Rambo Board Electrical issue

[Sheaffer]

First, some Context. I have been printing heavily on my Rostock Max V2 since I finished assembling about 4 weeks ago. When I say heavily, I mean about 20 hours a day on average. Since the original build the bed has taken what seems to be a long time to heat (10-20 minutes), but I really don't know how long it is supposed to take and the printer is close to a window in a moderately cool room (ambient temp maybe 60F). I also had a prior problem with the Rambo board that was entirely my fault -- I arced a circuit on the hot end and popped one of the little ceramic fuses. This was about a week ago and replacing the fuse easily solved the problem.

Now the Problem. All of sudden the bed stopped heating. I figured it was just the 15A fuse...I wish. The positive connection on the Rambo board to the bed had completely melted.

See pic here http://imgur.com/hnlf4ZM.

And here http://imgur.com/bc1DICH

The positive and negative wire got so hot their insulators melted together, but didn't appear to short. and the positive wire actually melted to the internal metal in the connector. I'm damn lucky this didn't start a fire.

So...I have a few questions. First and foremost, what could have caused this? Even after the short described above, the printer was performing perfectly other than what I perceived as the long heating time for the bed.

Assuming I can diagnose and fix the underlying cause, do you all have thoughts on replacing the connector vs. replacing the whole Rambo Board?

Thanks in advance for any advice the group can provide.

[Eric]

It's happened to others, here's a couple previous threads about it (first couple hits searching for "melted connector" with forum search function):
http://forum.seemecnc.com/viewtopic.php ... 09&p=74702
http://forum.seemecnc.com/viewtopic.php ... 10&p=53336

Your options range from sending the board in for repair (you'd have to ask SeeMe if it's a warranty repair or not), removing/replacing the connector yourself, or even soldering wires directly to the traces on the board and using an in-line connector on the wire so it's still removable.

[mhackney]

If the RAMBo can be repaired under warranty, go for it. If not, contact Ultimachine directly. In the past they have fixed RAMBos for me that I blew in multiple & creative ways. Once you have the board back in hand, a better end result is to use an SSR to power the bed and not connect high current to RAMBo directly.

[Sheaffer]

Thanks...I opened a ticket with SeeMeCNC...it seems to me that if there have been at least 3 nearly identical failures (mine plus the two others noted above), there is a design defect. Any recommendations for the SSR? This solution seems to make sense to me.

[mhackney]

The problem is that the quality of the soldering, wire gauge and other factors have an effect (negative). The connectors on these boards are not really designed for high current applications. An SSR is much safer. As far as recommendations for SSRs, it's a great question. Not all SSRs are created equal. Here's something I wrote on the google group recently:

First, David Crocker wrote this:

The cheap DC-DC SSRS all have high voltage drops as described above, so they need a lot of cooling. It is possible to get DC-DC SSRS with a much lower voltage drop, but they are expensive, for example Crydom DC100D40 costing about USD 100. A less expensive alternative is a 5-terminal SSR such as the one sold by reprap.me that they call a power expansder.

The issue is that most of the cheap SSRs have a large voltage drop. This results in significant heating. I was aware of this from another application I had a few years ago. I discovered these Auber SSRs which are both inexpensive and do not have a large voltage drop:

I ran my Onyx on 24v directly from the RAMBo for a year with no problems. When I switched to a different controller I decided to be "safe" and use an SSR. I used a cheap Fotek 25A relay which literally caught fire within a few days of putting in service (it was on a heatsink). I replaced it with this: http://www.auberins.com/index.php?main_ ... cts_id=288

Auber 100A #MGR-1DD80D100. Here's the product description:
This SSR is specifically designed for low voltage DC applications such as thermoelectric cooler (TEC or Peltier cooler), LED illumination and low voltage DC heater. A unique feature of the design is its extremely low on-resistance (only 0.006 ohm@ 70A, 10V). The main benefits of using this design are
1) Low heat generation from the SSR. Removing the heat generated by the SSR itself can be a headache when high current is passed.
2) Low voltage drop. There will be negligible voltage drop at the SSR. In comparison, other DC SSRs may have 1 - 3 volt drop with a 12V power supply. The voltage drop can become a significant issue when low voltage DC power source is used.
3) Low cost. The price of this SSR is much lower than other same current rating DC SSR with higher voltage rating. In addition, you will also save the cost and space for a heat sink.
The maximum working voltage is 80VDC.
Since the heat generated by this SSR is about 10 to 100 times less than other DC SSRs with higher voltage ratings, there is no need to use a heat sink when load is a 12VDC 10A TEC or heater. When used with our HS25 heat sink, it can handle a 60A load. When used with our HS25ET (rated for 25A AC SSR) external heat sink, it can take a 80A load. And when used with our HS40ET (rated for 40A AC SSR) external heat sink, it can handle a 100A load.

I've used this every day for a year now (6+ hrs of printing a day) with a heat sink and no problem. I have a thermistor attached to the base so I keep track of the SSR's temp. It has never been more than a few of degrees above room temp. I've NEVER printed unattended, if I had I would have had a major fire most likely with that Fotek failure. I use this relay on all 5 of my printer's beds now.

[IMBoring25]

It's not unique to SMC or the Rambo. Every kit I've ever seen is on the ragged edge of the limit of the heat bed terminal capacity. My RepRapPro Mendel melted down the heat bed ground terminal on its original Melzi twice before I went to an SSR. All it takes is a bed with a resistance a little below spec or an imperfect attachment at the screw terminal.

I think my Rambo is still intact but the Max is on sabbatical until I can get the SSR installed since my old-style ATX PSU doesn't have the chutzpah to properly drive both the bed and the hot end at the same time.

[geneb]

One thing (of many) that mhackney's got going for him are those crimp-on ferrules that he uses. They provide great contact area for the compression terminals used on the RAMBo and other boards. I'm going to write up their use later, but they basically take a squishy bare wire and turn it into a hard, square post that can be more securely held in a compression terminal connection.

It's one of those things that's such a good idea I wish I could insist builders do it when they first assemble their kit.

g.

[mhackney]

I had a good idea?

[Eaglezsoar]

I searched for a bare squishy wire and could not find one. Is this a GeneB only wire or is it available on the web somewhere?
Inquiring minds want to know!

[Sheaffer]

#1 SeeMeCNC is awesome...they got back to my ticket and authorized and RMA even though I didn't ask for it.

#2 You guys/gals are awesome...I appreciate all the advice.

#3 The SSR idea is great...going to order the one recommended above while the RMA is underway....thanks again!

[Sheaffer]

I received the new RAMBO board from JJ. Got the relay recommended above. Got a heatsink on Amazon for 5 bucks. Rewired everything over the weekend and am back up and running.

Thanks again everyone...I truly appreciate all the help and advice. I think we can mark this thread as "solved".