Replicator 2 active cooling fan not working properly

[niike....@gmail.com]

  Hello, everyone,

Recently I've encountered a problem with my Makerbot Replicator 2. Active cooling fan was not running during the print. I've forced it to begin moving with my finger and it began rotating, but the rotation was very very slow. After the print was finished it hadn't stopped rotating. The situation is the same even if I am not printing anything; If the printer is on I just move the fan a little with my fingers and it begins rotating. I've had sailfish firmware, now I've restored the original one but the problem is still there. I've also replaced the fan with a new one from a local store, which has 12V but it has the same issues as the original fan had. Usually I am using the Makerware for printing, but looks like its a hardware issue to me.

I've never suffered any problem like that. What can it be? Can anyone advise?

Thank you.

[Dan Newman]

Standard problem with Rep 2's and 2X's: the MOSFET controlling the fan has failed
and needs to be replaced. Easy to do if you are comfortable soldering. Search
this forum for posts on failed Rep 2 MOSFETS (or failed MOSFETs on mightyboard rev G).

Dan

[Scott Booker]

12V replacement fan for a Rep2???

Nope, should be a 24VDC unit.

[Jake Clark]

Here's the correct blower fan: http://www.fargo3dprinting.com/products/replacement-blower-fan-makerbot-replicator-2/

Should have a quick-connect version of this up within the next few weeks: 

-J

[Roy Kusuma]

Hello! I'm having a similar issue and it looks like replacing the mosfet is the way to go. Just wondering, for a replicator 2 is there a way to switch the main extruder to the perfectly good b port meant for a 2nd extruder? I imagine it would just be an adjustment in the firmware programming no?

Thanks!

[Dan Newman]

There's only one "port" for the print cooling fan. So please be more specific: is it the extruder
heatsink cooling fan which is impacted or is it the print cooling fan which is impacted?

Either way, sure you can re-jigger the pin assignments in Configuration.h and then rebuild.
But if it is a heatsink cooling fan as you suggest by mentioning "port b" (as in the outputs
for the second, left, extruder), then you can also just wire the heat sink fans power directly
to the 24V board input and just have it on all the time that the printer is turned on. No need
to play with firmware at all.

Dan

[Roy Kusuma]

Yes, it's the extruder heatsink fan. So my fan runs as soon as I turn on my rep 2. The LED next to Q11 (mightyboard rev H)FET is on as soon as I turn the board on. Which wouldn't really be a problem except that the voltage drops from 24v to ~18v as soon as it hits 50c with an audible rpm drop on the fan. The LED next to Q11 also dims ever so slightly when this happens. Quite convinced it's the MOSFET causing these issues. For some reason this slower fan speed is okay for 300 micron resolution but once I start doing higher resolution prints like 100 micron, there is a very good chance it would air print after about 30 minutes.

I don't really have a good explanation for this except maybe the slower feed rate allows for the filament to heat up higher in the tube and causing jams? Whereas if I do 300 micron prints technically it's pulling filament 3x faster than 100 micron prints and it might be fast enough where the filament softening issues are avoided.

I have ordered some new FETs and plan to resolder the q11 FET that I think is causing the issue.

So I don't actually have much experience in circuits but I can understand basics. So how would you go about wiring the fan to the 24v input if I were to follow your suggestion? Also, lets say I don't want to solder anything on my board, and I switch the extruder connection (with the extruder fan, stepper motor and heating rod) from port A to B, what do I need to do to update the firmware to think B is A?

Running Sailfish 7.7.

Thanks!!

[Dan Newman]

On 08/05/2016 5:07 PM, Roy Kusuma wrote:
> Yes, it's the extruder heatsink fan. So my fan runs as soon as I turn on my rep 2. The LED next to Q11
> (mightyboard rev H)FET is on as soon as I turn the board on. Which wouldn't really be a problem except
> that the voltage drops from 24v to ~18v as soon as it hits 50c with an audible rpm drop on the fan.
> The LED next to Q11 also dims ever so slightly when this happens. Quite convinced it's the MOSFET causing
> these issues.

No doubts there: they are absolutely known to fail and do so depressingly often. Just sticking
something into the fan whilst it is running can kill that MOSFET with the induced EMF (and physically
damage the fan as well by breaking fan blades).

> For some reason this slower fan speed is okay for 300 micron resolution

Do a big enough print and it will likely still fail -- just takes longer is all.

> I have ordered some new FETs and plan to resolder the q11 FET that I think is causing the issue.

There's old postings here on doing the part swap. I just cut the three legs, pull the body
of the part, unsolder the three amputated legs, then solder in the new part holding it down
while tacking one leg down. (I do it with the machine on it's side; no need to remove the
board.)

> So how would you go about wiring the fan to the 24v input if I were to follow your suggestion?

Find the two wires for the fan in that unpleasant harness MBI made, cut those two wires,
then run them to the +24V and GND inputs on the board. If you get the polarity wrong,
the fan won't turn. If you get the polarity correct, it will. Since this involves cutting
wires and later patching them back, you may not want to do this. It also involves
figuring out which two wires in the harness. (And don't cut the wires too close to the
header; otherwise, it will be difficult to resplice later.) You may just want to wait
to receive your new MOSFETs.

Dan

[Alex Reid]

I just ran into this problem with a Rep 2 Rev. G board. Got out the soldering iron and replaced Q4 (fan_cool) and Q7 (extruder 1) and my fans are running properly again.

In case anyone is interested, I used the DMN3404L (Digikey, Datasheet) and it works well. 

On Friday, April 29, 2016 at 11:40:21 AM UTC-7, Carl wrote:

Or... you could get a fan like this one that is ready to plug and play... for a quarter of the price! :-)

To repair the MOSFET on a Rev G Mightyboard... Have a look at 'Q7' (Q11 on a Rev H Board) on your controller board... you will probably see evidence of the failure...

I believe the original MOSFET used is a 2N7002K... it is rated for 300mA @ 60V... so it might be worth trying something that is able to handle a little more current... and then replacing both 'Q7' (Q11 on a Rev H Board), 'Q8' (Q9 on a Rev H Board)and possibly even 'Q4' (Q7 on a Rev H Board) at the same time. 

Two alternative options that have been successfully tested include the DMN3033LSN-7 and the MMBF170... but if these are difficult to get hold of... there are many others that could also be used! :-)

 

[Fred von Stein]

Hi,

  I just joined this forum, I've been out of the loop for several years since the MBI Google Operator's group was taken down. 

I have a Rep2X that hasn't been used in some time... I dug it out and tried to start it up with a 'preheat' and noticed the fans weren't spinning. 

I googled my way here, and found a likely culprit suggested for my Rev H board - the FETs were indeed blown when I looked.

I have two questions:

1) I have a 50W weller digital soldering station with pencil tip and decent soldering skills. Is soldering a SMD FET possible with this setup? Or must I really use a hot air setup? Or put another way, since I likely won't be buying a hot air reflow setup, what are my odds of pulling off a successful desoldering with my current setup? 

2) in addition to the exploded FET body, the leads look charred. The LEDs next to the FETs are on when I turn the board on, (I don't know why, since I would think this would indicate the fans were trying to run, even if I'm not heating) but the LED next to Q9 is pretty dim. Is there any info of upstream damage that I should investigate? I attached an image below for more info. 

[Daniel Newman]

On 29/10/2016 9:24 AM, Fred von Stein wrote:
> Hi,
> I just joined this forum, I've been out of the loop for several years since the MBI Google Operator's group was taken down.
> I have a Rep2X that hasn't been used in some time... I dug it out and tried to start it up with a 'preheat' and noticed the fans weren't spinning.
> I googled my way here, and found a likely culprit suggested for my Rev H board - the FETs were indeed blown when I looked.
>
> I have two questions:
> 1) I have a 50W weller digital soldering station with pencil tip and decent soldering skills. Is soldering a SMD FET possible with
> this setup? Or must I really use a hot air setup? Or put another way, since I likely won't be buying a hot air reflow setup, what are
> my odds of pulling off a successful desoldering with my current setup?

Use your Weller: it'll do just fine. I've replaced a number of these. I just turn the printer on its side, remove the bottom cover,
snip two of the leads on the FET, unsolder the third, unsolder the two stubs left behind, put a little solder flux on the pads, then
tag solder one lead of the new FET on, then solder down the other two leads, then touch up. All with a soldering iron no different
than yours; all with the printer on its side. Of course, you can simply turn it upside down if you want -- likely is a little easier.

Having good lighting and then checking your work with a magnifying glass is a good idea.

> 2) in addition to the exploded FET body, the leads look charred. Is there any info of upstream damage that I should investigate?

I've never encountered upstream damage myself. I don't recall hearing/reading of any either. However, others may
chime in with their experiences.

Dan

[Joseph Chiu (Toybuilder)]

A little bit of ChipQuik low-temperature solder can make the removal process even easier.  Soldering with a small pencil tip might give you a slight bit of trouble, as there is a pad that goes to the ground plane which sucks a lot of thermal energy.  You might find that blobbing a fresh blob of solder on the tip might be needed to give you a little extra heat capacity on your tip while you do this.

[Rich Webb]

+1 to both Dan's and Joseph's suggestions. ChipQuik alloys with the solder and lowers its melting point considerably, so it stays molten longer as it cools down. This lets you desolder/remove relatively large packages with just an iron and without damaging them. Very handy to have in your tookit.

But, if this is a one time deal and since you're not intending to reuse the parts, Dan's suggestion to just clip the leads is fine and what I'd probably do even having some ChipQuik and a hot air station on the home workbench. Blown parts, who cares. Just be a bit gentle clipping the leads; I've seen more than one instance where a tech acted like he was cutting 4-ought wire and managed to pop the device pads right off of the board. Eschew the oops!

[Fred von Stein]

Thanks for the advice,

  I removed the old fan FETs: I had success breaking the package and removing two legs, but a third leg (the one going to ground) was difficult to remove since it required a lot of heat. I eventually got the third leg off, but the pad did not feel as smooth as the other two - like there was still something there..?

 The LED next to Q9 that was previously dim is now bright. :o)

 I attempted to solder on two new FETs. I tried preheating the extruders again, but neither fan would run. :o(

 I removed the FETs and tried again with two more. Again, no luck with the fan coming on during preheat (I waited to 75°). Also, the LEDs near the MOSFETs are on, which I assume means the board is sending a signal to turn on the fans. I suspect I have a cold solder joint and I'm just not making a good connection.

 I am also worried that I keep applying more heat to these FET legs and might be perhaps damaging the FETs. I only have 4 left (I ordered extras). But should I be unsuccessful, is there any alternative I can pursue to fix my Rev H mightboard for my Rep 2X?

Thanks,

Fred

[Jetguy]

No, the LED near the fans is an indicator that the FET is actually working and is turned on. That LED is on the FET output in parallel with the output terminal of the board.
Your wiring or fan may now be suspect. Also, I just looked at the board since we do not have schematics and the inductors in series on the final output could be blown.

[Jetguy]

You can just solder bridge short the inductors, they are there for FCC EMI/RFI compliance. I wouldn't waste the time to replace them with exact parts if they had failed.
Again, if the LED lights, then the FET is working. It's on the output side so start walking the output part of the circuit.

[Jetguy]

Just trying to teach you how I know what I know.
#1 the basic premise that the same folks that built this previous board- made the board you have now https://cdn.thingiverse.com/assets/6d/94/65/9b/90/MakerBot_MightyBoard_REVE_Schematic.pdf
So while that is not the schematic of this board, conventions largely are followed.
#2 Googling that resistor value number "2471 resistor" http://www.mouser.com/ProductDetail/KOA-Speer/RK73H1ETTP2741F/?qs=BT1Sl5l6DqSGgRAVlgHhEw%3D%3D
It's a 2.74K Ohm resistor.
#3 If I know the resistor value, and I assume the source and function (current limiting for an LED, source is 24V) then if I do basic Ohm's law 24V divided by 2740 Ohms, I get 0.00876 Amps of current. Again, we take previous knowledge of a function (LED current limiting, most LEDs we drive to roughly 10mA AKA 0.01A. Also note, most turn on as low as 3mA and typical 5mA is you aren't driving for maximum brightness.
#4 So what I'm getting at, is that from assumptions of just looking at where components are in relation, trying to see the traces, we validate those assumptions with information and math. Again, I don't know the exact schematic, I may not be able to see all traces. However, if I take marked values, have a little intuition about the circuit, you can turn assumptions into better hard facts. That's how I can say I know the LED is indicating the true output of the FET.
The previous generation circuit worked that way
The same company built both boards- unlikely to change conventions
The values of components in the circuit match the assumptions and are valid proving even if I don't have a perfect schematic and cannot validate the traces or board path, the "function" is sound.
#5 Another talked about aspect of the Replicator 2 and 2X was Makerbot needed to get certifications for shipping worldwide. As such, we knew that certain parts of the circuit- especially any output had EMI/RFI limiting added as a function of the new boards. This is because every wire acts like an antenna and radiates any radio frequency noise generated by the board. Inductors on outputs choke off high frequency EMI/RFI. So knowing they did this to every wire output, this is how L33 and L34 (L is a symbol for inductor). So that means the inductors are like tiny fuses, they are just a coil of wire around an inductive core and if they are that small of a surface mount device, they probably cannot handle much current.
But the other aspect is, they are not "strictly" required for function of the circuit. They are just a filter on the output, a through device and due to construction and design, could potentially blow open circuit just like a fuse.

[Fred von Stein]

Hi Jetguy, or anyone else who might still be active here.

I recently got around to fixing my MightBoard Rev G for my Replicator 2X (does anyone use these anymore)? Well, I finally had a friend with better soldering skills than I microsolder on new FETs after I failed myself several times and gave up on this board/printer.

I now get the heat LED to light up when I go into preheat, and I get the fan LED to light up when the temperature registers 50°C. However, the fans don't turn on. I did a continuity test between the wire at the top of the printer that the fan plugs into and the connector at the board, the wire has no break in it. I then tried probing the third output pin on the connector, which would correspond to the red wire for the fan, but I have no output voltage.

I appreciate your picture of the inductors and mentioning that they may have blown to an open circuit, but my board is slightly different, and there is continuity across my inductors. How would I go about testing where my problem is to get the fans to working?

[Jetguy]

I circled the blown inductor. That's why you do not have 24V source on the red wire. Just solder blob to short the inductor.

[Jetguy]

Notice how that one is more brown than the others when you zoom in on the photo? Then also follow the red wire trace and it's right to it.

[Fred von Stein]

Thanks for the help, Jetguy.

  I measured the resistance of the inductors, and they were both 0.2ohms, which was similar to other inductors in that area. But I removed the red circled inductor and bridged the pads with a solder blob. I verified i have continuity from the output of the FET to the leg on the board connector. I turned on the printer, started preheat to 55° and the heat LEDs came on, at around 45° the heat LEDs went off and the fan LEDs came on. However, the fans did not spin. 

Could my fans be defective too? Or is there something else I need to check? And what caused all this to blow up in the first place?

Thanks.

[Jetguy]

Fan could be defective. That's what blows this up. I tried to explain, this circuit has NO margin for error. It is rated and sized just enough to be on the edge of blowing up normally powering a brand new fan. As a fan ages and the bearing dry up and stuff happens (broke blades, dust buildup, things stopping the fan) it starts drawing more and more power to spin. That's why the FET blew.

This is why I  warn to check your fan, ensure it spins free, ensure it's lubed and not making noise. Replace it before you blow a mainboard.