MakerBot Industries

Hmm, I am using an Antec EA 500d green too, same model. Thanks for the info. I havent lost any stepper drivers or anything. But if this problem starts up again I will check the voltages. Was your problem intermittent?

Did you have these problems while the Antec was installed or did you switch from the stock supply to the antec because of the problem? How quickly did you loose your first stepper driver with the Antec? I believe Jetguy also uses this power supply, has he had any problems?

i have some of my stepper drivers plugged in to adapted sata to molex power connects (not the X and Y axis which is having the issue)? Would this cause any problems?

Hmm, maybe I should switch back too… I don't want my stepper drivers to break. I like the Antec though, it is so much quieter, my LEDs don't dim, it just seems to be so much more reliable. I don't know what to do.

Well, the Antec is staying in for now. The only thing I modified on this Antec PSU is putting a hole in the case (being careful not to hit anything inside of course) so I can secure my heater core ground to it. I haven't changed any of the leads, I just kind of wrapped them up. I do have some SATA power to 4pin molex connectors though, for the purpose of giving the EC its own lead. Not that shortening the non needed cables would cause problems. But you (tomlombardi) and MBI are right, best to get rid of the unnecessary variable.

As for makerbot support… so far Ethan says, "That is really weird and not good". Then he said to try the beta firmware and see if it still happens. I am not going to try the beta firmware until the problem starts up again though. Right now I am printing perfectly!

Go into repg, preferences, then change the firmware update url to this:

http://firmware.makerbot.com/beta/firmware.xml

And it will download the new beta firmware when you go to update.

Or so I think, I haven't tried it yet.

I think I may have had a breakthrough with my problem thanks to help from the google MakerBot group. Someone mentioned something that made things click. I am still going to buy the pololu drivers just in case the problem comes back.

Zgbot said this:

"anything that can cause electrical noise? florescent lights plugged
into the same power strip?"

At that question something clicked. The first time I had this problem, last thursday night at around 8 PM… I had just moved my printer to a brand new printer cart I had picked up. Well it was on a different power outlet. I had my ToM plugged directly into the wall with no surge protection or UPS or anything. I noticed every time I would turn my printer on, the UPS on a different outlet across the room would beep for a second. So the problem started happening as soon as I moved to this outlet, without surge protection or ups, on a circuit in my house that obviously has some issues or is overloaded or something.

Funny thing is, it happened around 8 pm the first night it happened and it happened again last night at around 8 pm. Which is peak usage time for electricity in my house hold and just about any other household in the world.

So, I don't know if this is what is causing my issue, I really hope so. I have since moved to a completely different room, which is on a different circuit and now I am hooked up to a UPS.

I still am going to buy those Pololu stepper drivers though. Are they hard to hook up? Do I have to buy a specific one for the ToM? Are there any guides out there for hooking them up?

So far so good here too, but that doesn't say much, it was working on the old outlet too before I moved it today. It also worked for a week between the first time it happened and the second time it happened.

I really hoped the problem is fixed for the both of us. I don't know if this switching to a different outlet is going to help, but it can't hurt. Should have had it on a UPS anyway. When I moved to the printer cart I didn't bring a ups along with it. I would have never guessed that my outlet could be the issue. I probably would have kept going on, turning my printer on, UPS's beeping all over the house and me sitting here oblivious thinking "what the hell is wrong with my stupid printer".

I shouldn't jinx myself thinking the problem is fixed though, it could be looming around the corner.

Makergear is sold out of kits of course. They are always sold out of stuff…. I guess I am going to have to buy one some where else. Thanks for the info though. Noone sells a prebuilt kit? With all the problems I am having with the last "kit" I put together, I am a little bit nervous about building stuff….

Yup, I have four spares (two on old/unused my Mk5 hot end and two replacement ones I got from MBI to repair the last failure). I guess I'll wait for it to start failing again and then replace the resistors.

Why would it only happen during a printing operation? When I move in control panel, it usually moves without issue. I think it may have happened once or twice during control panel movement, but it would move perfectly for the most part (99% of the time). And when it occurs during homing, it usually fails towards the end of the range, near the endstops.

Why would it only happen during a printing operation?

That's the nature of transient problems. Vibration from the motors to the Z stage? Vibration from stepper extruder to Core? Time-at-temperature plus vibration? All of the above plus length of a single move?

For all I know, those resistors have nothing to do with it.

However, we've seen many instances of short-to-Core failures, it would explain what's happening, and you may just be the first person cough fortunate enough to notice an impending failure.

This could turn into Science if you're not careful…

@Ed I wasn't doubting your explanation, I was just curious. It is just so strange to me, that like, for example, It will start to home, get pretty far, then once it gets to the endstop corners, it starts doing this. But once it reaches the endstop, everything moves fine to center start position. I can also stop it in the middle of failing, immediately go into control panel and everything will move fine.

Your explanation is the best one yet, because it would seem that those are times, for the most part, when the resistors are activated. They turn on and start heating right before the machine homes. They are usually on when it is printing. They aren't on when I am messing with control panel. It is just so weird that it mostly happens when X and Y are headed towards the end stops and not away. It seems to do this with the prints too, when it messes up, it always seems to be when the X and Y axis is moving in that direction. Do the steppers draw more power when moving in one direction over another?

Strange problem.

Do the steppers draw more power when moving in one direction over another?

Nope, the driver chip sets the current level for each winding and that's what it delivers.

What does differ: mechanical resonances. Toothed belts: turning one way puts tension on the short section directly to the stage, turning the other way puts tension on the long section that loops around the far pulley. Steppers: a particular speed excites a resonance only when the stage is at a particular location.

Plywood & acrylic sheets, steel rods, brass bushings: there's not a lot of damping in a Thing-O-Matic frame, so anything that vibrates here affects stuff over there.

And that's not getting into the sources of electrical noise that can give an Arduino a headache.

Admittedly, the breeze from all that handwaving might ruffle your hair, but the general idea is that everything is deeply intertwingled, seemingly unrelated events really have close coupling, and looking at anything in isolation won't reproduce the problem.

Debugging something like that is really hard. Designing a system that avoids those problems is even harder… but you must do it only once!

I have been having massive problems with skipping on my T-O-M also. Things that printed before fine now have skips. I also upgraded to the Antec supply, but I had the skipping problem before on the stock supply too. This was my primary reason for developing the servo drive.

I'm going to try stealing my Pololus from my Cupcake and use them in the T-O-M to show it's something retarded with the stepper drivers. I've got enough gen 4 stepper drivers around to make 3 bots but all of them have shown some form of a skipped step issue. They will test fine and even work for a while-but all develop issues at some point. This is a really sore subject with me because asked Makerbot when they released the T-O-M why they chose those chips over others in the market.

At first I suspected solder joints, but others have said the chips themselves fail internally. I'd rma them the 12 stepper drivers but like I said, the problem is they work almost everytime at startup-it's a longer time based operation when they flake out. It would make me really upset to send them all and get them right back. With Pololus at $13 a piece, it's hard to justify even messing with shipping to RMA on the stock ones-only to get one that may exihibit the same issues. 12 drivers at the $39 Makerbot charges is nothing to sneeze at ($468) vs 12 Pololus at $156. All you need extra is a single strip board and some connectors.
Also, one key advantage is the single pot for current adjustment rather than a bunch of them for all the microstep adjustment functions. The only thing I am concerned with it the extruder motor because the Pololus have a 2 amp max current. That may have to remain on a stock driver for now.

I have a decent sample size and am not impressed so far.

As I stated in my post, I have 2 of the Antec supplies. I would honestly throw the power supply away if I thought it was the issue-but being I have the LED strips and could see a power glitch, I'm pretty sure I have the steadiest power the bot as seen to date.

The real problem is I have 12 of those stepper drivers and don't trust any of them. I'm willing to bet TomLombardi is in the same boat as I have suggested-replacing the drivers with the same ones solves nothing nor does changing the supply. It will likely work for a while then go back to the same problem.

The only other theory (if you believe the supply is the problem) is that the Antec has worse 5 volt regulation than the stock supply due to the tiny load presented by the motherboard. In theory, the power resistor on the board is the only real power drawn at the 5 volt line. I have no reason to believe this is true, and even less reason to think the stock power supply had better regulation. I will put a scope on this thing over the weekend as well as swap at least X,Y, and Z to Pololu and see what happens. I have both supplys (Antec and stock) and really 2 bots so I should be able to do side by side comparisons.

Ok, I trust your opinion Jetguy. I am gonna leave the Antec Psu in. I am also considering buying some pololu steper drivers in case of failure. I have no idea what
I need to buy though. Any chance you can point me to a place to buy or some kind of installation guide?

Also, the only way I even really could tell was the clear case and the fact the infrared probe lets me see the temp through my case. My first good print on a job longer than 3 hours.

After seeing the temp rise and having tried the print now 3 times before (something like 9-10 hours of printing wasted). The only change was adding the fan.

Trust me, I was about to ship all my electronics to Makerbot over the same thing.

My X and Y motors are running at 40C now.

X fan mod- note that's a 30mm not 40 for this exact mod.

No, you give some of the best advice here. I trust you. Wouldn't have done the Z axis brass tube mod or the PSU swap if I didn't value your opinion. You really know your stuff. I wish I knew half the stuff you do about this stuff.

Just, I have done a lot of troubleshooting (computers) in my day. And of course, you understand that sometimes, you just gotta try every thing. Specifically things you did the day problems started.

I have a question for you though… When the problems came back tonight, my Y Axis would get all crazy. Like I'd go into control panel, try to move it, grind grind grind fail. But then all of a sudden, no matter which button I would push Y would go in the min direction. It is very much like the some signals are getting crossed or some kind EMI or the motherboard is just spouting out gibberish every so often instead of commands to the steppers.

The one other thing I could do is unground the Extruder. I had run a wire from the bolt on the heater core down to the power supply case. But I doubt that is the problem, because it did it even while the extruder was completely off.

I was thinking about getting an STP CAT 5 cable for EMI protection but I don't think that will help either. Because that only interfaces between the extruder controller and the Motherboard. And the extruder works fine.

I also only have issues with X and Y. Z never seemed to fail, nor did the stepper extruder.

I just swapped the stock supply in. It grinded on Y first try. But I made sure to power off, then power back on with the power chord unplugged and the usb cable removed. Then I shook the Y platform to make sure it didn't get knocked out of place during the move. Now everything is working.

That is worrisome, that it did grind with the stock power supply. Now I play the waiting game. If it comes back… we know the issue is some where else. I really hope it doesn't come back!

This 40mm fan made the difference from not being able to make a print without skipped steps on any job over 3 hours to working perfectly. It's just dangling in the bot by the power cord-blowing in the general direction of the x motor passing air underneath and around the end of the stage. It dropped the temps from 80C to 40C!

I went from this

To this by adding a fan dangling in the bot-using the Antec supply.

Imma gonna reply down here now.

Yeah. that tall ToM is sooo cool.

Hey I got fans for this type of situation when I ordered my Led strips. Maybe I'll try dangling a fan in the electronics bay. I wish I had gotten little heatsinks though. I guess I can chop some up that I have laying around.

With that acrylic body… do you have any problems dissipating heat? Like any more so than wood?

So far so good printing new servo brackets for my delta bot. No skips yet.

Oh and I had / have my power for my EC coming of a SATA connector on its own lead. On both the Antec and now this stock one. Wouldn't that be just as good as using the PCI+E one?

About heat- I don't think it's any different than stock thermal wise.

Oops -should have made it clear I was answering your power question

Nope, and here's why. The stock power supply only has one rail. No matter what wires you connect it to, the entire bot is affected by both the voltage droop and or any noise from the EC or the resistors. Also, the chief way to reduce the voltage drop is to use more parallel wires to decrease the resistance between the load and the source. the 6 pin connector has 3 +12V wires and 3 grounds. The SATA has 2 grounds but only one 12 volt wire just like the standard Molex.

At least on the Antec, you get both the advantage of 6 wires and it's a totally separate 12 volt rail inside the supply.

The difference is that my 1 meter white led strip dims when the heaters turn on on the stock supply or when I had the EC on the Molex or even the 4 Pin CPU plug (all the same rail), but now, my lights are always steady using the PCIe rail for the EC.
Next mod posted is how to kick the fan on full speed in the Antec.

Yeah, I know, I have that problem too with the LEDs dimming. Not that it was a real problem. Rather have that then what I have been having. That is why I switched Power Supplies to begin with. And you can tell that Antec Supply is much better quality than the stock one just by the weight difference alone. I think the Antec one weighs 3 times as much!

I honestly don't think my problem is fixed with moving back to the stock supply… but I have to try. We'll see what happens.

I was just thinking that the SATA power connectors might run off their own rail too, because they were on seperate leads, I guess that is more for organization purposes, or ease of assembly at the factory. I couldn't say for certain because I never cracked it open.

I had a volt meter on the power supply this time during failure…. it was acting the same way it did during successful operation. Never dipped below 11.98 volts. Didn't spike or anything I was expecting to happen.

I guess now that I have it out I can take a peak inside the Antec. My bot just finished a successful print with the stock supply.

What resolve the issue for you?