kbob and Impulse, the Rostock MAX

[kbob]

I just wanted to ensure a good round hole in the RTV applicator. It's not like i used a torque wrench on the drill chuck or anything.

Not overkill at all.

In other news, the bed thermistor, the hot end thermistor, and the hot end resistors are encased in RTV. They should be cured now, but they're at the maker space and I'm not.

Better Way to Bond Resistors?

I estimated 24 turns of aluminum foil around those resistors. I wound it as tightly as I could, but 24 successive air gaps, even tiny air gaps, make a pretty good insulator. I wish I knew a better way to put the resistors in there. I thought about the thermal paste that you use with heatsinks, and I thought about some kind of oil bath, but I wasn't convinced either one would be compatible with the RTV.

But I'll be interested if someone finds a better way to bond the resistors to the hot end.

[kbob]

These worked very well at the Cheapskate end of the delta arms to eliminate play. I got 36 (that's enough for 12 printers) for $2.49 at Target.

Still sorry about the white balance.

[jetpad]

I don't know if it is a better way but it was a different way. I've had no problems with the hot end since I did it. I just used Silicon tape for everything on my hot end. Here are a couple pictures of it.

http://forum.seemecnc.com/viewtopic.php ... 5&start=10

[Eaglezsoar]

Now I am confused. You apparently are running two build logs at the same time, this one and the David Smith's (Jetpad) Build?

[Eaglezsoar]

I just wanted to ensure a good round hole in the RTV applicator. It's not like i used a torque wrench on the drill chuck or anything.

IMG_9706.JPG

In other news, the bed thermistor, the hot end thermistor, and the hot end resistors are encased in RTV. They should be cured now, but they're at the maker space and I'm not.

Better Way to Bond Resistors?

I estimated 24 turns of aluminum foil around those resistors. I wound it as tightly as I could, but 24 successive air gaps, even tiny air gaps, make a pretty good insulator. I wish I knew a better way to put the resistors in there. I thought about the thermal paste that you use with heatsinks, and I thought about some kind of oil bath, but I wasn't convinced either one would be compatible with the RTV.

But I'll be interested if someone finds a better way to bond the resistors to the hot end.

The best way is to invest in one of the cartridge type heaters. They are 40 watt and 6mm in diameter. Much better than using resistors.

[kbob]

The best way is to invest in one of the cartridge type heaters. They are 40 watt and 6mm in diameter. Much better than using resistors.

Oooh, goody! Another upgrade to look forward to!

But wouldn't cartridges have the same problem? How do you bond them to the hot end cavity for good heat transfer? My cavities are full of RTV, but I measured them at 0.25" (6.35mm) yesterday.

[Eaglezsoar]

Some hot ends are being made with a set screw that holds the cartridge in place, for those that do not you would
need to wrap it with aluminum foil until it is snug. You should not have to use RTV and it virtually eliminates shorts.

[kbob]

I finished assembling the printer early this morning. At lunch time, I set the printer on my desk, plugged it in, and promised myself I wouldn't turn it on until I finish work tonight. About 2:30 PM, I left my office for a couple of minutes and came back to thick smoke pouring out of the power supply.

No photos, sorry. By the time I got the power supply out of the chassis, it was just a little warm. The only visible damage outside the power supply case is that it is leaking brown fluid. Would that be a burst capacitor?

There's a 1000 watt Corsair power supply somewhere around here that's just over a year old. I think I'll dig that one out of whichever machine it's in and cut its connectors off. I still want to get this machine running for this Saturday's regional Maker Faire, but it just got a little less likely.

[kbob]

I pulled a Seasonic SSR-550RM out of a PC here. After a few hours remaking the wiring harness, the new power supply is in. I've powered the printer on, I've tested the limit switches and have just started testing the motors. I had to reverse the leads on all three motors -- I suspect an error in the manual. (-:

Not only is the new power supply a little bigger, it is also over two pounds heavier.

Which box still holds magic smoke?

it was tight getting the bigger brick into the Rostock base, but it did fit.

It has a single +12V rail rated for 45A (vs. 16A for the Logisys). I am no longer worried about overloading the power supply. (-:

[kbob]

The Rostock MAX also makes a fine soldering vise.

A Panavise is easier to set up.

[ApacheXMD]

Glad everything is ok. Electrical fires are scary.

Was there a power switch on the old power supply and was that on at the time?

Any insight on what may have happened?

[kbob]

Glad everything is ok. Electrical fires are scary.

Was there a power switch on the old power supply and was that on at the time?

Any insight on what may have happened?

The Logisys power supply did have a hard power switch. It was on. The soft power switch on the Rostock front panel was off, so the power supply was in standby mode. It had been plugged in for two hours when it started smoking. It had not been turned on at all. I was not doing anything with it, just using a laptop near it.

I already mentioned the brown fluid that leaked out afterwards.

I think one of two things happened. Either I'd managed to get some metal into the power supply (e.g., a wire snipping) or a component gradually failed into a short. I haven't looked inside the supply.

Meanwhile, at this very moment I am watching my Rostock print a thing. Wheee!

[geneb]

Sounds to me like you had a catastrophic capacitor failure. I bet if you open it up, you'll find a split one. Have you contacted SeeMeCNC about it?

g.

[kbob]

Sounds to me like you had a catastrophic capacitor failure. I bet if you open it up, you'll find a split one. Have you contacted SeeMeCNC about it?

g.

Yeah, I exchanged some email with John. By the time he got back to me, I had already gotten the new power supply installed and all the electricals debugged except the fan. (I owe you more errata about the fan.) I told him there was no point in sending me a new PS, so he didn't. (-:

[kbob]

Last night I was printing long after dark. Since there isn't a task light near my printer, I installed an LED lighting system. Sort of.

Photographed in the daytime?

The light.

Those colorful clothespins are from Horrible Fright. http://www.harborfreight.com/22-piece-s ... 69374.html I use them to hold the Cheapskates in place when I transport the printer in the car.

Note the Onyx nail polish remover bottle. Only use genuine Onyx brand acetone; it was formulated specifically for your Onyx bed!

Photographed in the daytime?

[ApacheXMD]

Off topic: Nice flashlight! I was sooo tempted to get one of those hexbrights, but I have too many flashlights as is.

[kbob]

Off topic: Nice flashlight! I was sooo tempted to get one of those hexbrights, but I have too many flashlights as is.

Thanks. I find myself using it a lot more than I thought I would.

[kbob]

Now the printer is operational, it's time to dial it in.

I've read Gene's manual and followed the procedures.

I've read polygonhell's tips on tuning the amount of extrusion. I don't have a caliper with sufficient resolution yet (UPS is delivering it tomorrow). My digital caliper says the single layer wall thickness is 0.6 mm. Or 0.8mm. In lieu of the right tool, I tried printing the solid blocks with different infill percentages. There are visible gaps all the way up to 90% infill, so I set Slic3r's Extrusion Multiplier to 0.95 and called it good. I will revisit that when I get a new caliper.

20x20mm blocks, 50-100% infill

Same blocks, side view

Am I right? Am I way off? These were all printed with 0.2mm layer height and 0.55mm default extrusion width.

Is there another calibration document I should know about?

So that's what I've got. But here are the problems I know of.

• I have brown spots (burns, I think) on the underside of many prints. Here are some of those same blocks agin.
  

  

  Same blocks, brown spots on bottom

• I get strings. Here is the Lava Vase at 1/4 scale.
  

  

  Lava Vase, 1/4 scale, very stringy

  I've set retract length to 3mm. Is that too much? Not enough? Are there other retract parameters to set?
• I printed the 50mm tower with a single layer perimeter and no infill. It came out crooked with a bead down its back. Should I be concerned? What is this object supposed to test?
  

  

  50mm tower. Crooked.

  

  Same tower, close up.

• I have no idea whether my hot end temperature or bed temperature is high, low, or just right. I have not yet bought a thermocouple-based thermometer to calibrate the thermistor. If I did, what temperature would I aim for? (ABS)
• I am using a lot of ABS juice. Anything else and the prints curl right up off the glass. Elmer's glue stick did not work well. I've tried bed temperatures up to 95℃ (uncalibrated); the Onyx can't get to 100℃. Would a different bed temperature help?

So I guess my top level question is, how do I dial this printer in now that it's sort of working? Where do I go from here?

[geneb]

Printing ABS with an un-verified thermistor is risky. If it's under-reading, you run the risk of ruining the PEEK section of the hot end - Harbor Fright sells a nice little meter that has a good thermocouple on it for under $40. (I'm lucky, I've got three HFs within easy driving distance! )

I've found that I'm getting MUCH better prints from PLA than I ever got from ABS. I think it's a combination of different properties in the material and maybe PLA is a bit more forgiving of my uncoordinated stumbling around.

Blobbing and stringing boils down to slicer adjustments - I'm not good enough at that yet to recommend anything to anybody.

When you print a cube of known dimensions, how accurate is the result you're getting? BTW, a 20%-25% infill is plenty for most things you'll be printing.

Congrats on finishing your build and thank you very much for the feedback you've given on the 2nd edition!

g.

[Polygonhell]

The 100% fill clube looks pretty good at 100%
The tower is a notoriously difficult print, the fact it even looks like a tower is a win. The issue is heat, if you really want to print a better one, pint 4 at a time, it gives them time to cool between layers, active cooling can also help.
You will get seams on parts, the start and endpoints have to overlap, the size of the seam is dependent on the slicer, and the settings in it.
The strings can be cleaned up with the retraction settings, 3mm is probably too little for the stock SeeMeCNC hotend, try 5mm of retract at 30mm/s and 5mm of Wipe in KisSlicer.

[kbob]

Printing ABS with an un-verified thermistor is risky. If it's under-reading, you run the risk of ruining the PEEK section of the hot end - Harbor Fright sells a nice little meter that has a good thermocouple on it for under $40. (I'm lucky, I've got three HFs within easy driving distance! )

I've found that I'm getting MUCH better prints from PLA than I ever got from ABS. I think it's a combination of different properties in the material and maybe PLA is a bit more forgiving of my uncoordinated stumbling around.

Blobbing and stringing boils down to slicer adjustments - I'm not good enough at that yet to recommend anything to anybody.

When you print a cube of known dimensions, how accurate is the result you're getting? BTW, a 20%-25% infill is plenty for most things you'll be printing.

Congrats on finishing your build and thank you very much for the feedback you've given on the 2nd edition!

g.

Old Chinese proverb: Man with Harbor Freight in town always know where to buy tools. Man with three Harbor Freight never sure.

I didn't make it clear -- I don't have a thermocouple thermistor YET. I haven't had a chance this week to visit the solitary, lonely HF that's nearby.

[kbob]

The 100% fill clube looks pretty good at 100%
The tower is a notoriously difficult print, the fact it even looks like a tower is a win. The issue is heat, if you really want to print a better one, pint 4 at a time, it gives them time to cool between layers, active cooling can also help.
You will get seams on parts, the start and endpoints have to overlap, the size of the seam is dependent on the slicer, and the settings in it.
The strings can be cleaned up with the retraction settings, 3mm is probably too little for the stock SeeMeCNC hotend, try 5mm of retract at 30mm/s and 5mm of Wipe in KisSlicer.

Thanks, Polygon. I will try 5mm at 30mm/s.

My questions were all over the map in that last post. Here's something a little more focused.

What are the interesting retract settings? Distance and speed, is that it?

Is there some set of tests that will tell me if my retract is too little, too much, too slow, or too fast? Ideally, something that iterates toward some good settings?

[Polygonhell]

If you use Slic3r or Cura the only settings you have to control retract are distance and speed, in KisSlicer you also get wipe, plus some control when retract happens.
The way I usually tune retract is to create a cylinder model 20-30mm in diameter, I put 2 of them on the build plate and print them as single wall objects.
You are looking for 3 things the strings between them, whether the perimeter is closed, and if extrusion is consistent.
I would suggest starting with 5mm retract/prime at 30mm/s with 5mm of wipe if using KISSlicer, change one thing at a time, as you increase retract you'll find that the strings lessen, too much and you start having gaps appear at the start of the layers, too slow and you get blobs because the head sits in one location retracting for too long.

[kbob]

Thanks. I will run some cylinder tests this weekend.

But first, the printer is going to Maker Faire tomorrow!

[kbob]

A lot has happened since my last post.

First, my Rostock MAX was on the local TV news. It was a story on 3D printers in the area. You can see the Rostock (and my left ear) on screen for about 1.5 seconds; the MakerBot was the main attraction. (Rick gives much better interviews than I do.) http://www.kval.com/news/tech/211477411.html

Second, my Rostock MAX went to Eugene Maker Faire. It was completely mobbed all day. Fortunately, it was printing well (if too slowly), considering that it was less than four days old at that point (and was my first 3D printer). I made a single-wall Teethy Tiki ( http://www.thingiverse.com/thing:49334 ). That print ran for three hours of the six hour show. Single-wall prints are easy.

Since then, I've continued to print. My thermocouple thermometer came from Amazon on Wednesday. I finally got around to calibrating the thermistor tonight. My thermistor reads 4℃ low at 150℃ and 11℃ low at 260.

Now I have two problems.

• I need to work through the math to figure out how much to adjust GENERIC_THERM1_BETA to give the right amount of compensation. That's no big deal.
• I can't reflash my RAMBo. Apparently my RAMBo has the missing bootloader described here. http://forum.seemecnc.com/viewtopic.php ... =10#p13887

For now, I'll just manually adjust temperatures. I will need to reflash that board eventually, though.