. i like this guy's suggestion of a fail-safe that will cut power to the hotend if it gets a "cold" signal from the thermistor which is not rectified in a certain time limit. one could disable this if they wanted but i think in general its a good idea...

recently I got a misadventure which could put fire to my home. the kapton holding the thermistor start to peel off so for marlin the temp go down and he use the full power to heat the hot end. which go far over the peek possibilities so the hot part pop out the thread in the peek, the tubing go out of the peek , befor to cool in a weird shape. til now it is not abnormal, the following is not the optimal behaviour on a safety point of view.

the thermistor peel out completly so the cold extrusion prevention kick in and stop extrusion. But the power is still on full power , the hot end was getting hotter and hotter making a mess of the part partially printed (which i don't care) but making a lot of smelly and irritating smoke and couls have taking fire.

Is there a possibility to add as a safety that if the cold extrusion prevention kick in, one minute after, the hot end power is stopped and the print ended ? the printed part is ruined anyway.

I don't know if my story is a frequent one, but if a thermistor wire break down the result will be the same. even if is not frequent the risks are enough to make the software with a fail safe against it.

thank you

by **lwalkera** » 2013-Mar-Wed-20-Mar

AFAIK, Marlin already has that. Search the config for MIN_TEMP.

by **plexus** » 2013-Mar-Wed-22-Mar

lwalkera wrote:AFAIK, Marlin already has that. Search the config for MIN_TEMP.

you know, you are right! of course. if the extruder goes below the min temp and its turned on in the FW, it will give an extruder temp too low error! i guess this guy didnt have that set. thanks Laine!

by **plexus** » 2013-Mar-Thu-00-Mar

I have something odd to report that may have resulted in the issue that guy had with the extruder heating up even thought the temp was low. in my issues with flashing the FW in my PBrd, i switched between a version using Lincomatic's fork with the Viki LCD support that he posted not that long ago and the previous version before he made the LCD updates. the newer version does not give me the all caps extruder temp too low warning. when i switch to the previous version, it does! I am thinking this might not be something Lincomatic did because this other guy reported it and I got that from Erik's Marlin github. i am wondering if something changed in the current version of Marlin that disables this fail safe?????

by **phil_roberts** » 2013-Apr-Sat-16-Apr

My plug melted as well. I have an LC (not plus) so I don't even have the Y connector issue. I've only run my bot for ~50 hours, mostly with the heated bed at a nominal 65 deg C. I did not have a fire, but I think this means no more Printrbot on the wood desk in the living room. PS is a Diablotek ATX 250W unit.

I wanted to move to nylon anyway- looks like I'll have to learn the finer points of canvas sooner rather than later.

IMG_20130413_150709.jpg

by **REPRAP SQUAD** » 2013-Apr-Mon-12-Apr

This is what I've done for an early warning of fire issues. I have heat sensors tied into a loud alarm horn. If 2 or more of the heat sensors reach a pre set temperature it sounds the alarm as well as send me a text. I also have an IP setup camera to visually inspect as its printing with a wireless tablet or I can use my phone as well as my computer. That's teamed up with readily available fire extinguisher. I never leave it printing unless someone is at home. It works well, so far only a few alarms when I was getting the sensors setup to the pre set temps. This kind of issue will always be around no matter what reprap you own or have built. You can only be proactive by using correct wiring and correct wire diameters. Be safe and keep a fire extinguisher around no matter what and make sure it is rated correctly. Also a smoke alarm should be standard running gear, happy building my friends!!

by **bosh34** » 2013-Jul-Thu-02-Jul

yes,thanks,but can show some more photos for us to check?

by **plexus** » 2013-Aug-Wed-20-Aug

Just reporting that the power connector on my PB+ fried. one of the ground pins burned up. looks like it was over heating for some time. i replaced it with an XT60 same as the hot bed which also burned up.

by **Jason** » 2013-Sep-Thu-13-Sep

Mine fried as well, right at the connection between the board and the board mounted connector. Seems like The fire also took the two mosfet with it. Looking for new mosfet to order now, but the fact that the printrboard wont respond to my computer worries me a lot, hope the mosfet and the connector are the only fried part

by **Michele31415** » 2013-Oct-Thu-00-Oct

FWIW, I had the same problem as the OP with my Printrbot+. There was nothing in what laughably passed for "instructions" that explained you had to use the "Y" to connect both connectors to the power supply. I saw that thing and figured it was just an option depending on what style connector your power supply had, so I just plugged in one connector. And the wire got hot enough to melt the plastic mat the printer was sitting on.

IMAO, this is a poor design. Something this safety-critical should only be able to be assembled one way - the right way. And even then the safety margins on the wires are sketchy as Plexus pointed out.

I don't know how Brooke can sleep at night knowing there could very well be a lawsuit in his future after a Printrbot burns down someone's house. I notice this machine isn't UL listed. I think my next accessory will be a smoke alarm to sit right next to it.

by **REPRAP SQUAD** » 2013-Oct-Tue-16-Oct

That's why I made a mount for this little guy. I also have text messaging sent when smoke or fire starts as well as IP camera. I also recently added a full graded fire extinguisher within 15 FT of my prototyping lab at all times. It's just a good idea when dealing with these temps and power requirements.

by **Porter** » 2013-Oct-Thu-19-Oct

Just reading through this post, and I don't know if this applies to the newer PSU that ship with the Printrbot+ V2, but I didn't get the Y adaptor in my kit.

Has something been put in place in the newer models or should I be looking at purchasing one of those cables so that I the change of a fire is reduced?

Daniel

by **plexus** » 2013-Oct-Thu-19-Oct

Porter wrote:Just reading through this post, and I don't know if this applies to the newer PSU that ship with the Printrbot+ V2, but I didn't get the Y adaptor in my kit.

Has something been put in place in the newer models or should I be looking at purchasing one of those cables so that I the change of a fire is reduced?

Daniel

It's science, so you don't have to guess. the PB+ draw max about 200W or so (I have a thread on it somewhere). we know its working off 12V so that's about 17A. so the wires and connectors from the PSU to the Prboard have to be able to carry say 20A to be safe. look up a wire gauge chart. I don't have one in front of me. but lets say for example sake that a 24AWG wire can handle 6A then you need a total of 20A / 6A wires in parallel on each of positive and negative. so that's 4 wires. so if you have 4 24AWG wires on each side then you are good. if only 2 on each side and there is another 12V rail on the PSU you can usually connect them together to get a total of 4 wires of amperage capability. if you don't have thick enough wire to handle the current, the wire and connectors heat. the heat ends up deteriorating the metal which then gets so resistive it heats up so much stuff catches on fire. so don't mess with this. make sure all cabling and connectors can safely easily handled the current you throw at it. unless you want to live dangerously.

by **REPRAP SQUAD** » 2013-Oct-Thu-20-Oct

Much agreed Plexus. I actually got a kit for wiring repraps and I think the wiring is like 24 gauge but it just looks like it isn't sufficient so I'm in the midst of upgrading all my wiring with a bit heavier gauge then I would ever need . Also adding custom wiring couplers that organize and connect eachother cleanly. Well worth the extra time and effort.

I think wiring is something that people commonly skimp on or don't even take into consideration when building bots.

by **Porter** » 2013-Oct-Thu-22-Oct

As a matter of course, I'll double check everything tonight. I'm just using what I got from Printrbot, which I don't beleive included the Y cable. If that is infact the case, should Printrbot be delivering these adaptors with the built printers?

Daniel

by **Porter** » 2013-Nov-Fri-20-Nov

OK, I checked it out this morning after reading up a bit on cables, etc. When I looked at the printer, the Y adapter is inside the Plus V2. I had only plugged in the square plug when I moved it last. I'll post pictures when I'm at my computer for future readers of this thread.

Daniel

by **REPRAP SQUAD** » 2013-Nov-Thu-11-Nov

I recently decided to upgrade all my wiring throughout my bot. This is definitely overkill as the smallest gauge wiring is now about 20 gauge used for homing and the largest is about 14 to 16 gauge for power. Here's a quick pic of the new wiring set up.

by **Mooselake** » 2013-Nov-Thu-17-Nov

Looks nice, well laid out! 14ga isn't really overkill, particularly if you have a Plus headed bed. I'd add a fuse or two, both for mains and DC power.

I'm seriously considering building a base for my Plus and putting the power supply, bed relay, etc. in it. Main power switch, LED panel lights, a voltmeter on the front, and a power entry module on the back. It'll be an excuse to get some wire ferrules and a crimper, too.

Kirk

by **REPRAP SQUAD** » 2013-Nov-Thu-17-Nov

I haven't added the heated bed stuff yet and that will be the highest gauge at 8-10. It's great for debugging experimental setups. Soon I will be using quadruple extruders as well. You have given me a few ideas. I'll be interested in what comes outta the moose cave. I just bought a few panel mount fuses as well as inline ones.

by **jordan314** » 2014-Apr-Sun-23-Apr

Arrgh the current getting started guide says NOTHING about plugging in two sets of power to the printrbot.
http://printrbot.com/wp-content/uploads ... -Guide.pdf
I only had the 6 pin connector plugged in, and wondered why the yellow wire got so hot. Now I know to plug in the 4 pin connector too. Luckily I had no fire or damage. I'm going to open a support ticket, this is a hazard for those of us that bought the bot pre-assembled.

by **EddB** » 2014-Apr-Mon-02-Apr

REPRAP SQUAD wrote:That's why I made a mount for this little guy. I also have text messaging sent when smoke or fire starts as well as IP camera. I also recently added a full graded fire extinguisher within 15 FT of my prototyping lab at all times. It's just a good idea when dealing with these temps and power requirements.

You could always connect a thermal switch right against the power connector so that the power shuts down (or some other action) if the temp reaches a certain point. You can salvage these out of rechargable drills, usually connected to the batteries to prevent overcharging or fires. Can also get 3 of them for a $1 from my favorite surplus store (http://www.goldmine-elec-products.com/p ... ber=G14366) they trip at 65c and re-enable the power once it drops. Could wire a relay to an alarm as a warning as well if you don't want the power cut. Might be able to stop this type of problem before it starts. Anyone know how hot it gets before that connector melts?
I grabbed a couple 20amp resettable circuit breakers (same source) for the bed and bot. They will help protect the board, but if the connector melts enough to short a fuse, it may be too late.

by **plexus** » 2014-Apr-Mon-08-Apr

If its an issue of wires/connectors heating, you don't know where the fire is going to take place. i suppose you could measure the temp of your 12V power harness when in full load and then set up some kind of temp sensor to do the switching if it exceeds that. but your point is well take - in order to be safe with unattended printing, there should be fail safes in place. its a lot of work to set that up because of the number of different fail safes that would be required.

the best thing to do is just be there while it prints so you can address any issues that come up at the time.

by **russianwizard** » 2014-Apr-Fri-20-Apr

I ended up getting a corsair CX430 (Super overkill but I've been through at least 2 cheapo power supplies in the past year so why not.) and I wired the 8 pin cpu connector up to the printer and the wires hardly even get warm. Since I have my power supply on a shelf on the bottom of a custom enclosure and the wires going through a conduit I beefed up the extension wires to 4x 16 awg wire. (2 for +, 2 for -) Here's a handy chart to show you what gauge wire you should use for your printer based on the amps it draws as well as the length:

200 watts (peak power draw of the printrbot) = ~17amps

currentdrawdc.PNG

Also when you are looking for a new power supply keep in mind most power supplies are optimally designed for 80% of their peak utilization. So if your printrbot draws 200 watts, you at least want a 250 watt power supply. Also I'd steer clear of any no name power supply brand. I personally reccomend atx power supplies made by Corsair, Antec, Seasonic and Silverstone. Also look for power supplies with the 80 plus bronze silver and gold ratings since these are made to be more efficient at 80% utilization like I mentioned above.

by **Seraphim** » 2014-Apr-Wed-02-Apr

So from reading this thread I have gathered that the weakest links in the chain are the molex connectors and wires on the PSU, which can be solved by using a Y-connector.

How about the 4 pin molex on the heat bed output, the mosfet and the trace wiring on the PCB on the printrboard? Can that withstand the 10-12 amps from an 8x8 MK2 heat bed?
I have seen people use a relay and run the high amp supply from the heatbed through the relay and directly to PSU, bypassing the printrboard. Is that necessary?

by **RetireeJay** » 2014-Apr-Wed-09-Apr

After much searching, I found a Molex datasheet http://www.molex.com/molex/products/datasheet.jsp?part=active/0022272041_PCB_HEADERS.xml that says the board headers are rated at 4A per contact. So with 2 contacts used for + and 2 used for Switched, the maximum current you should pull through the Molex header for the heated bed is a total of 8 amps. The MOSFET can handle more than that. The plated wiring on the board may or may not; some Chinese clone boards definitely did not have adequate copper thickness to handle 8 amps, but apparently genuine Printrboards purchased from Printrbot do have enough plating thickness.

by **Seraphim** » 2014-Apr-Wed-21-Apr

Thank you for the reply.
It seems that adding a Y-connector just moves the weakest link to the heatbed molex on the board.

What you have found is actually very concerning. What it implies is that if the heat bed draws enough current to necessitate a Y-connector, it means it is too much current for the heatbed molex as well. Is that right?
If that is right, it means you should probably not use a Y-connector (unless you want to be on the safe side when using a small bed) and ALWAYS use an external relay to power the bed.
For example, according to the figures Plexus supplied above, the heatbed he is using would draw more than the 8amp limit of the molex connector on the board and he thinks he is safe when just using parallel or oversized wiring.

This should really come from Printrbot. They supply a board that has a heatbed output on it so they need to specify the maximum current. I noticed someone has asked that question on their technical support page around a week ago but there has not been a response.

by **pvyParts** » 2014-Apr-Wed-22-Apr

Food for thought, but if the Heated bed is being driven with PWM instead of On/off ( Bang bang ? )

IMO ( not an enginner ) the apparent current will be less then if it was just constantly on.

Having said all of that I would probably still say that a external relay is probably the best way forward anyway cause i am not sure i would want that sort of current flowing through the board on top of what it already has to push out. last thing you want is a voltage dip or something play havok with the rest of the board, when you are a long way into a build and it messes it all up.

by **plexus** » 2014-May-Fri-12-May

when thinking about current, connectors and wire there is much to consider. its not just an issue of the current capacity of saif connectors and wires. the other consideration is the mechanical connection made in the connector. it seems that the board traces can handle the heat bed current as we've not heard of a case of the board burning up. we know that the wires can be a problem and need to be addressed with higher current capacity wiring. the main issue i think is with the actual connection in the stock molex connectors. that is not to say the connector cant handle the current but rather the mechanical connection between the male and female can not. this seems to be the case considering those molex connectors that burned, burned at the actual pin connection of male to female. the increased resistance due to the bad connection caused heat which caused corrosion which caused more resistance, your typical recursive failure.

the solution for me was to solder a high current connector to the existing connector, making sure to use both pins for + and - and a good solid solder connection. the added solder provides more current capacity. now the replacement connector, an XT60, provides the mechanical connection with better effect. that coupled with appropriately rated wires and both the connector and wires gets only moderately warm, i'd estimate no more than 30-30C.

i think the failure was not with the stock molex connectors being under spec for the current draw, but rather the mechanical connection made. not uncommon with cheaply made connectors. if you replace them make sure to use genuine whatever you decide, i suggest the XT60 as its more than enough and provides a good solid mechanical connection.

as long as you ensure your wires, connectors and power supply are all well within operating specs, you dont need to use the external relay solution. my PB+ with 150W qu-bd heater, 1/4" aluminum heat bed, XT60 connectors and 10 AWG wire has been humming along for a year now with no issues and nothing getting too warm, that isnt supposed to.

by **RetireeJay** » 2014-May-Fri-12-May

@Plexus,

Did you somehow solder your XT60 connector directly to the original Molex connector pins? (It seems to me that would be risky because applying force to connect or disconnect the XT60 halves would be putting mechanical strain on the solder joint.) Or did you use short pigtails of wire wrapped and soldered around the Molex pins and soldered on the other end to the XT60 connector - making a "floating" connection with the wire coming from the heated bed?

by **jr57k** » 2014-May-Sun-16-May

WOW! I had the SAME thing happen today. Looks like it happened in the exact same spot, between the PSU and the 4 pin adapter for the main board. I started as soon as I hit the bed AND the end, I was swapping out some ABS for PLA and the printr had been idle for about 4 months. I've had it two years and my gut tells me that there is corrosion between the connectors (I'll know once things cool off!)

WARNING! YOUR PRINTRBOT MAY CATCH FIRE DUE TO INSUFFICIENT CURRENT HEAD ROOM ON THE CABLE!

I'm going to replace with XT60s, just like I've already done with the heat bed connector (see earlier thread about FIRE).

by **jr57k** » 2014-May-Sun-17-May

Replaced connector with XT60.. The wires on the 4 Pin CPU cable from the PSU are still HOT. Almost too hot to touch. I may crack open the PSU and wire in some beefy wire for the connection, I'm not happy AT ALL.

by **plexus** » 2014-May-Sun-18-May

RetireeJay wrote:@Plexus,

Did you somehow solder your XT60 connector directly to the original Molex connector pins? (It seems to me that would be risky because applying force to connect or disconnect the XT60 halves would be putting mechanical strain on the solder joint.) Or did you use short pigtails of wire wrapped and soldered around the Molex pins and soldered on the other end to the XT60 connector - making a "floating" connection with the wire coming from the heated bed?

RJ I soldered the XT60 to the original Molex connector pins. It's not risky because there is no force on the cable at that point. You could also solder wires to them to move the XT60 out further just make sure the wire is sufficiently thick. I use 10AWG stranded for the heat bed.

by **RetireeJay** » 2014-May-Sat-19-May

I noticed that it takes a lot of force to mate & unmate XT60 connectors, so I wanted to make sure I could hold the two halve without putting any force on the board. Therefore, I used 14ga zip cord to move the connection point away from the board. This has the further advantage that if I want to I can unplug the heated bed without turning the Printrbot up on its side. (On the Plus, the board is underneath the center base platform. This cord moves the connector out to where it's accessible.)

Although I never had any visible heat damage at the bed connector, I now feel more confident that my setup is safe.

By the way, zip cord is made with a very high number of very fine strands - so that makes it an excellent choice for connecting to something that moves; it is both flexible and fatigue-resistant.

ZipCordSolderedToHeader.jpg

ZipCordAndXT60Conn.jpg

XT60Connector.jpg

by **plexus** » 2014-May-Tue-17-May

My strategy was to remove as much wiring and solder joints as possible. not that it might matter but when optimizing current paths this is a best practice. since I rarely have to unplug my heat bed and the cabling is strain relieved before heading to where its static, theres no issue with solder or wire mechanical stress. depending what you are going for.

by **REPRAP SQUAD** » 2014-May-Thu-11-May

I went with a similar setup. I made sure that the connections coming off of the heated bed are static as well. I also added a quick disconnect using a terminal block because I commonly take off my heated bed to use the full 17.5 inches that my y axis has to offer. It works great. I added an auto relay with diode so I didn't need to worry too much about the printrboard hb plug. Everything works great on it and I havent had any issues.

by **pomonabill221** » 2014-Jun-Thu-21-Jun

Since hear and reading of all the problems with the stock power supply connector, and even my connector getting hot, I have gone through extensive mods to my PB+ lc also.
Some time back, I bought some cable drag and rewired everything... X/Y/hot end/auto level probe/bed heater wiring and FET(I have the RAMPS board now so I relocated the FET to the print bed and used a small heatsink).

z axis drag cable.JPG

y axis cable guide.JPG

x axis cable drag.JPG

I also bought a 12v 20A SMPS and modified it for remote sensing to compensate for cable drop.
I also used #14 zip cord (malibu low voltage wiring), and use a 15 amp / 6 pin Cinch Jones connector for power.
Now, the voltage at the bed varies by only about 100mV and the voltage at the power connector stays rock solid. This is because I am sensing at the printer's connector and only the positive voltage (power return is not sensed as it would take major changes to the power supply), and still have a little drop between the connector and the bed heater, but very minimal.

bed connections.JPG

The bed wiring is also #14.
I am not using the RAMPS connector for the bed heater either, but run directly to the triac, and brought out the fet's gate drive.

pwr and usb.JPG

power and usb.JPG

by **REPRAP SQUAD** » 2014-Jun-Fri-12-Jun

Nice setup. I've done some similar mods to my plus. I had to run dual drag cables due to having dual extruders. I probably could have fit all of the cables in one big drag cable but I like using two small ones instead.

I decided to move my printrboard and extrudrboard to the outside for proper cooling. Here's a couple more recent pics of it.

You really did a great job and gave me a few ideas for future modifications.

by **pomonabill221** » 2014-Jun-Fri-16-Jun

Thanks!
It is fun to see all the different mods and ideas that "we" have come up with!

The drag chains are great but you don't want to get too many wires/cables or stiff wire in them or they will be a little hard to bend.
I think since you have dual extruders, the 2 drag chains was the best way to go for you as a single would have been larger and probably a little stiff.

I agree with moving the board outside, or at least install a small fan to move some air over the board. I have the RAMPS board and the motor drivers do get a little warm even with the heatsinks (stick on fins.... kinda mickey mouse I think).

I like your air intake ""nozzles" for your extruder fans! Looks like the intake for a jet engine.

Question... having dual extruders on your printer... I gather that the X motors have enough torque to move the mass without missing steps?

I'm getting the "twitch" to pursue (read buy), the laser cutter modification. There is a post here about someone looking into that, but hey, we have the printer with the appropriate drive electronics and the accuracy is very good, so why not? It's going to be a while but I have thought about doing something like this so just maybe.......

Oh the mods...........

by **REPRAP SQUAD** » 2014-Jun-Sat-11-Jun

Like you said, having two drag cables just makes since. My x axis has the large nema 17 kysan with lots of power. I've ran this setup really fast and never missed a step. I had a small amount of overshoot on the sharp corners but, that's it.

I myself have always wanted to add a laser attachment but I've heard discouraging things. Plexus is a moderator of these forums and he's a laser enthusiast. He also has lasers you could buy. He can tell you of the abilities of using a laser on a RepRap.

by **pomonabill221** » 2014-Jun-Sat-15-Jun

Good to hear that the steppers have enough torque to move your extruder "cluster". I kinda thought it would, but there is alot of mass there with two motors and all the extruder pieces. Starting and stopping I would think would be the hard part.

It would be nice if we could start a thread about laser mods to our printers and what people have done so far, and any recommendations that others would have, like Plexus.

by **REPRAP SQUAD** » 2014-Jun-Sun-11-Jun

Ya they have plenty of power to move efficiently. My x axis motor runs really cool with the smaller heat sink on it.

by **Michele31415** » 2014-Aug-Sat-16-Aug

Well it may not have actually exuded flames, but my Printrbot+ v.1 did get toasty today. I turned it on and set the bed to 108 to get it warmed up. Only it wasn't heating at all. I checked the +12 v. power supply and it was fine. Then I turned it over and found this:

ClipBoard.jpg

I pulled the connector off (had to use a pair of pliers because it was really stuck) and saw this:

ClipBoard1.jpg

That's as far as I've gone so far. It looks like the mosfet that drives the bed (Q2) just burned up. The question is why. One possibility is a short in the bed. I'm measuring a resistance of just 1.0 ohms across the heater element. That would mean a 12 A. draw. Can someone tell me what the resistance there should be?

The other possibility is that the 100 ohm resistor R11 gave up. Or maybe the mosfet just had enough. The hot end is working fine. Any suggestions of what to do next would be greatly appreciated. It was working fine until today.

by **plexus** » 2014-Sep-Wed-20-Sep

Wow. nice. not! looks like the connector failed. the FET might be still ok. I'd carefully desolder and remove the connector. check the FET for obvious signs of damage. Its an easy enough part to repair and so you could order a new part and replace it, regardless of if it works or not. Play it safe. then, on one of my boards, I ran 10 AWG wire soldered to the board, each polarity across both pads for the connector. this goes to an XT60 which mates to the PSU input.

by **amarisJones** » 2016-Mar-Tue-01-Mar

I think I may have missed something here. Because I am interested in this topic.
As far as I know.Replaced connector with XT60 ! The wires on the 4 Pin CPU cable from the PSU are still HOT.

by **bbrown64** » 2016-Mar-Tue-06-Mar

I am glad that you where able to save your printer and house from a potentially hazardous situation. I have one suggestion to make. Because long print times build up a lot of heat under the printer. Add a cooling fan blowing directly onto the circuit board. IMO I do not believe that these printers where ever designed for long build times. If so there would already be cooling fans where needed, as well as a better connections. Add that cooling fan should contribute to the overall safety of your printer.

by **RetireeJay** » 2016-Mar-Tue-06-Mar

amarisJones wrote:I think I may have missed something here. Because I am interested in this topic.
As far as I know.Replaced connector with XT60 ! The wires on the 4 Pin CPU cable from the PSU are still HOT.

If you have wires that are hot, then you need more copper. That is, more wires in parallel, or heavier-gauge wires. Because not only are you running a fire risk with hot wires, you also are losing power (and thus losing voltage) in those wires. Which means, among other things, that the voltage at the board can be spiking up and down as the load changes (as heaters turn off and on). So it's hard on your circuitry and can potentially cause glitches or unexpected behavior in your Printrboard.

If you have a Rev D board with a 4-pin connector for power, know that Printrbot used to supply a "Y" jumper that plugged into the board but allowed you to connect two different cables from the ATX power supply. Hopefully they can still provide that for you, or maybe you can find one with a Google search (there are other companies who still supply the Rev D board, and they should carry those jumpers).

by **Mooselake** » 2016-Mar-Tue-07-Mar

Necroposting through a VPN, doesn't look good...

Kirk

by **belgiantonio** » 2017-Nov-Wed-13-Nov

Got a similar problem today on my assembled metal plus. The power connector (the square one with six wires, three black, three yellow) just melted, the wires failed. I hope I don't have any problem on the electronic card.

This Printrbot was bought assembled, and if I can tolerate a need for tweaking, we have here a serious safety issue. Has the issue been adressed since (ie is PBHQ selecting a connector that does not create a risk of burning your house down ? Just saying if this problem is not solved, that makes the printer is not fit for sale in Europe). It should be quite easy to find a connector with the correct rating....

Anyway if someone has a suggestion for a suited connector available in EU,I'd be glad for it. Thank you!

by **RetireeJay** » 2017-Nov-Wed-20-Nov

The original boards rev A through D used a power connector with 4 pins. It was, worst case, being used close to the margin of its capability. But the 6-pin connector on the Rev F board should be very safe. If it's been assembled correctly and used correctly. If, for any reason, there are less that three solid connections from the power supply to the board's +12V side (yellow), or less than three solid connections to the ground side (black) then it stresses the connector's electrical capability.

Are all six of the connection cavities on the plug equally discolored from excess heat? Or was the heat focused on just one or two pins?

by **belgiantonio** » 2017-Nov-Sun-13-Nov

Hi Jay, and thank you for your answer. The three black wires seem to have suffered from heat, but the most affected was the one on the right side (when looking into the white part of the connector with black cables up). I did a new connection today, and my computer doesn't recognize the bot anymore. I hope the board is not fried. I'll try reinjecting firmware, then I'll check again.

I have a spare ramps card that I bought for another project, could it do the job if needed ? Will I loose any feature doing that ?

by **belgiantonio** » 2017-Nov-Mon-03-Nov

The LED on the board is lit when the board is plugged on USB only. The board is not recognised and I didn't succeed in making FLIP work properly (java error). I think a diode somewhere died in the fuss. I'll have to buy a new board I guess :-(

by **RetireeJay** » 2017-Nov-Mon-07-Nov

belgiantonio wrote:The LED on the board is lit when the board is plugged on USB only. The board is not recognised and I didn't succeed in making FLIP work properly (java error). I think a diode somewhere died in the fuss. I'll have to buy a new board I guess

If the LED on the board is not lit when you supposedly are providing power from the power supply, that means that no power is coming in through the connector (or very unlikely, a problem with the 12V to 5V converter on the board). It may be that the power supply connector on your board was never soldered properly to the board; that would explain the generation of heat and the eventual total failure to deliver power. So if you, or a friend, have a soldering iron and some experience making good solder joints, it might be worth looking at that option. Otherwise, you're right; your board as it stands is defective. Don't get a Rev G board; it's totally different and has different firmware and usage requirements. Stick with Ref F.

by **belgiantonio** » 2017-Nov-Mon-08-Nov

OH maybe I did not expressed myself properly (not a native english speaker, sorry). I mean, when I have no power connected to the board, but only the usb wire, the led light is still on (weakly, because it is 5v supply, not 12v). Is it normal or is it a sign my printrboard has dieded ?

I don't think there is a problem on this level, as the LED is normally lit when power is connected, but I'll have a closer look anyway.

Anyway, thank you for your time and answers (including all those on other topic which have been already helpful to me a couple of times)

by **RetireeJay** » 2017-Nov-Mon-11-Nov

Yes, the LED can light up with power coming in through the USB port. That does not indicate a defective board.
The LED should also light up when you have 12V connected, either with the USB cable connected or not connected.

by **belgiantonio** » 2017-Nov-Mon-13-Nov

I found the culprit to be the USB wire! It went damaged in the accident of the power wires, because it was working fine before that. I am relieved that my board is fine. One sad thing about PBHQ, and they can do nothing about that, is that they are in the US, and I am in Europe, meaning extra fee everytime I need a specific item

.

Thanks again for the help and keep up the good job!

by **RetireeJay** » 2017-Nov-Mon-13-Nov

I am glad to hear your board is fine.

The USB cable is "generic" so you should be able to buy it locally.

by **belgiantonio** » 2017-Nov-Mon-14-Nov

Yup. It is even "universal" probably meaning one can find them on Mars or the Death Star. 8-)

Printrbot Talk Forum

Printrbot on fire!

Who is online