3D Printed Barrel Servo/Fan


 

Tom Kole

TVWBB Pro
I'm updating this thread to reflect a few observations that I've made regarding the servo dampers. The barrel servo has worked well for me, but in the last few months I have found that the rotary valve design is much easier to print and probably works even better. I will leave the barrel servo information up but am also including the source files and pictures of my rotary designs. There is an offset and in-line version.

Offset rotary valve

I designed this based on the rotary valve concept and a idea that Bryan had started but I think he abandoned along the way. My goals were to have a physically attached blower with completely separate air ducts and motorized components, a wider opening at the blower inlet, and to make it so that I could mostly shield the electrical parts from the weather since it seems to rain every time I bbq. This is what I got:

12150983415_688279bcd6_b.jpg


12150961975_abe3ef22e4_b.jpg


12150961945_5452a98da4_b.jpg


12150961895_fb4792335d_o.jpg


It prints in 3 pieces without any supports. The cover piece pops on and can be secured along with the rest of the unit to the 2" x 1" square tube that attaches to the smoker. The cover doesn't actually play any role in sealing the air ducts, it's just there to keep water out of the electrical/mechanical chamber.

Here is the graph of my torture test. Bear in mind that there is no food on the smoker which could alter the response times.

12151627096_1dc8f236a1_b.jpg


Large BGE
B: 0
P: 3
I: 0.008
D: 5
min fan: 10%
max fan: 100%
both fan and servo on

Anyone with access to a 3D printer that wants to try it, here are the files. The raw files were created in Inventor fusion and I also put up the corrected .stl files.

The blower is different than the stock blower listed in the BOM. I found a smaller footprint blower on digikey that matched the original blower in static pressure, airflow, and current draw:

http://www.digikey.com/product-search/en?x=0&y=0&lang=en&site=us&KeyWords=603-1370-ND

I couldn't find it's 2-wire version so I just cut off the tach wire.

In-line rotary valve

13798732244_1d5f2ebb28.jpg


13798425585_21faa6f4f0.jpg


13798427465_ae8258a76b.jpg


Here are the files.

Barrel servo

I'm starting a new thread for my new 3d printed barrel servo/fan combo instead of hijacking the 3d print thread. It is designed for the standard blower listed in the heatermeter BOM and the servo's used by D Peart in his original design. A homedepot TECH brand RJ11 jack for 6-conductor wire is incorporated to carry fan and servo control over a single wire. It is designed to mount onto 1" square tubing. All screw holes are designed for M3 hardware.

As of version 2, I started using MG90s servos for better torque and durability. You can find them here, or ebay. They are slightly different in size than the SG90s so I have two different designs. If you don't have any servos yet, I'd go with the MG90s. Version 2 will be the last version to support SG90s.

Version 2

I tweaked the barrel cover so that the barrel and electronics are completely covered and created a conduit for the fan wires to travel to the rj11 jack.

10567215313_182a443a82_b.jpg


Here is a video demonstrating it's capabilities. It starts with the fan on 100%/servo valve open, then goes to fan off/servo valve closed, then back to fan on 100%/servo valve open, and finally to fan on 100%/servo valve closed.

Barrel damper

Here are the files.

The .stls are already repaired via netfabb. I suggest using the support feature in slic3r for printing the cylinder.

Mount
The mount can be fabricated in the following manner:

Supplies: aluminum sheet (I forget the exact gauge but is should be fairly bendable to conform to the curvature of the smoker), 1" square aluminum tubing (barrel damper) or 1" X 2" aluminum tubing (rotary damper), a 1" chisel, hammer, metal snips, JB weld or high temp silicone, and a drill with 1/8" drill bit.

1. Cut out a 77 mm square (1" square conduit) or a 105 mm X 77 mm rectangle (2" X 1" conduit) from the sheet metal (for large BGE, adjust dimensions for your smoker)
2. Draw an "H" in the middle with dimensions of 1"x 1" (1" square conduit) or 2" X 1" (for 2" X 1" conduit, make the "H" two inches tall)
3. Use chisel to pierce through metal along the borders of the "H"
4. Use hammer to bend back the newly formed tabs at 90 degree angles
5. For the barrel damper, insert a 2" piece of 1" square aluminum tube through the square hole until the edges line up with the end of the tabs. For the rorary damper, insert a 3" piece of 1" X 2" aluminum tube through the square hole until the edges line up with the end of the tabs.
6. Use 1/8" drill bit to drill a single hole through the tabs and metal tube
7. Use an M3 or 1/8" screw to fasten together
8. Place JB weld or high temp silicone along the tabs and edge of the tube that is protruding through the sheet metal to form a nice airtight seal
 
Last edited:
Graphs

First trial run using version 1 mount on a 2 inch piece of 1" square aluminum tubing. The PID settings are unchanged from my Auber setup.

PID settings
B: 4
P: 3
I: 0.005
D: 5

Fan 10% to 100%, servo 0:90 degrees

317gzzt.jpg


Here is a pic of the mount and barrel servo in BGE color. There is JB weld on the inside of the mount to seal around the edges.

10566847183_99b96b556e_b.jpg



Latest PID settings

1. Large BGE
B: 0
P: 2.5
I: 0.0035
D: 6

Min fan: 10%
Max fan 40%
servo: on/off

Top Vent:

10566505395_4b4d105b9e_b.jpg


10566521776_4ecfa2140f_b.jpg


Graph:

10572730963_a8c2d721c0_b.jpg


When the area that I circled in red started to happen, I checked the BPID contributions and didn't like the way the heatermeter was responding to the perturbation so I adjusted to the following:

B: 0
P: 2.5
I: 0.0035 (I'm not sure that it even goes to the 10,000ths place)
D: 6


By setting the heatermeter this way, the fan only kicks in if the blower reaches >= 10%, but the servo is full open at any value greater than 0. You can see that after the initial rise to temp, the fan has not kicked in at all and the servo has been able to hold the set point very tightly just by going on/off. The bonus is that if the system gets overwhelmed at some point and needs a little extra oxygen, it can call on the fan and still do so in a controlled manner without the fan coming on at 100%. You can also leave this setting to control higher temp cooks as well with only a set point adjustment.


2. Vision Kamado
B: 0
P: 1.5
I: 0.0035
D: 6

Min fan: 10%
Max fan 100%
Running in fan and servo mode

Top vent:
ztRqVBI.jpg


Graph:
NBvMbe1.png
 
Last edited:
As always, excellent 3D modeling work there Tom....
It's been a while since I've printed anything, I'm wondering if you're gonna share the stl's with us? I'd love to give this a go on the printer and test it out on my FauxMado...
 
As always, excellent 3D modeling work there Tom....
It's been a while since I've printed anything, I'm wondering if you're gonna share the stl's with us? I'd love to give this a go on the printer and test it out on my FauxMado...

Thanks. I knew I forgot something. I linked the files in the first post.
 
Kool, I'll give the print a shot sometime soon and let you know how it goes. Thx for sharing the 123D files as well, I'm gonna have to modify the output from square to round to fit the 3/4" copper pipe I use on my vent plate... Thanks again for you hard work on this and the case(s), and again for sharing...
 
Tom, your 1" square tubing on your egg....did you fabricate that or buy it somewhere?

Fabricated from aluminum tubing and thin aluminum sheet purchased from home depot. It's basically a 77 x 74 mm rectangle with an "H" pattern that I cut into the middle of it. The "H" is made so that you have two 1/2" pieces of the aluminum sheet that you bang out into right angles. You then slide the piece of tubing over them, drill a 3 mm hole right through, and then bolt on. I'm using JB weld around the perimeter to create an airtight seal. It's not the fanciest design, but it's simple and should work. I'm using my CNC mill for the cutouts because I have one, but a pair of snips and a sharp 1" wide chisel should work perfectly. It fits into the vent guide of the BGE and then you just close the vent to hold it in place. I just copied the idea from the Auber fan mount.

If you know how to weld (need to learn this) you can make something much better, or if you could find the parts in copper, you could solder them together much like Ralph has demonstrated in some of his designs.
 
Last edited:
Tom, is there a possibility that you will be making this housing for sale for those of us who do no have a printer? TIA
 
Tom, is there a possibility that you will be making this housing for sale for those of us who do no have a printer? TIA

I want to put it through some real world tests before doing that. If all the 3d print guys try it out, we should get some good data using different smokers.
 
What are you using for layer height(s) when you slice the parts? I sliced at .3 first layer and .4 on the rest and printed it up last night. I haven't completely cleaned up the parts yet, but they seem to mate with a bit of space around the barrel and the valve offers little resistance to air flow when it is closed. Not sure if it will close better when the parts are cleaned up, or sliced with thinner layers....

When your valve is in the closed position, if you put suction on the output (grill) side of the valve, how much will air flow through the closed valve on your build?
 
I use 0.3 mm layer heights for everything. There is 0.25 mm space built in laterally between the base and the housing to allow for variability in print and tube accommodation. That gap closes upon tightening of the mount screw. There is also a preferred orientation of the valve, The side with the rounded edges should point towards the fan. Lastly, you need to make sure that your limits are set properly in the software. I didn't end up cleaning the parts at all with acetone because it worked so well right off the bat. I will hopefully mount it to my square tube tonight so I can answer your suction question.
 
Also, the metal tube is actually part if the design.

How so? Then I am missing that part for sure, I just converted the square output to round so I could fit it to my 3/4" copper adapter. Testing my valve by drawing air through the output (without the copper tube connected) I find it offers very little resistance to air flow when closed, that is why I am wondering if your valve is the same or if it is somehow tighter?

The fan blow test doesn't really say much, there are lots of places for the air to go rather than through the output and the static pressure of the HM fan isn't very high to start with. A suction test is a more real world measurement of how it will perform, as the valve needs to prevent draft from the fire which is essentially a suction....
 
Last edited:
The blower test simply tests the barrel portion. You are correct in saying the real world scenario is suction. However, I think you are a bit off in your estimation of the negative pressure generated by the smoker. I designed the valve so that the metal tube is held in place by the base and the housing to create a solid conduit to the barrel valve. Without the tube, there will be leaks along the length of the device. So as I'm stating for the third time, I'll get a better idea of the performance once I mount the tube. Thank you for reminding me that I should never release anything until I've completely finished all tests.
 
I can't really tell from the photos Tom, but what is the slot for next to the RJ11 slot where the servo goes? I'm assuming it has something to do with the servos?
 

 

Back
Top