Printer cabinet

[conny_g 251]

Here’s my new printer cabinets to keep noise in my office down and to store them in a neat & tidy way.

Made by my favorite furniture maker.

PSU, cables, Raspberry Pi with OctoPrint are stowed in the drawers underneath the printers.

The printers can slide out for filament change or other maintenance.

The doors have gaskets so the noise can’t escape.

 

Edited September 2, 2018 by conny_g

[izzy 33]

Looks a really nice setup.

[KevinB 1]

Very space age, clean and proffessional!

[geert_2 558]

And what about the heat? Don't they overheat? Or is there extraction or cooling?

 

[conny_g 251]

No extraction / cooling yet. As the ones who print ABS (me not yet as most tries failed) close their cases to increase the temperature I am assuming / hoping that the printers do not mind a slightly increased temperature. In case this is not true for PLA and other low temp materials I can still add ventilation.

 

Did the first short test print today and the noise reduction is phenomenal:

The Synology disk array in the cabinet next to the printers is louder than the Ultimaker!

 

[gr5 2,239]

For PLA it will get too hot - you want the air temp below 30C.  But for ABS (or anything other than PLA) it should be great.  You want air temp around 35C to 40C.  This will work great for CPE/nGen/PET and nylon, etc.

 

If the PCB under the printer gets too hot the first thing to have trouble are the stepper drivers. These have built in temp overload circuits and will shut down typically only for 1 second but enough to miss steps if it happens in X or Y and if it happens in Z typically the bed will fall suddenly about 1mm (or more!) causing an underextruded layer.

 

In other words - worst case your part fails.  The printer will not be damaged unless them temp gets really extreme.  So the only modification (for non-pla) would possibly to add some cooling to the electronics under the printer only.

[foehnsturm 970]

Super nice.?

I already knew, you're well very organized person. ?

[conny_g 251]

8 hours ago, gr5 said:

For PLA it will get too hot - you want the air temp below 30C.  But for ABS (or anything other than PLA) it should be great.  You want air temp around 35C to 40C.  This will work great for CPE/nGen/PET and nylon, etc.

 

 

Just testing this, longer print in PLA.
After some 2 hours of printing the temperature in the cabinet reached 30 degrees, now let's see what happens in the remaining 3 hours.

With that slow increase of temperature there will not be much ventilation required to keep it down.

Some two 40mm fans on the back top of the cabinet pulling the warm air outside should be far enough.
 

[conny_g 251]

Ok, the first thing that was happening after 2,5-3 hours is that the feeder stopped feeding.

The thermometer in the cabinet showed 32 C and the 2nd extruder had 38 C.

So it's very well possible that the 32-38 C air temperature plus the feeder motor temperature (it's a BondTech) made the PLA too soft to be extruded.
As the PLA is in the printer unused already for 2 months I forwarded it by 50cm and restarted the print.

My expectation is, though,  that the same will happen and it will stop feeding at some point.

 

[conny_g 251]

After the 2nd try these learnings:

• After a few hours the temperature stabilizes at 35 C air (outside UM) and 40 C 2nd (unused) nozzle
• Together with the heat from the feeder motor the PLA is too soft for the feeder and the prints fail for underextrusion resp. "non-extrusion"

So I need ventilation for PLA.

Edited September 4, 2018 by conny_g

[owen 19]

Try one of these thermal switches. https://www.ebay.com/itm/9-C-99-C-DC-12V-Intelligent-Digital-Led-Thermostat-Temperature-Controller-/321723841989

You just need to hook up 12V and a good fan.

I've been using one with a 4cm fan for a while and it works great but on a hot day it's hard to keep the temperature down for the lower temp filaments so now I have hooked up an 8cm fan.

I set it to 45C for ABS and 30C for Ngen. I take my my enclosure off for PLA but if I had to use it I'd set it for 10C to 20C to that the fan is on all the time.

[conny_g 251]

2 minutes ago, owen said:

Try one of these thermal switches. https://www.ebay.com/itm/9-C-99-C-DC-12V-Intelligent-Digital-Led-Thermostat-Temperature-Controller-/321723841989

You just need to hook up 12V and a good fan.

I've been using one with a 4cm fan for a while and it works great but on a hot day it's hard to keep the temperature down for the lower temp filaments so now I have hooked up an 8cm fan.

I set it to 45C for ABS and 30C for Ngen. I take my my enclosure off for PLA but if I had to use it I'd set it for 10C to 20C to that the fan is on all the time.

I was planning with one or two ultrasilent 50mm fans per cabinet with 12m^3 / h each, that should do it. The increase in temperature was very slow, so there does not seem to be much air I have to remove to keep the temp down.

Would mount it/them on the top or top back side, back bottom there is already a hole for the cables where cool air can be pulled in.

http://www.blacknoise.com/site/de/produkte/noiseblocker-it-luefter/nb-blacksilentfan-series/50x50x10mm.php

 

Thanks for the tipp with the temperatur controller, looks like a good option. Was thinking about something I can integrate in Cura / G-Code or Octoprint like a self-made circuit attached to the Raspberry Pi (Octopi) that would be controlled by an Octoprint plugin that looks for a certain thing in the G-Code, which is configured in Cura in some way. But the Cura part is even optional, it's not much effort to open Octoprint and set the fan temp.
 

But maybe simple solution first and nice second and your proposed controller for the start... ?

 

[geert_2 558]

I would also suggest installing big diameter silent fans. Preferably with air canals running towards the ceiling, similar to the fumes extraction system in a kitchen. But then you would have to drill big holes in the cabinets above it, and install tubing in it. Just like in the kitchen. Maybe also increase opening in the back for a good equal flow at low speed?

 

This would also help extracting fumes.

 

Anyway, it looks really nice and clean, professional.

 

[KevinB 1]

I would also suggest grabbing a brand new roll of filament...fresh out of the package,

It will make trouble shooting easier.

 

And try a simple test cube..

 

With my enclosure..i have only seen problems with some low temp filaments like Ninjaflex or T-glase.

I have a thermometer on the top of my enclosure and it usually runs about 90deg F any higher for PLA and i pop open the top and front door.

 

Maybe try a print with the doors open, just as a test.

 

 

[conny_g 251]

This is what happens to the feeder with PLA if the environment temperature is too high:

The filament is torn and it starts to wrap around the feeder wheel...

 

Wanted to print some quick mounting bracket today and for that change the filament to silver. And somehow it wouldn’t move at all. So I dismantled the feeder and found this.

 

But also found that the feeder motor was rather hot even with the cabinet door open, checked the motor current and it was set to 1.200mA which is unnecessary much. 

I think to remember that they recommend 1000mA.

 

My 50mm Fans arrived, will add them to the cabinet this weekend and test. Have a plan B with 120mm fan if that does not work.

 

Edited September 8, 2018 by conny_g

[conny_g 251]

Mounted 2 pieces of 50mm fans in one cabinet for a test. That should extract some 24 m^3 per hour of warm air to the the cabinet above it - that has holes to the outside anyway (IKEA already has these type of cabinets open for any electric devices you might put there). As the warming of the cabinet was very slow - 10 degrees in 2 hours - I would hope that's enough to keep the temperature down enough for PLA.

 

Edited September 9, 2018 by conny_g

[conny_g 251]

After 5 hours the temperature in the printer cabinet stabilizes at 31 C, the 2nd extruder shows 36 C, the cabinet above the printer it's 28,9 C.

Print successful so far, 5 hours of 7 are done.

 

So it seems the fans and lowering the extruder current did the job.

But it feels kind of tight at the limit - the temperatures are not very much lower. So I will do another test with a 120mm fan which has double the air pressure and 5x the air volume than the two small fans.

 

[conny_g 251]

2 prints of this girl successful. But I am not completely happy with the temperature curve yet - see image.
 

The "exhaust cabinet" temperature (the cabinet above the printers where the fans are sucking the heat) is not going really high and the temperature in the printer cabinets is not really much lower than before.
I think so far the feeder current was the dominant driver as this generates more than 5 C more heat directly at the feeder. I think that this plus the +10C air temp caused the filament to become too soft to feed.

 

So I am trying the other thing I had in mind. Put an enclosure on the two holes for the small fans, but use a 120mm fan with 2x the static pressure and 5x the airflow of the two small ones.
(Hehe, that's fun: use a laser cutter to improve the 3D printer enclosure ?).

If this has slightly better performance as an indication I would drill another hole for more airflow or even make a complete cut-out of 200x50 which is about the same area than a circle of 120mm diameter.

 

[tinkergnome 927]

Perhaps a silly question:
While you use the fan to extract the warm air, is there also an inlet for fresh air somewhere? I guess this is quite essential...
And +1 for the 120mm fan!

[conny_g 251]

2 minutes ago, tinkergnome said:

Perhaps a silly question:
While you use the fan to extract the warm air, is there also an inlet for fresh air somewhere? I guess this is quite essential...
And +1 for the 120mm fan!

 

Yes, there's a hole at the bottom back, underneath the sliding table for the cables going into the cabinet. It's not that big (50-60mm) and one of the next optimization steps. But first I want some indication that this principle works at all. Don't want to put lots of holes into the expensive & nice cabinets before having some proof of concept ?. So I am trying to approach it in small steps carefully.
For this I now want to see the two lines to change their profile: the printer cabinet line needs to stay lower and the "exhaust cabinet" one needs to go steeper. 
 

[conny_g 251]

Not exactly "thrilled" by the result ?

The 120mm fan did perform slightly better, but more in a negligible range.

The cabinet temperature increased slower and stayed 2-2,5 C below the temperature with the 50mm fans.

And the cabinet above increased slightly quicker and higher, so more heat was put there.

But 2-3 degrees does not really make a big difference. Now it would be interesting to do a final test without any fan.

(And btw, there is another difference in that last diagram: the temp curve of the 2nd extruder - gray line - is very different, goes up rather steep in comparison the the last test. That's probably because I had the 120mm fan in the cabinet to mix the air and avoid layering of the air temperature and that fan I used for the next test. And it seems there the temperature up in the cabinet at the height of the 2nd extruder was higher than for the thermometer at the bottom, 4 degrees of difference.)

 

It seems that it needs more / larger ventilation holes to make this work with PC fans.
The problem is obviously their low static pressure - their airflow decreases drastically with pressure they have to build up, that's at least telling me value of static pressure and the corresponding diagrams "pressure vs. airflow".
But I don't really want to make the case that open, that will harm the noise reduction effect which is the key reason why I have them.

Another option would be a test with a real 230V ventilation / duct fan. They seem to have higher air flow (2-5x of the 120mm) and like 100x the static pressure. That should also work with small holes for ventilation. But they are much noisier and much more expensive. The quiet ones (25db and less) are beyond 100 Euros.
Or would anyone know a duct fan that is like 12V/24V and low noise, i.e. a small variant of the 230V building installation duct fans?

I think I will do a final test without fan to know the difference. And if the 50mm or 120mm fans do make a notable difference I will just keep that until I run into any issues with longer prints. A temperature of 33-35 C does look doable at this time, will just collect more experience and react if necessary.
 

Edited September 11, 2018 by conny_g

[geert_2 558]

Personally, I have the feeling that those little holes are not enough, and neither are the small fans. I would rather go for 2 silent 120mm fans, or one 200mm fan, or so.

 

But it could also be that the extracted air volume would be enough, but that the airflow does not reach the parts to cool? For best results, you would need a diagonal flow. Maybe try inserting a few cardboard pieces to guide the flow?

 

What you could also try: temporarily replace the front door with a piece of cardboard, in which you mount the fans. Fold and glue it in place with removable tape. So you can try the effect of different types of fans, and locations, without drilling holes everywhere and ruining the cabinets.

 

And of course, the cabinets above and below also need big enough openings on their other sides, and between walls and cabinet. You have to add all these resistances in series. So total resistance is probably 4x higher than that of the one of the printer alone.

 

Concerning current, I have set mine to 900mA, if I remember well (UM2 non-plus). But that was to prevent the X-, Y- and Z-steppers to skip steps when too hot. Not the feeder.

[conny_g 251]

Yes, you are absolutely right. The fans and holes are too small. Did the control test with fan off yesterday and to my surprise and frustration the temperature values don’t differ too much. Maybe 1 degree.

It seems it has much more impact if the test was run right after another one or after 10/12 hours of break, the curves are very different in these cases. Means they are 2-3 C higher if there was a test right before, even if the starting temp is the same ~25 C. Logical, the hole cabinet stored heat and will absorb less right from the start.

 

Looking into datasheets of fans some more I am finding that real house ventilation duct vents have 10-20 times the static pressure vs the best pressure optimized PC fans to get air through the 10cm holes in the wall. And for that they are 50dB mostly instead of 20-25 ? 

Thats the key issue here why not too much air is taken from the cabinet. It needs to be 5-10 times as much air to make a real difference.

Thinking about making a test with a duct fan now ?

Or with a vacuum cleaner, they are designed for strong pressure resp. suction ?

Alternatively I build a small A/C into the cabinets, i.e. active cooling....

 

But on the other hand even this last print was successful. So it seems that the reduction of the feeder motor current brought the level of the filament temperature in the feeder to an acceptable level and there is no issue with the air temp of 35  C at all.

I should just keep on printing and watch out for symptoms before spending more time on this.

 

[geert_2 558]

For comparison, my printers are sitting in the fume extraction cabinet in my laboratory. This extraction cabinet has an exhaust pipe diameter of 200mm. And the rectangular front opening is ca. 1100mm x 50mm. All this to get a smooth equal air flow.

 

Although this is designed for extraction of chemical fumes, it is just standard lab equipment, I guess the concept can be ported to extraction of heat too, since both are about getting enough air flow, and good equal distritbution.

 

I have no idea what fan is in there, since it is a huge central fan on the roof for all cabinets. But you can really hear and feel the airflow.

 

Yes, you could disassemble a mini-fridge (like those in hotel rooms), put the cooling radiator inside and the heat radiator outside. But then you are going to have that constant compressor rattle. Or you could disassemble a cooling box based on peltier-elements and use these. But I doubt if this has enough capacity? It has to remove 100 up to 200 Watt of heat. And you may get condensation and corrosion. Probably a big fan is still easier...

 

[Julian2801 1]

Looking really nice! I'm planning to build something similar (only for one printer though). Are you using some noise-canceling materials or is the wood cabinet doing the job alone?