Using Actuator to Automate Open/Close of Computer Side Panel

Community Projects
1

Hi, so I already got my PC side panel on hinges!:

http://www.youtube.com/watch?v=QE1ucislSxU

I would like to be able to press a momentary push button to open it, and then push it again to close it. I’d like it to run off battery (think of it as a mod to the case, not the system), but then run off the PSU when it can. The hinges itself take nearly zero force to push open/close (and I haven’t even lubricated them yet), but the 3/16 rare earth magnets that keep the panel shut tight do. I don’t know if it’s more or less force depending where you push, but I pushed a scale at the halfway point and it was ~25oz to pop it through the magnets.

Bear in mind that I have a lot of experience with modding computers, psu’s, wiring and such, but my experience with servos is an RC plane that never got off the ground when I was 14, and zero experience with microcontrollers/programming/arduinos/etc.

So far, this is what I’ve got:

  • Adafruit Trinket 5v $10 shipped. Cheapest microcontroller, just to program the pushbutton control. I originally was going to go with a pololu. And is it possible to make a programmable servo work so i dont need a microcontroller (although that may not be economically practical, i dont know exactly the costs of a servo programmer)

  • High Torque Micro Servo ~$20. I think I’ll need high torque to push through those magnets. The rod has to be mounted below 1.25" or else it’ll end up on the window itself (which is okay I guess, but I’d rather not). I guess I could buy a full size servo and just have the rod pointed downward, not sure if that’s a good idea though.

  • Using 4xAA Batteries in Series for 6v, and I’d buy a 4xAA battery housing unit from ebay for $1 so replacing batteries is really simple (not for me, but say if i sold the case).

  • Mounting hardware from local hobby shop

  • SPST Switch

Now, I am a little confused on battery. Ideally, I’d like to do something where I ran off the battery when my PSU was uninstalled/unplugged, and then PSU when it was plugged in (ie 5VSB line, instead of 5v rail, so i’d get 5v even when pc was off). Or anything similar to this (ie run off battery and then battery is charged by usb/5v, etc). However, I can’t find any battery charging circuits above 3.7v or 4.2v.

So I have a lot of options here I guess. I’d appreciate any suggestions on anything here really. Thanks!

2

Forgetting about the magnets for a moment, how is the servo going to apply force on the door to open and close it ? Some crude drawing could be posted and would help. Here’s my concern, when I see door openers they usually are mounted and apply force near the hinges. Meaning you’ll need a LOT of torque there to overcome the magnet’s force waaaaay at the other end of the door. It’s almost as if you need another (momentary) “pusher” near the magnets … or something other than magnets.

Can I proper assume the door is steel (or some other ferrous metal) ? That the frame magnets latch to the door itself and not other magnets in the door ?

3

I would suggest that instead of using magnets that you use actuators leveraging this project as your baseline: http://www.youtube.com/watch?v=cvMiXJl_mjg

You can even utilize his RFID method for security.

4

The servo would be by the PSU on the bottom. The horn would be pointed forward, and it’d have a pushrod on it attached backward, mounting to the side panel on a point close to thinges behind the servo. The servo would turn clockwise and the bend would straighten.

The magnets, when I applied it in a calc, was ~.62lb force (3/16 N35 magnets) each, but I have them slightly distant, so their force should be alot less than that.

I mean the mg90 is like 35oz of torque?

5

Belial88:
The servo would be by the PSU on the bottom. The horn would be pointed forward, and it’d have a pushrod on it attached backward, mounting to the side panel on a point close to thinges behind the servo. The servo would turn clockwise and the bend would straighten.

The magnets, when I applied it in a calc, was ~.62lb force (3/16 N35 magnets) each, but I have them slightly distant, so their force should be alot less than that.

I mean the mg90 is like 35oz of torque?

So if you have 35 oz-in of torque (your #, not mine), how much is needed to over come the magnets at whatever distance they are from the hinge ? If the magnets are .62 lbs of force, aka ~10 oz but they're at 18" (a SWAG) from the hinge, you'll need 180 oz-in of torque to over come their force. You are woefully short unless the true magnet force is (as you say) a lot less. And that's assuming the servo can apply all it's torque. Depending on where the servo pushes on the door you may gain back some of the leverage lost to the magnets location. See the illustration below.

(click on to open)

(by the PSU on the bottom assumes I know where the PSU is. I took my guess above)

6

That image is close. I’d have the servo at the far left of it’s movement, so it wouldn’t be going from 0* to 90*, but -90* to X, however far was necessary, probably something like 60*. And I think I’d have the pushrod mounted further behind the servo. I know that doesnt do the ‘optimal’ 90* rod/horn relationship, but it creates a shaper angle, as the extension of the door relies on the angle between horn/rod goes from small angle to large angle as it opens up, since I don’t really have much room to work with here (ie i cant use a huge horn).

Maybe the magnets are a lot of force, I don’t know. If my towerpro mg-90 ends up not beign strong enough, I’ll get an hs-85bb and use a ball bearing to help compensate for the rod being slightly angled downward.

I ordered an

Adafruit Trinket 5v $11

Towerpro MG90 Servo 35oz 6v Ebay $11

4xAA 6v battery pack ebay $1

I’ll hook up the trinket and servo in parallel to the battery pack. I wish I could do the charger thing but I think it’s impossible to find a charger or something that switches between the battery or psu if psu is on.

7

I still don’t think you can overcome that force, especially if you are moving the servo to the left even farther. :? You should probably consider other options because even though the door moves “freely” you still have to overcome the door’s weight in addition to the magnets. And while it may only be an oz of force or two, it adds up.

Suggestion would be to use a latch like in the video I posted above.

For power, you could always use: https://www.sparkfun.com/products/10711

8

As I understand that lipower shield correctly, it’ll take a 3.7v battery and make it output 5v, and it also can be charged via usb/5v at the same time? Sounds great, though it’s a tad expensive.

I’m going to hold off a sec on alternative batteries, as since there is worry the servo I got might not be strong enough, I’ll see if 6v is good or not. Can’t use some 5v supply if 6v doesn’t work.

9

Belial88:
That image is close. I’d have the servo at the far left of it’s movement, so it wouldn’t be going from 0* to 90*, but -90* to X, however far was necessary, probably something like 60*. And I think I’d have the pushrod mounted further behind the servo. I know that doesnt do the ‘optimal’ 90* rod/horn relationship, but it creates a shaper angle, as the extension of the door relies on the angle between horn/rod goes from small angle to large angle as it opens up, since I don’t really have much room to work with here (ie i cant use a huge horn).

My GUESS is that your single servo will need some help to overcome the magnets. Two ideas for your consideration are :

  • some solenoids that push the door open or

  • another servo (or gear motor) that runs a “cam” that pushes on and forces open the door. Both would be simple to add on (space permitting) after the initial trial.

10

Or remove the magnets altogether and rely on the servo to keep the door closed.

Mee_n_Mac:

Belial88:
That image is close. I’d have the servo at the far left of it’s movement, so it wouldn’t be going from 0* to 90*, but -90* to X, however far was necessary, probably something like 60*. And I think I’d have the pushrod mounted further behind the servo. I know that doesnt do the ‘optimal’ 90* rod/horn relationship, but it creates a shaper angle, as the extension of the door relies on the angle between horn/rod goes from small angle to large angle as it opens up, since I don’t really have much room to work with here (ie i cant use a huge horn).

My GUESS is that your single servo will need some help to overcome the magnets. Two ideas for your consideration are :

  • some solenoids that push the door open or

  • another servo (or gear motor) that runs a “cam” that pushes on and forces open the door. Both would be simple to add on (space permitting) after the initial trial.

11

The magnets definitely need to be there, but I can have the servo slightly pull on the door so that the magnets dont need to be as close, which drastically reduce the force of them. Like if I push/pull the door tight by even the halfway point, it still hangs barely open at the edges.

I don’t know what you mean by solenoids to push the door open, and I’m not sure how I’d do the cam thing for another servo on the door.

12

Belial88:
I don’t know what you mean by solenoids to push the door open, and I’m not sure how I’d do the cam thing for another servo on the door.

I mean you put a solenoid (or 2) mounted to the case where the magnets are. It's piston, when activated, pushes through a hole in the case and onto the door, forcing it away from the magnets ... helping the servo to open the door. A motor or servo using a cam lobe that rotates and acts through a slot in the case does the same thing ... pushes on the door end where the magnets are. Both need only be activated for the initial push to overcome the magnetic force. Your servo and linkage (not shown below) take over from there.

(click on to open)

13

What about an electromagnet to neutralize the field of the earth magnet during the unlock sequence.

14

I have no idea how to use an electromagnet to neutralize the field of the earth magnets lol…

Very cool idea, basically set up a helper servo to just push the door. I think if the magnets are too strong though, I’ll just get a micro servo like an hs-255hb, i mean the rod doesnt have to be perfectly straight and i was just wrapped up with the servo having to be under the door window and didnt realize that using a coupler would allow the rod to be even with the arm, rather than mounted on top.

15

Following this closely for a mod or two of my own that I have in mind. I have a HAF932, so pretty similar case. I’m planning on a very custom ‘skin’ for my case following an overall theme. Having the door do this would be a very nice addition, so I’m hopeful you/we find a way to make this work. I’m fairly new to all of this stuff, mostly hanging around case modding. Good luck, and this place is such an awesome resource.

16

Valor958:
Following this closely for a mod or two of my own that I have in mind. I have a HAF932, so pretty similar case. I’m planning on a very custom ‘skin’ for my case following an overall theme. Having the door do this would be a very nice addition, so I’m hopeful you/we find a way to make this work. I’m fairly new to all of this stuff, mostly hanging around case modding. Good luck, and this place is such an awesome resource.

If I were to do this and were to use the magnet design like the OP, I would use the two solenoid actuators as Mac suggested to over come the force of the magnet to allow the servos to move the door. If you do some creative positioning , you could likely have just enough force that it doesn’t snap the door open hard and allows the servos to create a nice smooth motion.

17

I’ve got the hinges already working, I just got all the parts so I’ll be installing it over the week.

That said, haf 932? Hope you didn’t buy that recently ;/ It was THE case about 5 years ago when it came out, but nowadays the haf cases are far outdated and overpriced compared to new budget cases. The HAF932 incorporated features only found in high end cases back when it was released, forcing every other company to do the same, but since then many newer cases have come along and done it (Corsair, fractal, bitfenix, zalman atm).

18

Yeah, I got it about 3 years ago now… how time flies lol. I used to be awful proud of it. It’s still a great case if you already have it, but no… right now I wouldn’t buy it again. How hard was it to get the hinges working?

I’ll start a project log on my case eventually. I’m renewing my wedding vows, and re-doing our wedding (since the first run kinda sucked… we both agree lol) on Nov. 9 and we have a gift registry this time. Hopefully I get some of the tools I requested so i can get back into my modding. All my old tools were my grandfathers and he took them when he moved :\ Just gimme a Dremel and I’m set :smiley: lol

Glad to hear you’re making progress. I’ll update my post here as I can, and hopefully I get some insight into what I’m trying to do. I’m VERY new to the whole arduino bit and all ‘this stuff’…

19

I’ve come to learn that side panels are actually manufactured very poorly, but it looks snug and tight because you are essentially brute forcing the panel and warping it when you slide it in those notches + putting the screws in (you might notice on the top of the 90* bend on the side panel, it’s flush but the other side isn’t flush until you really screw it in). Putting on the hinges is as simple as gluing them on (you can put them anywhere, but the easiest way imo is just glue them inside to that 90* bend and the outside corner of your ccase).

I’ve seen a few people doing the side panel mod and actuator mod, use supports behind the door to make it sit straight, like metal bars, but I don’t know how effective it is.You need to use something that ‘pulls’ the door flush, and forces that little rail on the bottom to slide up against the case (hard to explain, look at the bottom rail of your side panel on the inside and you’ll see what I mean, it makes sure you put the panel on right). Rubber band, magnets, a lotch of some kind, etc. Rare earth magnets are cheap and imo, pretty awesome. I could just stop at the hinges mod with magnets and that’s awesome in and of itself.

Just use epoxy of some kind, and use lots of painters tape to hold things in place.

Dremel is pretty awesome, I’m just looking for excuses to use it. But I’ve never really had a use for it before - I cut my 140mm fan holes with a hole saw (that’s $40… just buy it and return it), I’ve always cut my windows with a jigsaw (large, straight cuts, shinobi and source 210 have very straight angles), and all the heatsinks I’ve modded onto things I’ve cut with a jigsaw. I got a dremel 4000 because it’s got some thing where when it cuts and slows down, it’ll speed up reactively, that and it’s just supposed to be very high quality. I paid i think ~$60-70 for it on ebay new with a ton of stuff included, vs the cheapest dremels you can find are maybe $40-50 so it was kinda no brainer. You can sometimes find some on craigslist for ~$20-30, but all the ads I messaged didn’t respond.

I do a lot of budget modding, I don’t have much money (or rather, I don’t spend much money on computers because I don’t think you need to). I really enjoy making great mods for super cheap, or getting high performance out of cheap stuff (ie finding hynix ram for $50 that can overclock to 3ghz+ like $300 ram, modding $20 closed loops to fit on graphics cards to outperform $100 aftermarket gpu coolers, window mod, etc).

20

Okay! Got the materials!

So I’ve got my adafruit trinket and it’s all ready to go, I set up arduino IDE program, and put an adafruit softservo master in the library.

  • i am totally lost on code… I understand that each push of the momentary SPST 2 pin switch I have will ‘change the state’ but I have zero clue how to code this stuff.

  • I’m also a bit lost on how exactly to wire the momentary switch. Do I need a resistor or something here? I’m using just a simple computer power button for this (i like it’s look) but I do have proper momentary smt spst’s if necessary…

*8mhz vs 16mhz? As I understand, 8mhz is 33% more efficient (9mah vs 12mah) so wouldn’t I rather run at 8mhz since I’m running off 4xAA@6v batteries?