TurboCAD Forums

The Ultimate Resource for TurboCAD Knowledge

Register
 
Interested in some really terrific mobile apps? Visit www.turboapps.com for details.

Real World Info Needed
Read 11081 times
February 11, 2010, 12:40:45 AM
It looks like I will have to fabricate a NEMA 1-15 Receptacle for a project (I can't find a PCB-mountable model anywhere, and the wired models I have found are too darn big).

A 1-15 Receptacle accepts 2-pin plugs on those wall-mounted transformers used to charge cell phones, iPods, cordless drills, etc.

A typical prong is 0.6"-0.7" long, 0.25" wide and 0.058" thick.  To mate with it, I want to make a U-shaped piece that has a gap of 0.05" with a flange on the forward face to guide the prong into place (see approximation in attached image).

What are standard thicknesses for aluminum sheeting?  I have found sites like http://www.precisionsheetmetal.com/home/thickness.htm and http://www.onlinemetals.com/merchant.cfm?id=76&step=2&top_cat=60.  A cursory review shows that the listed values are not the same.

Does anyone have the experience to recommend which aluminum to use?  I usually saw 6061-T6 specified on the fabs made at the last two small companies I worked for.  Is this suitable for use as a clamping connector in an electronic application?

Also, what is the smallest bend radii for the recommended material?  


Thanks in advance,

Jeff



[attachment deleted by admin]
« Last Edit: February 11, 2010, 12:44:17 AM by Jeffin90620 »

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


* February 11, 2010, 03:16:22 AM
#1
6061 T6 describes an extrusion alloy.  It's slightly higher in silicon content than the 6063 commonly used for window extrusions and the "T" part of it refers to period in a tempering oven.  It's more brittle than 6063 untempered, tempering T6 would probably make it unworkable for the folds you're describing, Jeff.  I don't know whether aluminium is the right material for an electrical connector like that, it would probably be reactive, most electrical/electronic prongs/barbs are steel, nickel, brass or gold plated, which might invite electrolytic differential.  Most I've seen are brass or other copper alloy if not plated steel.

Logged


* February 11, 2010, 08:40:06 AM
#2
I agree with Murray (except that 6061-T6 is available in sheets). That particular alloy won't bend as tightly as your sketch indicates. If you must use aluminum, try to find alloy 1100-O, which is soft enough to accept tight bends and has  excellent corrosion resistance as well.

However, I agree also with Murray's skepticism about aluminum being suitable at all for an electrical connector. Although the material does have excellent conductivity, it tends to acquire an oxide coating which can seriously degrade the performance of an electrical connection. While alloy 1100-O might be better in this respect than other aluminum alloys -- and I can't say for sure that it is -- I really think you should consider a different material, such as plated brass.

Henry H
« Last Edit: February 11, 2010, 03:36:17 PM by Henry Hubich »

Logged


February 11, 2010, 09:26:47 AM
#3
I was thinking that the aluminum would be plated with something to provide long-term electrical connectivity, but it seems I still have some research to do.  Brass is, of course, ideal, but there are cost concerns (this will be mass-produced).

Thanks for the advice.

Jeff


Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


February 11, 2010, 10:32:39 AM
#4
to answer a couple of your questions, you can find information from the mill about available sizes.  Here's one I use often to get an idea of what materials are available.

http://www.ryerson.com/stocklist/StocklistServlet?COM=GetHome&REF=1

Just click on the metal you're interested in and look at the stock list available.  It's very useful.  There is also a Data section to get the mechanical and chemical properties of the materials.

I also like this website for formulas and better instruction about various topics.  You may find it useful for what you're doing as well.

http://www.engineersedge.com/mechanics_material_menu.shtml

As far as a bend radius for aluminum, I typically have used 2t for 3000 and 5000 series aluminum or 2.5t for any other as a general rule of thumb just to prevent cracking.

Logged


February 11, 2010, 10:50:49 AM
#5
Thanks for the links.  I will look through them.

Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


* February 15, 2010, 12:28:54 PM
#6
I was thinking that the aluminum would be plated with something to provide long-term electrical connectivity, but it seems I still have some research to do.  Brass is, of course, ideal, but there are cost concerns (this will be mass-produced).



Jeff,

The oxide problem others have mentioned is serious and not readily solved.  Aluminum was used for house wiring many years ago and many houses caught fire as a result of overheated connections due to oxide formation.  The use of Al in electrical applications is heavily restricted in most codes. Plating is not an answer since aluminum does not accept electroplating or soldering.

Yellow brass or nickel brass (white) would be your best bet.  Do not be fooled by material costs because the real cost you need to know is fabricated cost.  Brasses are widely used for electrical parts and most fab shops are very familiar with it.  It is likely the savings in fab costs will more than offset any difference in materials costs.

Logged


February 15, 2010, 12:54:51 PM
#7
I was thinking that the aluminum would be plated with something to provide long-term electrical connectivity, but it seems I still have some research to do.  Brass is, of course, ideal, but there are cost concerns (this will be mass-produced).

Thanks for the advice.

Jeff



Typically aluminum is not plated, it is usually anodized.

Logged
TurboCAD user since v3
 TurboCAD on flickr || My twitter ||


February 15, 2010, 01:36:17 PM
#8
The oxide problem others have mentioned is serious and not readily solved...

Yellow brass or nickel brass (white) would be your best bet.

Thanks.  I have already decided to go with one of the brass formulations.

Now I need to know the optimum thickness and length to make a reliable spring contact that doesn't require a non-conductive support frame.

Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


February 16, 2010, 04:01:43 PM
#9
Okay... I am considering a different clamp design for an AC power receptacle.  This new concept uses two brass pieces soldered to a PCB via the square pins.  The prong from the transformer pushes the springs aside as it is seated, providing the clamping force that keeps it in place.

In practical usage, is the curved piece long enough to allow repeated spring action?  Should I increase the gap between the two springs?

Keep in mind that this is for a Class II transformer, so the power will be less than 10 watts (and usually less than 3).

Jeff


[attachment deleted by admin]

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


February 16, 2010, 04:13:22 PM
#10
You know bud:

I would crack open a bunch of receptacles made my Leviton, (You can get these from home dope-o) and get some ideas. Also your PCB design to hold the clips are going to have to meet UL/CSA listings. I would find some rule books on that, and do some reading to make sure you don't have to start from scratch. Yes, definitely nix the aluminum idea. Plated nickel on steel clips is a consideration too. Frankly, there people that make these kind of parts off the shelf.. might want to google it.

Aram

Logged
1) Tcad V20 on HP Z10 Workstation 3.3ghz-4gb- 64bit- Nvidia Quadro 2000 Win 7
2) Tcad V20 on Lenovo D30 Workstation 2x- Xeon, 64bit, 16gb -Nvidia Quadro K2000 Win 8


February 16, 2010, 04:21:43 PM
#11
I spent hours over two days looking for a NEMA 1-15 receptacle that mounts to a PCB.  No luck.

I am aware of the UL certification requirement (even if I used UL certified receptacles and plugs, the box plugs directly into a wall socket so it will have to be certified).

Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


* February 16, 2010, 07:11:15 PM
#12
Okay... I am considering a different clamp design for an AC power receptacle.  This new concept uses two brass pieces soldered to a PCB via the square pins.  The prong from the transformer pushes the springs aside as it is seated, providing the clamping force that keeps it in place.

In practical usage, is the curved piece long enough to allow repeated spring action?  Should I increase the gap between the two springs?

Keep in mind that this is for a Class II transformer, so the power will be less than 10 watts (and usually less than 3).

Jeff


Not enough info in the drawing. Need another view to even guess whether it's workable. Even then, there's no way I would bless the design without actually playing with a prototype.

Henry H

Logged


February 16, 2010, 07:33:46 PM
#13
Another view?  It's a TCW file, so I don't follow.

What would you need to know to gauge the proper spring configuration for this application?

Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


* February 16, 2010, 07:42:45 PM
#14
Another view?  It's a TCW file, so I don't follow.

My bad. Don't need another view. My eyeball sez the square pins will fail.

Henry H

Logged


February 16, 2010, 07:48:38 PM
#15
I suppose I could round them off, but I am more concerned about the curve radius (and depth) as well as the gap between the two pieces.

Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


* February 16, 2010, 07:59:45 PM
#16
I suppose I could round them off, but I am more concerned about the curve radius (and depth) as well as the gap between the two pieces.

Jeff


Rounding them off won't help, Jeff. Look how small their cross section is compared to the cross section of the main part of your device. The pins will bend long before the rest of the part can deflect enough to admit the component being inserted. If you stay with this concept, you'll have to make the pins a lot more beefy than the "spring" portion of the device.

Henry H
« Last Edit: February 16, 2010, 08:02:20 PM by Henry Hubich »

Logged


February 16, 2010, 10:23:57 PM
#17
I do a bit of research on internet

<> a passive search using google.alert sends related links to e-mail each day <> mostly junk sites set up for adsense, but once a week there is a gem

Logged
Gene Haynes


February 16, 2010, 10:49:35 PM
#18
Oh, okay... I figured that a 0.032" pin would anchor a 0.032" thick spring that wasn't being bent a whole lot.

I'm hoping to stamp the parts for minimum cost, so making the pin thicker than the spring is a problem.  I may have to make a plastic frame to anchor each end of the springs, but that's going to make it a lot more expensive because assembly will be required.


Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


February 17, 2010, 05:46:48 AM
#19
Just a friendly note here, I see a lot of language that I find myself doing before a realize I could calculate the actual forces in the same time that I spend trying to eyeball and convince myself that it's strong enough.  At these points in my designs I generally stop and quantify the actual forces at play and determine exactly what cross section is necessary.  That being said maybe the pins will work, maybe not.  But at least you'll have a reason to believe one way or the other.  Another suggestion may be to not make the pins thicker, because as you say, that's not possible with the process you're attemping, but increasing the number of pins may be possible and will accomplish the same goal.

Logged


February 17, 2010, 09:24:24 AM
#20
I don't have the reference materials to do the math.  Increasing the number of pins would necessitate increasing the length of the clamp, taking more real estate and requiring a larger PCB (increasing cost), unless I can shorten the arc (but that seems like it would shorten the life due to increased bending stress).

If this design is inadequate to the task, then I should either come up with a new design or add a plastic frame to buttress the springs (in production quantities, that should be cheaper than increasing PCB size).

Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


* February 17, 2010, 02:29:46 PM
#21
I don't have the reference materials to do the math.  Increasing the number of pins would necessitate increasing the length of the clamp, taking more real estate and requiring a larger PCB (increasing cost), unless I can shorten the arc (but that seems like it would shorten the life due to increased bending stress).

If this design is inadequate to the task, then I should either come up with a new design or add a plastic frame to buttress the springs (in production quantities, that should be cheaper than increasing PCB size).

Jeff

Does the clamp have to be two separate pieces?

Henry H


Logged


February 17, 2010, 02:33:58 PM
#22
No.  In fact, I was planning on having one piece stamped and bent.

Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


* February 17, 2010, 04:06:46 PM
#23
Jeff
   I have attached info for a Receptacle that might work for your project.
AL

[attachment deleted by admin]

Logged


February 17, 2010, 04:18:32 PM
#24
Thanks, but I've had the specs for this very connector for weeks and even did a preliminary PCB layout with it.

However, it has two flaws.  One is that it is a panel-mount device (no PCB pins or pads), so manual assembly would be required to connect the tabs to the PCB and that adds a lot of expense to the unit cost of production.

The second problem is that it is just so big.  Re-sizing the PCB to accommodate two of them adds nearly $1 per board.  I am trying to keep the unit cost under $8, so I don't have any leeway.

Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


* February 17, 2010, 10:10:12 PM
#25
No.  In fact, I was planning on having one piece stamped and bent.

Jeff


Don't know how this configuration meshes with your requirements, but mechanically it's pretty good.

Henry H

[attachment deleted by admin]

Logged


February 17, 2010, 10:55:01 PM
#26
This looks pretty good.  The prong will enter from the opposite end from what I originally intended, but your design will probably provide a more satisfying sensation when the plug is inserted.

Thanks,

Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


* February 18, 2010, 02:21:20 AM
#27
Not trying to gazump you, Henry, but contacts are usually sprung against the moldings that surround them.  Jeff wants his to be board mounted so he doesn't have a molding to provide opposition.

[attachment deleted by admin]

Logged


February 18, 2010, 09:33:34 AM
#28
This is also a doable design.  Now I've got some thinking to do.

Thanks, Murray.

Jeff

Logged
TC Pro Platinum 2019, 2018, 2017, 2016 & 2015 (all with LightWorks & RedSDK) & V21
System: i7-5820K @ 3.30GHz, ASRock X99 Extreme4, 16GB DDR4-2133 RAM, Gigabyte GTX 970, Samsung NVMe SSD 950 (256GB), Windows 7 Pro (64-bit) SP1


* February 18, 2010, 11:52:25 AM
#29
Not trying to gazump you, Henry, but contacts are usually sprung against the moldings that surround them.  Jeff wants his to be board mounted so he doesn't have a molding to provide opposition.

Yeah, I realize that, Murray. But I don't mind being gazumped, having acquired thick calluses in the gazumporium.

Henry H

Logged


* February 18, 2010, 12:20:45 PM
#30
I saw something that looked like a little cheese grater for removing them when I was obliged to accompany my wife to a beauty product emporium! 

Logged